Skip to main content

User login

What is not?

 Impossible is not a fact, its an opinion 

Getting Started

March 31, 2010 by Dr.Inayath

Too many of us are not living the dreams, because we are living the fears of others -- Project N is about an effort by us to overcome the fear of venturing into an unexplored terrain, about overcoming the high tides and reaching the shore of success. Welcome.

A Medical Expert System is a computer software that generates diagnosis and management guidelines given the patients symptoms,signs and investigation results. It is intended to be an aid to physicians rather than replace them and their clinical judgement.


               " To create as a community, a reliable and robust medical expert system that will run on all major platforms, available in various international languages and compatible with other electronic health media. An expert system which rationalizes its diagnosis and treatment (i.e gives indepth analysis), is available with its peak performance round the clock, and by its unemotional, steady and infalliable nature help increase efficiency and prevent inadvertent mistakes. It also serves as a learning tool for students.We wish to learn during its development - the experts knowledge, examine them for correctness and completeness and also study in depth the protocols and medical architecture. "


For the uninitiated who are wondering how the MES can help you, and whether is it worth the trouble and expense?, here are a few advantages:


1. Availability and Permanence: With MES, you have an expert at your finger tip all the time. A human expert dies, a MES doesnt - the knowledge is preserved permanently. Thou shall die, thy knowledge never

2. Reduced cost: Since the replication of the software is free, there is a tremendous cut of costs. We are using all free, open source software to build MES, hence the cost of running the software will be virtually zero. Frugality is an article of faith for us.

3.Reliability, Reduced danger/risk : Prevent human mistakes especially by novice doctors, or when human is tired and overwhelmed with work.

4.Combined expertise :  We are developing  a system that has expertise of many experts, which stores different profile for each expert and by knowing which profile works better or which treatment is more effective, it can be given more preference.

5.Explanation : A physician may be tired, or may not have time to explain to the patient all the intricate details. MES can help here. Also, for a learning physician, MES can give step by step explanation of how it solved the case and decided the management.

6.Faster : MES shall take lesser time than experts, more so in the areas of calculations and extensive derivations.
7. Steadfast :  Is Steady,complete with  unemotional response all the time thereby available round the clock with its peak efficiency.

8.Tutor :  Students can run sample cases and learn how the expert reasons and verify those methods.  The open database serves as a great resource for learning examples.

9. Strong sense of Community : We are all alike minded developers who think technology can go a long way in improving our health care. This project is a place for us to network, and bring ideas together.

10. Development  advantages: As developers of this software we get to learn and examine an experts knowledge. This can then be tested with other experts for correctness,consistency and completeness. We are actively seeking and learning diagnostic paradigms that can solve cases or unison of various methods as a general problem solver. In the end, the whole process of assembling and building the software would be a great learning experience.



When I began this project, the primary aim was to build a system for rural India, however as this is an open source initiative i expect and hope the project will diverge to meet other uses too. Here are a few uses i can think off my head....

RURAL MEDICINE - I read on an internet article that the diagnostic accuracy in rural PHCs is very poor. If this surprises you, it shouldn't - the chaotic state of medical practice in rural india requires no testimonial. The state of affairs is such that the PHCs are provided with a set of drugs and the physician is required to treat all kinds of maladies with the same. This creates short circuited algorithms for management where differential diagnosis is rarely invited. No periodic examinations are held for physicians thereby leading to little or no motivation to gather new clinical protocols and update/revise ones knowledge. However this shouldn't be taken as a generalization for all doctors, there are many who are good doctors despite the system. However this criticism does apply to most of us (including me) who have come out of this rather poor education system. More about our great education system here. Anyways, as you can guess rural India is not the best of places to get sick at, nonetheless many do get sick here and many more suffer. But there is something we can do about it - the least is by working on Project N. And I am very confident that we will succeed, as we are not attempting to build the most perfect system (although we will try our best) but one good enough to give a diagnosis better than today's 50% (around 90%). In this formative stage of our project, we must also keep in mind extensibility issues to include various languages. Please read this page on how we plan to address this issue in development.

SYSTEM INTEGRATED INTO CLINICAL SETTINGS - The simplest integration would be to have the software by the doctors side, to act as his clinical decision support system. But it can be further expanded into a workflow where patients interact with system before meeting the doctor, so that the system collects useful information (all but physical examination) and pass on the report to the doctor who can verify, add findings and further use the MES. This can be particularly useful in busy centers. (Physicians may not be able to collect complete history in the very short time they get, moreover its not possible to do exhaustive history taking with each and every patient).

STUDENTS : The software akin to gaming allows students to input cases and study the various reports generated. As a part of this project, we plan to build a case question bank which can be easily converted into a quiz bank where students get to study real time cases.

What do you do when you want to conduct a research study on a certain disease? Yes - you make samples of diseased people and contrast group, follow them and conduct your study. But doesn't statistics teach us that the larger the sample size the more accurate the result is. Thousands of your sample cases walk in and walk out of clinics and their data is either not stored or is locked in traditional EMRs. Shouldn't there be an easy way to tap into a system and gather the information? The problem with building such a system is that although its easy to store information of patient in database, it is not possible to retrieve the exact piece of information from that database. To solve this problem we need a controlled vocