Classifying Study Designs

Content: Dr. Petitti gave us a "compendium" of study designs in the field of epidemiology and medicine.   It was interesting to learn about the different types and recognize the difference between them.  At first glance I would have put them into one of two groups, observational or experimental, and I would have left it at that.  In my past those are the only two groups that I have been taught.  It was interesting to see the breakdown of these two groups into more distinct and specific groups.  Again I liked the way I was able to see connections with past lectures, specifically Dr. Greenes lecture on August 31st on research methods.  He  concentrated more on what is and what is not research, but he also classified research designs and introduced us to case control studies, and randomized controlled trials. 

Posted by Laura Wojtulewicz

Data mining in BMI

Content:

The two classes of BMI 502 in this week cover the two topics in biomedical informatics study from quantitative perspectives, one is the data mining technology used in computer science, and the other is the technology used in statistics study.

The lecturer from computer science gave us a general introduction to the data mining, and introduce the usually used the data mining methods, which include the classification, cluster, association rule discovery, and regression. From this class, I understand that the data mining is a really inter-disciplinary subject, which covers the statistics and computer science. Actually, the principle of data mining is rooted in the statistics, and implemetation and developing of data mining rely on the computer science study. The principle and basic algorithm of the above methods were introduced briefly, and the practical applications of these method in our daily life are also discussed in some simple examples. Further more, the more detailed discussion in this class focused on the method on classification. The decision tree classification, rule based classification, nearest-neighbor classifiers, and artificial neural network methods are mentioned in this class. Although no further detialed discussion was around each methods in classification, the introduction also initiates my interesting to dig in deeper to learn about how these methods were realized and the particular advantages and disadvantages of these methods.

Because the classification has broad applications in artificial intelligence, and especially can be used in decision making procedure, so I think the classification is close related to the clinical decision making support study in biomedical informatics field. For example, for the traditional clinical diagnostic procedure conducted by clinical professionals, the decision making procedure is actually very like a decision tree classifing procedure. The clinical professional classify the disease of a patient from the symptoms of the patient, and think about if... then the patient probably is ..., and further from another symptom, the if... then... procedure is performed again, until the last conclusion can be obtained. This procedure is actually the same procedure of the decision tree classification. Therefore, the computer can also be trained to learn the rules to build the classification tree model, and based on this model to provide decisioin support in clinical practice.

There is one book about data mining is very classic for get a further idea about data mining method.
Introduction to data mining. Author: Pang-Ning Tan, Michael Steinbach, Vipin Kumar.

Hope it can be useful


Posted by  Di Pan

Week of 09/21/09

Content:

The first lecture was on machine learning.  The difference between supervised and unsupervised learning by machines was covered.  Another topic that was taught was a few tasks that machines use for learning.  Those tasks are classification, clustering, regression, and semi-supervised learning.  Classification can be performed by using a model for a class attribute.  That model is created by using a record training set to find a class attribute as a function of the values of other attributes.  Some applications and algorithms for classifications were explained.  A definition of clustering and some applications of clustering were instructed.  Machine learning seems like a useful but complicated topic.  At what point to set k values in the knn classification method seems complicated.  That complication is from finding a balance between avoiding having that k value be sensitive to noise points and including points from other classes in its neighborhood.  I think the presenter did a nice job at including a diverse number of applications of machine learning in the lecture.

The second lecture covered a variety of study designs.  The importance of knowing about an assortment of study designs was taught.  Knowledge of study designs can help with interpreting studies and the statistics they contain.  That knowledge can also help a researcher to select a study and understand its practicality.  Some key study design dimensions were instructed.  Those dimensions are purpose, a questionnaire's effect on subjects, the subjects' and researchers' views of time, and feasibility.  Classifications of study designs were also explained in the second lecture.  In epidemiology and medicine study designs can be classified as descriptive, observational, or experimental.  In behavioral science and evaluation research study designs can be classified as experimental or quasi-experimental.  Additionally, subtypes and examples of the studies were taught.  The study design lecture was interesting to me.  Using interrupted time series in quasi-experimental design seems like a particularly helpful type of study design to know about.  Researchers using a interrupted time series timeline can avoid spending their resources to collect middle data in a timeline if that data is not useful to them.  That could be a valuable way to save resources in a study.

The Journal of Machine Learning Research offers a large variety of articles on machine learning.  That journal's site can help anyone who is trying to gain an indepth understanding of machine learning.  The site provides an article search engine for anyone who is interested in specific machine learning topics.  The link for the journal's site is: http://jmlr.csail.mit.edu/ .

Posted by:

Nate

Research design

Content:
This Mednesday, Dr. Petitti introduced us classification of research design in epidemiology research. There are three types of epidemiology research--descriptive studies, observational studies and experimental studies. Descriptive studies aim at description, the goal of experimental studies is causual inference, while observational study can study causual relationship or not.

Observation studies can be devided in to cross-sectional studies, cohort studies and case-control studies. The difference between cross-sectional studies and cohort studies is that the former one measures prevalence while the latter one measures incidence. In my opinion, cross-sectional studies are more close to descriptive studies. Usually, cross-sectional studies just state a situation rather than analysis the reason of the situation. The difference between cohort studies and case-control studies is that cohort studies need researchers to on-going monitor the subject which expose to risk factor to ascertain the outcomes, while conducting case-control studies, researchers do not need to wait, what they need to do is to compare cases with controls to find out the causaul relationship between risk factor and outcome.

Experimental studies are very similar to cohort studies, the only difference is that in a experimental study, researchers manipulate variables, while in cohort studies, researchers just wait. Experimental studies can be devided into Post only design and Pre and Post design. The difference between them is that the latter add information of baseline to assure adequacy of randomization and comparability at baseline.

I like the way Dr. Petitti giving her lecture that she compared different types of studies to help us understand the trait of each research design.

Posted by
Jing Lu

Week Five: Machine Learning & Study Design

Content:

Machine Learning-
At first some of the machine learning lecture just flew over my head but after reviewing the lectures and the comments of colleagues, it looks like the lecture finally arrived. To summarize quickly since there is extensive commenting on Machine Learning in the previous posts, machine learning is a field where we make or give a machine the ability to learn. And this is accomplished through several methods that either classify or cluster. Classification methods includes k-nearest neighbor, decision trees, and support vectors. As Xiaoxiao mentions, Mr. Ji mentioned google as a cluster example and I specifically remember stumbling across an article/webpage that actually describes the algorithms of google as well as talks about how google clusters its data. I will have to review my BME Capstone documentation for the link and I will post it in the comments when I find it.

Study Design-
I found this lecture on study design very interesting since she covered the different types of studies that the different fields conduct (even though the studies may be similar they are called different things in different fields). I especially liked the classic studies that she presented since like she mentioned, these studies are older and are much simpler. They provide a very good basis for understanding the different types of studies. What I liked most was the study on salt (intersalt i believe). It was interesting how the study when it looked at individuals, did not find a correlation between sodium intake and hypertension but when it was performed like it was then the points all lined up and provided a correlation that indicated that salt correlated with blood pressure.

Posted by Eric

Study Design Overview

Dr. Petitti introduced the topic of study designs in such a way clearly outlined the purpose of each type.  As with programming, the hardest part of research is actually evaluating the problem and creating a design.  The hard part is never the implementation and testing.  Creating the idea and coming up with the appropriate steps to answer the problem in a consistent way always seems to take more effort.  This lecture defined the characteristics of each type of study design, including causality, time frame, and point of view.  I really liked the fact that specific study examples were provided with each type of design.  This really caused me to take a close look at what separates each type.  Some studies seemed similar in nature, but once we broke down the pieces of the design , they followed the rules that we were given and fit in the appropriate category.  I found myself thinking of my days involved in clinical trials versus my work in post-marketing surveillance and how these two environments contained similar types of data but were structured so differently based on the nature of each environment.  The causality and time frame really define the type of study in this case.  However, not all designs are this clear cut.  I will be taking a closer look at studies in articles and why the researchers chose that particular design.

Posted by Annie

Compendium of Study Designs Review

Dr. Pettiti lectures on Compendiums of Study Designs was helpful in understanding the different purposes of studies. She was a good presenter making the information on different study designs pertitnent and easily understood.  Dr. Pettiti included some well known “gold standard” studies to help us understand the differences. As a way to help me understand and remember the difference, I have tried to write my understanding of each of the study types:

• Descriptive studies – majority of medical studies fall in this category.

• Case – discussing a new or unusual finding about one thing
• Case Series – discussing something unusual or common several cases
• Surveillance – on going observation and monitoring over a period of time

• Observational studies

• Cross-sectional – studying two different findings at one point in time or across the same population (determines prevalence)
• Cohort – studying findings from now into a time of the future (determines incidence)
• Case-control – comparing two groups of population – one with a particular condition versus a group without the condition
• Ecologic – study of populations and epidemiology and the effects of environmental exposures.

• Experimental – applying changes to a person or randomized group of people and measuring the outcome. Also considered a randomized study.

• Quasi-Experimental – used when randomization of the group is not possible.

I attended a conference on Thursday where the presenter was giving data about risk factors for DVT, prevalence with different risk factors and new requirements coming from the government and found myself thinking of these different study types and how they came to these conclusions, what types of studies were performed, how they determined prevalence, probabilities, etc.  It's a lot of information to process but at least some of this is getting applied in my day to day work.

Posted by :  Debbie Carter

Machine Learning and Study Design

One of my teachers once said that computers can solve problems which humans think difficult, but computers are not able to do things that humans can do easily. For example, if we try to calculate 459287 x 979435 by hand, it might take a long time, whereas computers can solve it in less than a second. What if we want a computer to do an easy task, let's say recognize a person? It will take a long time (if it can give solution), but humans can do those recognition thing comfortably. The basic difference between man and machine is the same thing: humans can recognize something (that's need parallel processing) easily but can't perform serial computation where as machine is better at performing serial computations like arithmatic calculations, but they are very bad at parallel processing (must say the designers of computers aren't smart enough to design something like that). Human can classify and infer something, machines cannot. So, the basic difference lies on learning. Humans can learn and store those things in memory for a long time and are able to infer new things. Computers need some technique to "learn" and the procedure that we apply to train a computer in norder to perform some tasks in the future is called "Machine Learning".

There are mainly two types of learning: supervised and unsupervised. In supervised learning, there is something that will "guide" a computer to give the best result. In unsupervised learning, there is no such thing like "guiding". According to the lecture, there are mainly four major topics in machine learning: classification, clusturing, regression, and semi-supervised learning.In classification, we would want to classify things (anything: yes/no, good/bad, healthy/unhealthy and so on). We use several technique to train a computer to classify those things. We first create a model using training data. The model, which is designed based on the training data, is then used to classify new data (test data). In case of surgical training simulator, if a surgeon performs a task, we would want to classify his performance as good or bad (let's not consider fuzzy answers). To evaluate the result, we need to check the results of previous similar cases and notice some key parameters like how it's done, how long did it take etc. Based on the majority of the results, we can classify the performance of the new surgeon. The results obtained from previous surgery cases are referred to as "training set", and the one that we wanted to classify is "test set". There are so many tools that can be used for classification. Few of them, which are mentioned in the class are: k-nearest neighbor, neural networks (artificial), naive bayes classifier, svm etc. Naive bayes classifiers are used in spam filtering purposes (like in spam-assasin). SVM can be used to classify non linear classification problems (like XOR-gates). Neural Networks are mainly used in computer vision to train a model to recognize some parts in images. Classification falls under supervised learning category.

In unsupervised learning, clustering is one of the popular techniques.In this technique, we take observations and put into subsets(clusters) in a way that the observations are similar in some sense (wiki). We didn't go much in detail during the lecture. It would be interesting to know in detail about clustering and semi-supervised learning in his next class.

The next lecture by Dr. Petiti was on study design, which I found very informative. She talked about various design techniques and how we can manipulate the information in an experiment. We went in detail of descriptive, observational, experimental and quasi-experimental studies. The main thing that I understood from the lecture was that experiments are randomized. In fact, experiments mean randomization. If experiments are not randomized, we cannot trust the results from those experiments.  But, randomized experiments are difficult to do for a few reasons. The examples given in the lecture slides were practical reason ("we can't randomize smoking") and ethical reason("can't randomize cocaiine use). The classic studies presented during the lecture were interesting. For each of design study, there was a classic study, and they really made things easier to understand.

Posted by
Prabal Khanal

study design

Content:

The classifications of study design used in Dr Petitti’ lecture are epidemiology, medicine and program evaluation research.


There are three types of studies for epidemiology and medicine: descriptive studies, observational studies and experimental studies. And for evaluation research are experimental and quasi-experimental.

Descriptive studies are widely used in clinical medicine and public health for description. Dr Petitti gave the famous examples of AIDS and HIV.

Prevalence and incidence are two important terms for observational studies. Time is a critical element. So there are cross-sectional study, cohort study and case-control study. Much emphasis was put on case-control study, because it is efficient. You can find cases and controls and ascertain exposure and risk factors at same time. You can find the causal inference of stroke and Migraine headache by asking the survivals of stroke, but not waiting for stroke.

Randomization will affect the outcomes of the experimental studies. Because of its limitations, quasi-experimental design is introduced. No researchers’ assignments of intervention are involved in quasi-experimental design. And other design features are substituted for randomization process. Actually, I am a little confused with time series analysis and interrupted time series analysis. I google it and find that it is widely used for data analysis. There are two main goals of time series analysis: (a) identifying the nature of the phenomenon represented by the sequence of observations, and (b) forecasting (predicting future values of the time series variable). A common research questions in time series analysis is whether an outside event affected subsequent observations. For example, did the implementation of a new economic policy improve economic performance; did a new anti-crime law affect subsequent crime rates; and so on. In general, we would like to evaluate the impact of one or more discrete events on the values in the time series. So I think interrupted time series analysis is much more like to inspect the correctness of time series analysis.

http://www.statsoft.com/TEXTBOOK/sttimser.html#


Posted by  Xiaoxiao

Machine Learning Review

The lecture on machine learning was a new concept for me and rather difficult to follow. I have searched for machine learning and not found anything to help clarify but then searched for the individual components, ie K Nearest Neighbor (KNN) and had some better results.

 

My understanding of KNN is it's the easiest of the algorithms to classify data. The concept is to compare new data with test data which has already been analyzed by its attributes and given a class description. The idea of evaluating the distance is still a little vague but I can only imagine this is based on graphing somehow. The steps to determine KNN are the following:

 

1. Determine the number of nearest neighbors before making the analysis. This is 100% up to you to choose.
2. Calculate the distance of the new data from the trained data (the data already categorized)
3. Sort all of the distances of the K (the number of samples you chose) to the new data
4. Evaluate all of the class descriptions of the K values
5. Majority wins and this class becomes the class description of your new data set

 The KNN type of classification is considered “supervised” learning because you have already evaluated attributes and pre-classified the training data set. A good reference showing an example is located at: http://www.scholarpedia.org/article/K-nearest_neighbor

 

Support Vector involves drawing a line between groups of clusters and evaluating the margin distance of the data from the line. It was not clear how to determine where to draw the line or if a straight line is always draw, etc. This type of learning though seems to be common for speech recognition and hand writing analysis.

 

Clustering is another learning task where groups of data are combined together with similar attributes and then new data is compared to those groups. This is considered unsupervised learning because the groups of data have not be previously given a class label.

 

These concepts are really new and if these are to be understood in much more depth, it would be helpful to go over some more concrete examples in class.

 

 

Posted by :   Debbie Carter

Week 9/21

Content:  The lecture by Shuawi Ji was very interesting, everything took a while to sink in, but at the end it all made sense.  Machine learning is programming computers to learn by previous data.  The learning process can be supervised or unsupervised.  Supervised learning are using classifications and clustering.  The method of classification was really easy to understand.  In constrast, I had a bit of trouble understanding clusters, more specfically where to actually draw the border (choosing a k-value).

Dr. Petitti's Lecture was also very interested and easy to understand.  She gave brief introduction to study design, while giving classic examples. Dr. Petitti really did a great job at explaining the difference between observational study and a descriptive study.  I really think that the methods introduced in this lectures will be very helpful for my term project.  I just have to think how I can apply them.

Posted by P Ortiz

Content: When you are presented with topic, such as machine learning, that is not familiar, you certainly can appreciate how difficult is must be to appreciate clinical lectures, when this is all new. In the current case, it is with great appreciation that I review the blogs of my classmates, who show a significant understanding of this topic, essentially the methods of handling and analyzing data or data mining with various approaches. The article suggested by Annie is instructive in this regard and states that “classification is a data mining task that can use various statistical or machine learning algorithms to assign instances to different classes.” In fact, one of the foundations of clinical decision support is the sequence of machine learning by various approaches. Whether or not the Intern program for clinical evaluation and diagnosis is a form of machine learning or smart computer analysis is not completely clear to me. Greenes in his book documents a number of limitations and states: “most clinical decision support systems in current use do not learn from data and still rely on rule based paradigms.” There has been limited adoption by the biomedical community of techniques for determining data patterns and relationships (Greenes, 2006), which is the purpose of data mining and machine learning. Clinicians do not necessarily understand probabilistic models for generating and analyzing knowledge and we have seen the limitations of the Bayesian Theorems in clinical decision making. Would it be good or not so good to plug data in from different sources and different formats and get out analysis. However, this is highly dependent on the designs of machines and algorithms and programs by us. “Algorithms that can recognize regularities in data and can construct a model” can be utilized. “Data mining techniques are pattern recognition techniques intended to find correlations and relationships in plethora of data.” Greenes goes on to distinguish supervised versus unsupervised systems. The latter represent exploratory analysis without predefined hypotheses or conclusions, unveiling potential “hidden patterns.” These are not applicable to clinical situations at this time, but supervised models can be applied to clinical decision making to develop predictive models from data handling. In this regard, machine learning techniques such as classification trees and artificial neural networks have been most useful, particularly in the latter in critical care situations, i.e. in the ICU. However, from a clinical standpoint most of these models are research models rather than clinical models (Greenes). It is therefore instructive to start to evaluate some of the other models/methods of machine learning, discussed by Mr. Li and start to think about their utility in bioinformatics and also in biomedical informatics. “Machine learning refers to changes in systems that are concerned with artificial intelligence…recognition, robot control, diagnosis, planning, prediction.” (Nilsson, 1996). One wonders if the equipment and paradigms that Kanav is working on in the operating room would be classified as machine learning based on human interaction. Inputs are viewed as attributes and outputs as classes. I do not have enough knowledge on these subjects to determine how raw or organized medical data in different formats can be applied to algorithms that might be created by K-nearest neighbor classifier or support vector machines but I can see clinical application for decision tree analysis and naïve bayes classifier, as well as neural networks mentioned above. A key question is how the hypotheco-deductive approaches and other clinical decision paradigms might be linked to and based on the machine learning and data mining approaches above. It would be useful to hear other’s ideas on this. I suspect that K nearest neighbor and support vector protocols would be focused on unsupervised rather than supervised situations. But Support vector machines do fit data by lines and might apply to supervised situations. Neural networks, as defined by Nilsson, are dependent on non-linear threshold functions, similar to the way that nerve turn on may required threshold elicitation. I am not sure that this simplistic concept means that mathematical model can simulate the nervous system. The second lecture by Dr. Pettiti was succinct and quite useful for me, as a researcher. Our primary focus has been on experimental studies, but the plan is to move into epidemiological studies that are directly related to our animal models of human disease.

Posted by  Stuart

Machine Learning - Research design

Content:
Machine learning - A machine learns whenever it changes its structure, program or data based on its inputs or in response to external information in such a manner that its expected future performance improves. For example, when the performance of a speech recognition machine improves after hearing several samples of a person's speech, we feel quite justified in that case to say that the machine has learned. It is possible that hidden among large piles of data are important relationships and correlations. Machine learning methods can often be used to extract these relationships.

Classification Algorithms

The goal of classification is to build a set of models that can correctly predict the class of the different objects. The input to these methods is a set of objects (i.e., training data), the classes which these objects belong to (i.e., dependent variables), and a set of variables describing different characteristics of the objects (i.e., independent variables). Once such a predictive model is built, it can be used to predict the class of the objects for which class information is not known a priori. The key advantage of supervised learning methods over unsupervised methods (for example, clustering) is that by having an explicit knowledge of the classes the different objects belong to, these algorithms can perform an effective feature selection if that leads to better prediction accuracy.


k-Nearest Neighbor(KNN) Algorithm

KNN classifier is an instance-based learning algorithm that is based on a distance function for pairs of observations, such as the Euclidean distance or Cosine. In this classification paradigm, k nearest neighbors of a training data are computed first. Then the similarities of one sample from testing data to the k nearest neighbors are aggregated according to the class of the neighbors, and the testing sample is assigned to the most similar class.

Support vector machines
Support vector machines (SVMs) are a set of related supervised learning methods used for classification and regression. A support vector machine constructs a hyperplane or set of hyperplanes in a high-dimensional space, which can be used for classification, regression or other tasks. Intuitively, a good separation is achieved by the hyperplane that has the largest distance to the nearest training datapoints of any class (so-called functional margin), since in general the larger the margin the lower the generalization error of the classifier.
----
Various types of study designs were discussed in the lecture. They include-
Experimental Design - Experimental designs are often touted as the most "rigorous" of all research designs or, as the "gold standard" against which all other designs are judged. In one sense, they probably are. If you can implement an experimental design well (and that is a big "if" indeed), then the experiment is probably the strongest design with respect to internal validity.
I think this is a good reference -
http://www.socialresearchmethods.net/kb/desexper.php

A randomized controlled trial (RCT) is a type of scientific experiment most commonly used in testing the efficacy or effectiveness of healthcare services (such as medicine or nursing) or health technologies. RCTs involve the random allocation of different interventions (treatments or conditions) to subjects. As long as the numbers of subjects are sufficient, randomization is an effective method for balancing confounding factors between treatment groups.


A longitudinal study is a correlational research study that involves repeated observations of the same items over long periods of time — often many decades. It is a type of observational study.

Case-control is a type of epidemiological study design. Case-control studies are used to identify factors that may contribute to a medical condition by comparing subjects who have that condition (the 'cases') with patients who do not have the condition but are otherwise similar (the 'controls').

Posted by
Harsha Undapalli

Discussions from week 9/21-9/25

Content: The idea of machine learning is to observe or analyze data by the machine/computer and being able to take decision on that. Two types of machine learning have been covered in the 1st class by Dr. Shuiwang Ji - classification and clustering. in case of classification data has been divided in two sets - training set and test set. Training set is used to build the model and then using test set the model has been verified. There are several ways for classification, among them vector machine is the one that i used for my MS thesis. One part of my thesis was to detect a fall of a patient. i used accelerometer to collect regular and irregular movement of the patient and then using manifold learning algorithm, i tried to project them on a vector plane and then drew a decision boundary to classify fall or non-fall.


Now the part of machine learning that was discussed in the class was clustering. as far as i understood from the class, it tries to cluster the data in terms of shortest Euclidian distance between them. Hopefully we will learn about some applications of clustering in the next class which might lucid the idea more.



Dr. Pettitti's lecture will definitely help me to design an experiment or a study. Randomized focus group is one major element to choose in a study. for example, if I want to study the people who go to subway for lunch, then if target people only from ABC building then I might end up with wrong result. I have to take samples from every major building around this area to get the right observation. She also discussed different type of study designs and how it differs geographically. She discussed few case studies which have been very useful to comprehend the idea. Also she explained two major terms in study design- prevalence and incidence.

Posted by Gazi

Week of 9/21/09

Machine learning illustrates the statistical framework of BMI. The lecture was challenging, but upon review of the the materials, some clarity has been brought to the discussion. I understand from reading that some systems endeavor to eliminate the human component while others incorporate it. And, once again, we can see that Bayesian Theory is sometimes used within the framework. As I always appreciate real examples, I am interested to know that PET scans use cluster analysis in a 3 dimensional image, to distinguish between tissue and blood. I have tried to search for introductory information on the topic of machine language, with little success. Any readings that anyone has found would be appreciated. I will continue my search.

Study Designs by Dr. Petitti was a bit more tangible. Her examples were actual studies and reinforced the designs she discussed. I think she gave an excellent lecture and included some of the most well-known studies. I found her indentification of Kanav's studies as Time Series Design helpful and logical.

Lee

Classification methods in machine learning

This presentation focused on two types of machine learning-classification and clustering.  Machine learning uses algorithms to recognize patterns and make decisions based on data.  It is used in a wide arrary of fields-from statistics to artificial intelligence, and applied in many processes like object recognition, bioinformatics, and natural language processing.  Clustering techniques definitely  have a use for me, in regards to organizing genetic/genomic data.  If anyone has any resources on clustering techniques, I would be interested.

I found a really good article that compares different classification methods, as it is applied to clinical data:
http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=2232569&blobtype=pdf
Trying to extract meaningful information from narrative fields in medical databases can be difficult.  This articles compares different classification algorithms/methods, such as rule generation, decision trees, and Bayesian classifiers, when applied to the output of a natural language processor.  Extracting meaningful medical information was an ever-present issue in my last job, so it was interesting to see in which areas different algorithms performed better.  Overall, the article gave me more insight into mining narrative text.

Posted by Annie

Machine Learning

Content:  I am always glad to hear a lecture that starts to connect ideas together for me.  Like Ashutosh pointed out, this lecture linked to Dr. Shortliffe's lecture on decision making.  He used an example of a decision tree, which gave us another example to build upon when learning about classification algorithms.  This was a new topic for me, so any connections I could make were a bonus.  Mr. Ji's lecture was a great introduction into machine learning and I learned a lot, from the difference between supervised and unsupervised, between classification and clustering and many applications of machine learning.  
The only area that I was a little fuzzy on was that of support vector machines,  I know that he was trying to leave the math out, but it might have been easier to see how they choose a boundary if after the margins were defined an example with real numbers was done. 

Posted by Laura Wojtulewicz

machine learning

Content:
On monday we had a lecture on machine learning by Mr.Ji. Since I am from non computerscience background I kept my eyes and ears open to understand the tpoic. I learned a lot of things in lecture but I was still not clear about the whole concept, so I used web and tried to relate what I learned in class. Here is what  I managed to understand.machine learning is defined as changes  in system that perform recognition,diagnosis,planning, prediction etc.so that we can get a result as desired.Its main goal is to turn data which is given as an input into output.After learning from collection of data we want machine or the system to answer questions based on that data.The machine turns data into information by extracting patterns from the data.
A machine learning data can be supervised or unsupervised. Supervised data have labels that goes with the data features, while unsupervised data donot have labels.When the data has categories as labels then we are doing clssification and when the data is numeric we call that we ar doing regression i.e. we are trying to fit a numeric output given some numeric input data.When we have a data without the labels and when we want to check to which category the data falls then we do clustering.The goal is to group unlabelled data which are 'close' to.
The lecture was mainly based on three types of classification algorithms.I thought that k nearest neighbour was simplest among three and the use of duck examples made it easy to understand. What i think that we look  for maximunm likelihood. The second one being the decision tree algorithm which is a kind of decision support tool which helps in modelling a decision and its consequence.Please correct me if I am wrong , when Dr. Shortliffe discussed about the example of a 95 yr patient who had a tumor in lungs in the decision analysis lecture , I think the decision analysis for this patient was done through a decision tree. The advantage of decision tree is that it is simple to understand and interpret and it can easily be combined with other decision techniques.The third was Support vector machine which finds boundry to seperate data types .however I have not  fully understood the concept of support vector machine but I hope that maybe in his next lecture I can have better understanding.I look forward to his next lecture.

Posted by
Ashutosh Singraur

machine learning

Content:

In the class on Monday, we learned some general concepts of machine learning. Machine learning considers the important questions of how to make machines be able to “learn”. I think it is just like teaching a little child to learn. The instructor focused on two of the machine learning tasks-classification, clustering and classification algorithms.


Last week, my roommate asked me a question about endoplasmic reticulum: what organs of the body are abundant in ER? I remember that ER is related to production of protein, and only endocrine system produces protein which composes hormone. We can compare the process of solving the ER problem with the machine learning. The training set records for machine are the biology knowledge I learned in high school and first year of my college; finding the model for classification is the way I remember the knowledge; the result that I make sure it is endocrine system reach the goal that previously unseen records should be assigned a class as accurately as possible.

For the nearest neighbor classifiers, Dr. Ji gave an example of duck which is easy to be understood. As to the k-nearest neighbor model, in my opinion it is based on an assumption that like attracts like. The way to determine the unknown record is the distance to its nearest neighbor, so it is classified according to its neighborhood.

Decision tree is more logical to me. The method to determine the unknown record depends on other attributes. So it is the characteristics that label the class. And the main technique for decision tree is that how to build a tree with attributes.

Support vector machine algorithm has more mathematics. The simple way to explain it is that to find a decision boundary to separate different data. It is not so hard for us to do so because we have intelligent brain, but for machine you have to teach it everything.

Dr. Ji just gave google as an example for clustering. But it is enough to understand that cluster is found by similarity measure among the given data which have a set of attributes.


Posted by

Xiaoxiao

Machinelearning

Content:
This Monday, Dr. Ji introduced us the concept of Machinelearning and several algorithm of classification, which are K-Nearest-Neighbor classifiers, decision tree, and support vector machine. Machinelearning is to state problem of "how to make machine to learn". In my opinion, the process of machinelearning can be devided into two steps: inference and decision. In the step of inference, we need to make machine to build a model to explain or classify the given data, while the step of decision is comparatively easy, which is to give a response to the given data based to the model.

Dr. Ji put great emphysis on the introduction of three classification algorithms. Classification is the process of recognizing different datus.

K-Nearest-Neighbor classifier might be the simplest one. Maybe that's why it is called " lazy learning". And this algorithm is also very easy to understand. It is to classify an object based on its closest training examples. According to my understanding, the core idea of this classifier is that: We imagine that each feature of both traing and testing objects are different "dimension", our objects are in "N-dimensional space". If an object is close to several(N) other training examples in this "N-dimensional space", then this object should be regarded as of the same class of those training examples.

Decision tree is even easier than KNN classifier to understand, especially with a graph given by Dr. Ji. Actually, I use this algorithm in my daily life. In some cases, when I clean up our olding things, I would use this algorithm to decide which things can be threw away while which should be kept. But I think it is much more complex to apply decision tree in machinelearning than to understand, as there would be a lot of specific problems such as at which point should split the tree and which attribute should be the root.

I don't know why support vector machine is called this name. It seems that its algorithm has nothing to do with support vector. The core idea of this algorithm is to maximize the margine. Though the figure given by Dr. Ji is two-dimensional, this "margine" should be the margine in N-dimensional space, and N is the number of features.
Posted by
Jing Lu

Cost effectiveness

"Can ethics be monetized?" was the first question when I was going through the lecture slides and the note that I wrote during Dr. Johnson's lecture.

The lecture by Dr. Johnson was an eye-opener for me. This was the first time that I was attending an economics lecture. Dr. Johnson hover around the basic (but very important) concepts of economics. The first lecture started from cost-effectiveness. It's something like "is it worth buying at this price" rather than "is its price the cheapest?". He also mentioned that economics doesn't give any answer for acquity or income distribution, and also that ethical values don't affect economics.  Direct, Indirect, Opportunity Costs were discussed in detail. What is the difference between cost and transfer was very important. Perspective was introduced then. What I do might not be worth doing for others. The evaluation of cost effectiveness should always be done from a single perspective throughout a study. The examples given by Dr. Johnson were simple but effective. Then I realized that economics is everywhere. The first lecture was mainly on "who pays the cost and who gets the benefit".

The second lecture started from where the first lecture ended. There was discussion on transfer, efficiency, benefit, discount rate, and reservation price. I still remember the joke cracked by Dr. Kanav, which nobody understood at that time except himself. I went through the NY times website and I found it there. The difference in theories from Keynesian vision and Neoclassical purists. The article was worth reading. I still have to go through the terms that were mentioned in the class in detail. But overall it was a nice exposure to the field of economics.

Posted by
Prabal

Economic Evaluations in Healthcare

The topic for this week was a necessary one.  After both presentations and discussions from Dr. Johnson were complete, I was left with a structured outline showing how to evaluate different variables in healthcare.  I was really trying to over-analyze the situations presented to us in the beginning because that is the way I have been trained.  I realized that the concepts should remain simple and logical throughout the process and should serve as the foundation of the evaluation.  It is only then that more details can be added to the complexity of the problem.  Overall, the process of analyzing an economic situation has few rules--to maintain a perspective, define your variables, and appropriately measure your outcome.  I think I have had the most trouble identifying the perspective in some situations where the perspective does not include individuals that I thought would be included (government, taxpayers, etc.).  I think that will probably just make me take a closer look at perspective in future readings.  The exercise given was an appropriate way to get us all thinking about the economics behind what we are studying. 

Posted by Annie

Week Four: Economics

Content:

During this past week, Dr. Johnson has re-stimulated parts of my brain that I haven't been consciously using in the past years after i finished my sole economics class. Examples of such applications that I've been using over the years have been opportunity cost, marginal benefit and cost to purchase products both online and in stores. The only part i didn't really apply was term to the concepts.

More importantly, I've found that Dr. Johnson really helped to teach the concepts of economics by allowing us to break economics down to its basic form; no ethics, no opinions. Just plain old economics and its math-like nature. The part I found most interesting is that simply the distinction of perspectives can change the meaning of everything. It can change whether an event or occurrence is a benefit, cost, or simply a transfer.

I also found it interesting during Monday's class that he spoke of how the income is distributed and how it should be ideally distributed (I believe he said there is no right answer).  I did a quick google search of the L-Curve that represents the current distribution that shows the distribution in a context we can visualize.

http://www.lcurve.org/

As well from the reading and some before class discussion I took notice of the word "capitation" and wasn't too sure what it was or meant so I looked it up and PBS has a great article that explains it in detail just in case someone else was curious as well.

http://www.pbs.org/wgbh/pages/frontline/shows/doctor/costs/wonderdrug.html

Posted by Eric

EMR and economist!

Content:

The two lectures by Dr. Johnson were just magical. It made me pondering for the 1^st time why EMR being the “next big thing”, yet not getting enough positive support from the hospitals or doctors. From the discussions I was almost convinced that it is the impossible thing to implement EMR by looking at from the economical point of view of the hospitals. Truly they don’t care about the patient’s shake, profit is what they care. We cannot blame them as in this verge of economic downfall nobody will implement an expensive system which apparently involves patients safety, not enough bang for the bucks for the hospitals!

The lectures proved me wrong when I answered to one of the Dr. Johnsons question that wearing helmet can be optional, not mandatory. For example, I used to always speak for laws that prohibit using cell phone while driving, but not buckling up mandatory. Because the former involves other people’s safety on the road where the latter involves the life risk of the individual who chooses to do that. However, I did not think of the fact that if any accident causes severe injury to the driver for only not wearing the seatbelts (which is very obvious) and the stupid driver end up staying on life support for the rest of the life, his entire medical expenses is bore by the public medical insurance which uses the tax payers money to pay for it. So end of the day it incurs on the general people like us!

Now, it took me two entire lectures to realize the consequences of not wearing seatbelts. But the policymakers of AZ still didn’t pay attention hence wearing seatbelts is still an option! I wonder what it would take to convince the hospitals implementing EMR considering only public safety unless you make them all sit in the lectures given by economist like Dr. Johnson!

Posted by Gazi

Week of 09/14/09

Content:
Both lectures this week were on economics in Biomedical Informatics.  One of the topics covered was methods of analyzing costs and benefits.  Three of those methods were benefit-cost analysis, cost effectiveness analysis, and cost utility analysis.  Another topic covered was the steps that should be preformed to conduct an economic evaluation.  Some economic terms were defined such as the term perspective.  Perspective refers to which group of people's perspective the costs and benefits are occurring.  Some of those perspectives are of societies, governments, and health insurers.  Another term that was defined was opportunity cost.  Opportunity costs are the values of alternative resources that could be used instead of a resource.  Also, discount rate was a term discussed.  That rate refers to the difference between the future and present values of a resource.  Additionally, criteria that should be applied to reviewing articles was taught.  Specifically, applying that criteria to studies of health IT including EMRs was covered.  A article by the authors of Menachemi and Brooks analyzed health IT studies and suggested that there are relatively few health IT ROI studies because of the difficulty in estimating ROI.  An example of that type of difficulty is that in those ROI studies benefits can be harder to monetize than costs.  Due to that difference in monetization, benefits can be underrepresented.

I found the definitions of the economic terms to be interesting.  Opportunity costs are particularly interesting to me.  When I am thinking about buying something I sometimes consider that I like what I am thinking about buying but there may be a greater use of my money.  Researching alternatives has made a huge difference for me in the past in terms of understanding variations in products.  It has also made a huge difference for me in anticipating the large number of ways in which those products can be used over time.  Here is a web site that has given me valuable deals in the past and helped me with researching computer component purchases:  http://shopper.cnet.com/ .  In addition, the lectures also intrigued me to think about how to classify complex financial relationships between societies, governments, and individual citizens.  Furthermore, the homework articles impressed me in their accounting for the large number of types of costs and benefits that can exist in products.

Posted by:

Nate

Content: We have had two very stimulating lectures from Dr. Johnson and the perspective of an economist is different than a medical perspective. Key points include trying to find the most cost effective solution, which is not necessarily the cheapest alternative. I will now be able to plan my vacuum cleaner purchases in a better way. Cost benefit analysis is monetized by cost efficiency analysis may depend on non-monetized benefits. Driving from Tucson, twice a week, certainly emphasizes how much safer it might be if the speed limit was reduced. Each day, I see a lot of crazy drivers and I am concerned that they have a shortened life span. I have now focused my attention on not taking the 10 through Phoenix. I have become acquainted with the Hispanic and native American community on Avenida Yacqui which is actually longer in distance but shorter in time than taking the highway. Soon, I will consider stopping there for dinner, but I will have to review the safety, i.e. utility, of that decision. From a cost standpoint, I am spending 25% less on gas by going slower through Avenida Yacqui after taking 7th St. down instead of the 10 all the way to Tucson. This is my own personal perspective, but I am not sure that society would view this perspective differently. Initially, I would think that driving by a longer route in distance at a slower miles per hour, would have a higher indirect cost for me due to greater time. However, the cost of gas reduces my direct cost and the time during rush hour is not different. My opportunity cost is equivalent. Nevertheless, the indirect cost of driving no matter the route is significant from a time standpoint and indeed has a economic cost.




I have talked previously about the dilemma that I have with my present EMR, upon which my practice revolves. The software will not pass muster in the future based on the federal standards, but in February, despite 6000 users, the company stopped technical support and soon after, when there were complaints, stopped the software internet forum. Recently, with I suspect, a huge exodus of users, who did not want to upgrade for $20,000 to the company’s new software, the forum has gone back live and technical support has again been able to be reached. Nevertheless, the inducement to go to the better software that can not transfer my existing templates (at all) and notes (fully) and ? my data including all labs is difficult to balance against the promised transfer cost to me of $44,000 should I switch to their new software. Will I be able to net $24,000 starting in 2012? If so, What do I do with my current records if technical support ends in 2010 and follow my patients after this? In this case, there is a transfer payment, but my opportunity cost and indirect costs exceed this possible payment, particularly in terms of my ability to deliver quality care to my patients. In essence, I have a penalty for adopting this standard of EMR, before it was mainstream. The profit is significant to the EMR manufacturer but perhaps, they have underestimated, the proportion of the 6000 users that will not upgrade and will go to another supplier. Interestingly, this vendor is the vendor that provided the software that created an error and diminished usability and error checking that was discussed by Dr. Patel with one of her cases related to excessive potassium chloride in the ICU patient. Great. Sell the software and do not support it. I guess the concept that is so popular among American car dealers is alive and well in the EMR industry for doctors and hospitals. Unfortunately, the government provides this subsidy to physicians at the taxpayers expense for inducing me to purchase a program that does not allow me to take care of existing patients and these patient, who are the taxpayers suffer. This subsidy is not a neutral situation from a societal perspective. It is actually an indirect cost now. Further, the cost of the software, when initially purchased now has deteriorated from multiple 1000s of dollars to a value of zero—a real decline in its original present value.



Stuart

Posted by

Economics in health care - Health Economics

Content:

Firstly, I would like to thank Dr.Johnson for his series of lectures which envisioned me the scope of economics in health and health care. To me, Economics is how efficiently resources are utilized in a given place. Economics guides the rules which explain the behavior of people who pursue (and compete with others for) the limited resources (goods, currency, health, etc.) within a society.Many of the healthcare economic principles focus on how people make decisions related to expenditures for the health given competing alternative (e.g., food, clothing, housing, hobbies, travel, education, etc.).

Micro-Economic evaluation of a given health care system can be done using - Cost Effectiveness analysis, Cost Minimization analysis, Cost Utility analysis, Cost Benefit analysis. The workbook gave me a good understanding of these types of financial analysis. I accomplished a broad overview of various types of costs and benefits that define a health care event, various perspectives and concept of discount rate from the lecture. I have never got a chance to explore various cutting edges in economics and hope this lecture would drive me further in knowing much more about economics and health care.
Posted by
Harsha Undapalli

Economics

Content:
I took economic classes in midle school, high school and college (such is Chinese education that students learn everything, and give back all we learned to teachers after final exam), so the terms and concepts in this class are like old friends to me. But maybe I haven't seen them for such a long time that I hardly recognized them,  let alone they suddenly spoke English to me.

I really like the way Dr. Johnson giving his lecture, except that sometimes he spoke too fast to tell a joke, and I was really confused with waht the whole class were laughing about. I like the way that pre-class readings are given before lecture. In this way I would think about a lot of problems before class, and listen to class with questions.

I think it is very important to imply Economic methods in other fields. Economics makes people behave wisely. Take BMI for example. I believe that the maximum of economic benefits is not the final goal for most healthcare organizations and governments. But economic evaluation is still necessary before making any decision. Because healthcare resouses such as fundings, physicians and equipments are limited,  we need economic evaluation to allocate these resources, to achieve maximum effectiveness. Moreover, I always think people who know some economics can arrange their life more rationally, and more effectively. Dr. Johnson maybe an example, as he described himself.  I can not agree with one of his idea more, which is that we should make our decision only based on our current situation, and anything happened befire should not affect our decision. His example about healthcare equipment nicely illuminated  this idea and the necessity of  economic evaluation befor decision making.

Posted by  Jing Lu

Economics in EMR

Content:
In this week’s two classes, Dr. Johnson gave us two excellent lectures about the economic factors that influence the implementation and operation of health informatics system in health care practices. He introduce the basic economic terms involved in cost-benefits analysis on EMR system in hospital, such as the perspectives, measures of cost and benefits, the opportunity cost, the discount rate, and the analytical approaches commonly used. This is the first economics related class that I have ever officially taken in my life. But Dr. John’s humorous style and simplification of abstract problem to practical daily examples make me not feel so hard to understand the economic points involved in implementation of health informatics technology in health care field. The most important point I got from his class is that “who will get the benefits, who should pay”. I will talk about my understanding and the practical issues related to this point.

For this question, I first need to set the scope of the analysis from hospital perspective. From the hospital perspective, the hospitals get the most directive benefits, the money, from implementation of health informatics technologies. Such as many cost-benefits analysis of EMR systems in hospital illustrated, the hospital will get the increased money resulted from the helpfulness of EMR. Then the patients indirectly also get benefits from the increased quality of health service from hospitals after the implementation of EMR. Furthermore, the higher quality service received by patients would induce the patients more prefer to choose this hospital next time when he/she need, and thereby this hospital will earn more money compared to before implementation of EMR. Therefore, based on the above analysis, I think the hospitals are the main beneficiary from implementing EMR. So payment for the cost of EMR system should be paid by health care providers, and the monetary benefits they can earn is the direct incentives to thrust the implementation of EMR.

Posted by  Di Pan

economics

Content: I took macroeconomics about four years ago, and I remember that I really enjoyed the class, but sadly I really don't remember much of the concepts.  I really liked the lectures given by Dr. Johnson, he made really interesting, and he also emphasized how important economics is to BMI.  I like especially his vacuum cleaner example, and think it really summed up everything.  I have to admit, everything took a while to sink in. The concept that I have the most trouble with is putting the ethics aside.  It makes perfect sense why its done , but I really bothers me. Another part that is really confusing to me is the perspective  I think this is a very interesting topic and I wish we could go more in dept.  As for the assignment I think I've read all the articles about 20 times and every time I get a new insight.

Posted by P. Ortiz

Economics Lectures

Content:  Like others in the class, this was my first experience with economics in school.   I was so gratefull that we had the luxury of having Dr. Johnson speak two classes in a row, it would have been overwhelming to have all that information in one lecture and it wouldn't have been as effective to have the classes separated by other lectures.   I enjoyed the lectures and I think I am finally understanding the terms and concepts.  The most challenging term for me still is perspective.  The idea seems so simple, yet there are so many underlying issues that I want to throw in to complicate it.  I still don't understand how tax payers aren't taken into consideration for a societal perspective, but perhaps I can figure that out when I look into that further with the resources that you guys have posted.  I think the assignment is perfect for us to see if we can apply these terms and use them to analyze a Cost-benefit analysis.  I too like Debbie, have re-written mine, from prior to the lectures, and again after the first lecture and yet again after the third lecture, which I think is a testament to how much I learned in each lecture. 

Posted by Laura Wojtulewicz

Some More References-Kanav

Content:
for those really interested in drilling this down.. this is the WHO guide to Cost Effectiveness Analysis
http://www.who.int/choice/publications/p_2003_generalised_cea.pdf

Also dont forget the friendly NY Times introduction at
http://www.nytimes.com/2009/09/06/magazine/06Economic-t.html?_r=1&scp=1&sq=economist%20get%20so%20wrong&st=cse

Also think of what I could ask in this section

Posted by Kanav

Economics in a week

Wow - what a whirlwind of thought provoking discussions.  Just when I thought I understood the concepts, the rules and lines kept crossing.  But, given all the information we heard and tried to absorb, the importance of what is being taught can't be over emphasized.  As we deploy electronic health records, the requirement and push to implement is coming from the government but health care organizations who have not pre-planned for this time are now faced with many of the evaluations, calculations and analysis on what and how much to buy.  So, not only does informatics look into data and hopefully help us improve outcomes in healthcare, it also expands into multiple areas including economics.  Vendors will sell you as much as you want but not necessarily help guide you in the appropriate selections for your needs.  How much is enough?  Where do the benefits and costs cross lines and it becomes more costly for less benefit?  I will have to investigate the cost - efficiency analysis methods as Dr. Johnson pointed out because this is where health care organizations are always looking for answers.

The assignment for this week on evaluating the article has had me rewriting it twice now.  I can't decide if the study was good, fair or bad.  You can argue several points in the article but what I have found in second guessing myself is learning more of the terms and thinking more about the concepts.  Probably the most important concept, in my opinion, is a consistent perspective.  Even when trying to write a simple paper, keeping the perspective the same can be challenging so I can understand how studies mix and match different perspectives but now truly understand the dangers of reporting data under differing perspectives.

Dr. Johnson is so correct in his statements about the profession of informatics making the future decisions on electronic health records.  Of course we can't be experts in all aspects of the field but just understanding the complexity of evaluating an electronic solution, evaluating what is needed and then making sound business decisions starts an organization off in the right direction.

Posted by   Debbie Carter

Economy evaluation

Content:

Professor Johnson has given us three lectures. Though I get lost in the economic concept and his questions sometimes, anyway I like this wise man.

Actually I just read the paper for assignment before the class, while Dr Johnson introduced a lot of knowledge in the pre-reading. But I caught the word discount which is included in the assignment question that I could not figure out, then I listened carefully and totally confused at last. I discussed with Jing and she gave me an example: when she was born, her mother bought her a kind of marriage insurance for 2,000 which promise that she will get 10,000 when she get married. But the 10,000 is much less worthwhile than that in 20 years ago. I got the difference of value at that time. My father is a Ph.D in economy (I am a little down that I do not inherit his talent in economy), I asked him for the discount and he gave me a number line.
In a way, we can consider that the value of 2,000 and 10,000 are equal, but the amount is different. As the time goes, the amount will become bigger as a result of inflation. When you look from the view of 2,000, the difference is discounting; when you look from view of 10,000, the difference is interest.

And I read the materials carefully after class, it is really helpful. The concept, measurement and evaluation will help to give a comprehensive thinking. I quite agree with Dr. Johnson that economy will not tell you how to use money. It is an evaluation of distribution.

I also like the way that a professor provide two classes in a week for us to understand a topic more. I like the pre-readings and the reviewing in the class.

Maybe I can take a course of economy next semester.

Posted by Xiaoxiao

Economics and Ethics

Content:One of the most enjoyable lectures which I will remember throughout and even plagiarize parts of the same :)
I must confess though that this is the first lecture on economics I have ever sat through and actually enjoyed.
The most intriguing part of the lecture for me was the separation of economics and ethics. Dr Johnson simplified ideas in order to get the concepts right at which point mixing theories was not a good idea. However I think in the real world separating economics from ethics doesn't seem logical. A google search of economics and ethics led me to this webpage http://www.nd.edu/~cwilber/pub/recent/ethichbk.html
which discussed John Neville Keynes argument of dividing economics into three areas: positive (economic theory: dealing with what is), normative (welfare economics: dealing with what ought to be), and practical (economic policy: applying the first to the second). This is putting this entire argument in a nutshell.
The article argues a very good issue of surrogate motherhood which ends up using children as a commodity.
Interesting !!!!

Posted by Sheetal Shetty

Economics

Great lectures by Dr. Johnson. He certainly has a talent for “encouraging” class participation. A few particulars were drilled in. Cost effective is the most efficient, not necessarily related to pricing. I once believed that quality (not value-I was too smart for that) might be reflected in cost. Maturity, the recession, and this lecture, changed this perception. In addition, what non-economist would think that things like welfare payments were merely transfers, from a societal perspective.Then, there are opportunity costs, some we are not willing to pay. What students agreed to reducing the speed limit to 10mph to save lives?
The article by Krugman, was also an interesting read. I can only trust the credibility of the author (2008 Nobel Memorial Prize in Economics; refers to his NY Times blog as “The Conscience of a Liberal”), as this is referenced reading from Kanav. He gives us a historical perspective of the evolution of economics, discussing neoclassical economics to the birth of Keynesian theory following the Great Depression, a return to neoclassicism, and adoption of monetarism, to New Keynesians, with mention of behavioral finance (does cognitive science have a role in every application?).The author also provides a plan to "re-embrace" Keynesian economics, given the fact that recessions and depression cannot be denied or ignored.
I am leaving this week with ever more to consider. I did not approach BMI from an economic perspective. I am painfully aware that economics is a factor in most every healthcare decision. However, BMI is clearly a global science. Everything must be considered and evaluated, perhaps finances as much as anything.

Lee B.

Economics in BMI

Content:
To be very honest I never expected that I would be taking a economics class in BMI.I had never taken economics subject in my whole student life but after attending those lectures I really find it interesting.Dr.Johnson's lecture revolved around cost effectiveness.In the very beginning he cleared a misconception that cost effective does not means the cheapest but the most efficient.I did carried this misconception for a long time.The other very important thing that I learned today was about perpective.For doing any study,reaserch,survey the perpective should be well defined.What is good for me can be evil for others.Now i realise that this word has a huge impact in economic studies and that is why he was putting so much emphasis on this word.He also defined "cost" and also talked about different types of costs and lossees which to be very honest were quite difficult for me to understand and relation between them.Sometimes I also got confused and tangled to realte an entity with a individual, society,services and thinking about different perspectives.I know that I have dig my head and try to learn as much as i could so that I could write something constructive in the assingnment.I hope for the best.
Posted by
Ashutosh

Economics and its importance-Kanav

Content: Dr Johnson gave an excellent series of lectures and personally for me Economics is a very interesting and important subject.  A true story. In India while preparing for my glowing engineering career, I chose a path of studying Physics Chemistry, Math English and I had a choice of studying computer programming or economics. I thought Ill take economics as if I didnt become an engineer I thought Ill become a MBA and earn my bread that way. As it turned out the whole of 11 grade and 12th grade I didnt pay any attention to economics just reading enough to get by with a minor A or so. It was only in 12th grade towards the end, that I had to take it seriously to get good marks in my exams so I could get into some colleges. So in 15 days or so I read the entire text books of 11th grade and 12th grade. It wasnt my ability that made me do so but the sheer power of the subject and the implications that economics had on my day to day living. I was completely bowled over by macro-economics and the fact that too much supply can be an issue bewildered me till i understood it. Frankly it was mind blowing stuff!

I did end up scoring highly on my exam probably because I really liked it and then havent touched the subjet till now. However i think of it as being extremely important and something that everybody has to know. Healthcare industry was not designed to be an industry and mostly we have very little idea of how it works. This fact was reiterated by Dr Johnson who pointed to the fact that most papers published in Health journals on cost effectiveness are bogus. I couldnt agree more. In fact some people running the show could be very dangerous as they have only psuedo knowledge of the subject but know enough words to scare us all. Anyhow I hope that you guys follow up on this subject and read more. Trust me from personal experience it is an engrossing read. In fact I just started re-reading Keynes work on the General Theory of Employment, INterest and Money. Fabulous read!

For a little lighter read on the subject see this article published in NY Times Sunday two weeks ago.

http://www.nytimes.com/2009/09/06/magazine/06Economic-t.html?_r=1&sq=economist get so wrong&st=cse&scp=1&pagewanted=all

 Written by Paul Krugman it talks about the importance of returning to Keynesian economics and also explains in laymans terms what happenned in the current economic crisis. Think about projections to the healthcare industry too and how we can avoid some pitfalls.  Further you will find ways to expand your knowledge. Expect a lot of questions in the midterm on this guys!

Posted by

Kanav

HCI

Content:
The example I want to talk about is an e-dictionary which benefits me a lot in the English learning. It is a computer application with a small and simple interface and could translate Chinese to English and English to Chinese whatever your input is words, sentences or paragraphs. It can also web search if the word is not in its database. What I like most is the screen words function: if I put the cursor on the word for a while (less than a second), the translation of the word will show up immediately, so that I do not have to input every word occurred in reading.

Pros:
Compared with paper dictionary, it is more convenient and efficient for translation.
Compared with other e-dictionaries, its highlights are screen words and web search function.
Portable, if you have your laptop by your side

Cons:
No universality, benefits special groups of people, such as Chinese who is learning English and English-speaking people who are learning Chinese.
Because of culture differences and phases variation, it could not translate correctly all the time.

The discussion shows that despite the human-computer interface is perfect and the search algorithm is speedy, there are still many deficiency existing in the e-dictionary. The problem domain is not computer factors or human factors, and it addresses in culture and linguistics. So, I am more into the statement that human-computer interaction is an interdisciplinary area. Besides human and computer factors, it covers several fields such as industrial engineering and cognitive psychology.

It is no doubt that computer technology is the most important element of HCI. Through computer technology can we achieve the goals that usability, universality and usefulness for human. And the center of the design is human, so it is important to do requirements analysis. Because human-computer interaction involves transducers between human and machines and because human are sensitive to response times, viable human interfaces are more technology-sensitive than many parts of computer science.

Posted by Xiaoxiao

HCI definition

Content:
This class reminded me the feeling of confusion and stress the first time I use my first computer, under the OS of dos. Time has gone when people are afraid of computers. Now using computers is as easy as using televisions, thanks to the researchers and engineers in the field of HCI. I am very happy that Dr. Kanav provided me with the opportunity to be exposed to this field. In my simple way of thinking, HCI is to study the way of communication between human beings and machines. HCI is so important because it can affect effectiveness, productivity, moral and safety.

In the reading material given by Dr. Kanav, HCI is defined as “a discipline concerned with the design, evaluation and implementation of interactive computing systems for human use and with the study of major phenomena surrounding them”. This is a long sentence to read. But I don’t entirely agree with this definition. This definition of HCI puts great emphasis on machine, while in my opinion, the central component in HCI should be human beings. The definition given by SIGCHI may be correct for the early stage of HCI development. But now, more and more designs take psychology and anthropology into account. For example, when we flip the screen of e-reader, we expect to read the next page, as this is the natural reaction. Another example, when we shop online, we put items into a shopping bag before checking out, just as we do in a shopping mall. In this sense, HCI is not only about the communication and feed-back between human and computer, but is also concerned with human behavior and habits.  What is more, we also communicate with other people through websites or chatting software, such as posting comments on blogs. I also find that human-human communication on the internet is designed to simulate that in real life. For example, people’s homepage are designed as there “home”, and friends can visit this “home” just like they do in daily life. In this sense, the design of HCI has extended to human-human interaction rather than only between human and machine.
-------------------------------------------------------------------------

I’d like discuss the pros and cons of a GPS navigator.

Pros:
Portable
Universality: It guides me to my destination with voice prompts, arrows on the map, and directions at the top of the map.
Hands-Free phone feature
Easy to use: touch screen

Cons:
Cost
Map is not updated daily. It does not take into account situations such as road construction.


Posted by Jing Lu

HCI WEEK

Content:
I believe Dr.Kanav introduced the vast doamin of HCI in a very simple ,lucid and understandable manner.I found one more definition of HCI from SIGCHI which says "it is a descipline concerned with design,evaluation and implementation of interactive computer systems for human use and with the study of major phenomenon around them." I think it preety much sums up  about the whole concept of HCI.The three components of HCI are Human,Computer and their interaction which is very obvious.Humans have limited capacity to process to information which actually put implications and constraints on the design.Information is recieved by sensory system,stored in memory and processed and applied by reasoning,problem solvingand also errors.All the users share the same capabilities with a common differnce. A computer system  also comprises of many elements. The interaction models translates between what the user wants and what the system does.
Interface is also a key aspect of HCI.There are seven basic principles in creating user interface design which you can look in wikipedia. Modern design models give user the center stage in desinging of any interface.It has become essential for designers to observe users at work and developing a understanding of how they interact with systems.It is very important because as Eric mentioned in his blog that a good HCI for any product will attract userswhile the bad one will result in loss of users.
A mobile phone is a very common and handy product of HCI.I would like to share my observation that I made during my days in India.In India the usability of touch screen phones like the "iphone" is still not widespread.People normally use phones which has controls via keys or buttons. Although these phones have all latest features like high definition graphics,3D games,camera ,music player,web service and many more.All these features are operated by keys. Although the features of phones of different companies are more or less same but their key controls and key positions are different.As a result the position of "enter" key of one phone turns out to be the position of "delete" key in other.As a result the user gets frustated and they are unwilling to change their mobile brand even if the others offer you a better deal.In this way the users are distributed which in a way has its own advantages and disadvantages.
I also found a fascinating article on potential application of HCI for aiding children with mental disorders like autism and bipolar disorer.By using HCI it is possible to analyze  and influence human perception and behaviour.Its really a nice article telling about application of HCI for treatment of disorders.U can read this article through this link:www.acm.org/crossroads/xrds12-2/mental.html.This also reminds me of the fascanating seminar presentation by Dr.Kanav.I was quite amazed to know that we can derive lot of applications in medical field through 3D games.The work that he is doing with Banner in rehabilitation centers is also commendable.We all know the value of cost effectiveness in every field and he and his team is able to device methods and devices for surgery just for 400$ which i guess is nothing  as compared to cost of  the conventional devices,its amazing.I feel I am more intersted in this field than ever after listening to him.

Posted by
Ashutosh Singraur

HCI with portable media players

Content:
This post is about how portable media players have a variety of characteristics in their HCI design.  I am going to use the HCI presentation's usability requirements' goals of usability, universality, and usefulness that Debbie C. and Islam G. have also used.

Usability:

Pros:  Portable media players can be light weight and easy to transport.  Also their light weight can make them comfortable to hold over long periods of time.  Portable media players can be so small that they conveniently fit in a standard pants pocket.

Cons:   The devices can be so small that they can be easy to drop out of someone's hands and suffer heavy damage from falling on the ground.  Additionally, some have battery lives that are only a few hours.

Universality:

Pros:  Portable media players can be found at stores around the world.  If one gets broken, replacements are often not hard to find.  A lot of manufacturers make them and distribute them globally.  Competition between global manufacturers has been driving innovation and lower prices with them.

Cons:  New portable media players manufactures may find the barrier to entry in their market high because of all of the existing manufacturers of portable media players.

Usefulness:

Pros:  Some portable media players offer sharp and colorful video playback and high quality audio playback.

Cons:  They can move around in someone's hand when that person is trying to watch a video.  That moving viewing experience can be not as entertaining as if the video stayed still in one place while it was being watched.  In addition, the video size can be somewhat small which makes viewing videos not as interesting as on larger screens.

Posted by:

Nate

Content: A few weeks ago, I went to California to visit my son and his girl friend. When we walked into the home, they were standing in front of a large TV with blaring music and a concert; she was dancing and playing the drums and he was playing a wild guitar. It was their music that was being created and I do not know if I would have gone to the concert. However, I learned later that this was the WII. The Wednesday lecture made me rethink this experience that I really could not relate to at that time. This is the face of human computer interaction that makes a computer more than a computational machine. A whole multidisciplinary field with psychology, computer science, anthropology, software engineering, and sociology is evolving. It is not without risk as we develop “techno-dependency.” In turn, these “new kinds of digital tools” and novel devices lead to an independent new development. “Central into the new agenda is recognizing what it means to be human in a digital future.” (Sellen et al). As part of this, the article by Rich opens up new ideas about how we can interact with devices, even some of the simplest devices that we now have that are complex and also inconsitent. User interfaces with ANSI/CEA 2018 may help create task based menus that allow learning of the device instructions. This is a more limited scope than the interaction of human and computer, but nevertheless just as important. However, this type of concept also allows generalization to devices made by different manufacturers without diminishing the device uniqueness and differentiation that is demanded for profitability by the different manufacturers. We need to be thinking about human and computer/device interaction and how each can benefit from other. Dr. Kanov spoke of universality and extend the ability of disabled or blind individuals to experience. Bad interfaces are not useful but can be acceptable if beautiful or cool. Part of the designs of software and hardware must involve user input and alteration from the beginning as part of multidisciplinary groups. A key feature is that just as software is written in a task based language that the definition and development of these tools must also be task based and hierarchail. Unfortunately, security and privacy issues are part of the human equation that have great risk. Different types of application demand different types of usability, as illustrated for emergent versus less emergent uses, which may often sacrifice satisfaction and user friendliness. User centered design versus activity centered design also becomes a tradeoff. It is clear that things have evolved since I built a short wave radio and got excited about listening to Nairobi or Budapest. I did not understand what they were saying anyway and I can not predict the novel ways that interactions with any one device and the human users will evolve. It is clear that this field has great potential application now and in the future to medical care and that I understanding Java and software engineering will be useful.

Posted by Stuart

Free HCI for you

One example of HCI is the automatic sliding door.  I will speak of its traits in general terms, but as with any technology a standard can have customizations for enhancement.  The automatic sliding door can be found in different environments (personal, medical, business, security).  Because of its motion detection trait, it becomes a universal type of device-opening for those with their hands full, for those with visual impairments and physical disabilities.  For example, it is large enough for a wheelchair to go through with ease and enough space.  It is definitely usable and useful to everyone, as it requires nothing from the user, but provides the user with a service.  Also, good automatic doors will have numerous security and safety features.  A drawback to this technology could be that it malfunctions or does not open.  This could make the user lose a few minutes of his/her time, but the door would eventually be opened the old fashioned way, due to manual features.  However, in an emergency situation, those few minutes could be crucial.  Overall, it's a device that makes our lives a little easier.  Just think of all you could do with that extra second!   

Posted by Annie

HCI

Content: The seminar by Dr Kahol was a good follow up to the lecture on HCI where he introduced his research and very positive outlook on the future of computer simulation  in medicine.
Just as BMI is an integration of various fields, I see HCI as an amalgamation of computer science, ergonomics, industrial engineering, psychology (cognitive science), anthropology (social and cultural aspects of using computers for humans).
If you look at the Human Computer Interaction Institute at Carnegie Melon University you notice they have people from humanities and social sciences, industrial engineering, fine arts,software engineering, technology. Their research project list is so extensive that I was humbled by the scope of this field. This particular project on Alzhemiers Disease where they are trying to develop a memory logging device for people with recent memory loss which can probably help slower the progression of the disease. Its a really interesting idea (I think Stuart will be particularly interested) and here is a link to that project
http://www.cs.cmu.edu/~mllee/mem.html
Their other research projects can be found on this link:
http://www.hcii.cmu.edu/research/projects
(Just an an aside also do watch Randy Pausch's last lecture.....he is one of the most inspiring people I have ever read about/watched. He was a professor at this institute and died recently due to pancreatic cancer)
I definitely think that the human element of HCI is the most challenging part.
The SIGCHI document had an excellent diagram which puts this entire field in a nutshell. As defined by the document it explains "five interrelated aspects of human-computer interaction: (N) the nature of human-computer interaction, (U) the use and context of computers, (H) human characteristics, (C)computer system and interface architecture, and (D) the development process, project presentations and examinations (P)".
I definitely think that usability is the most important factor of any HCI along with the cost. If these alll development is focused around these two aspects, the results should be promising.
As stresses by Dr Kahol, I too think that universality is also an important aspect.
The use of gaming consoles and "Tangible User Interface" in medicine to design simulators was discussed in the seminar very effectively. Multitouch interfaces uses more than one touch from the user to interact with a system and I see the use of the same designing large touchscreens where the user can use both hands and all fingers to manipulate the image. I am still thinking of its uses in BMI.


Posted b: Sheetal Shetty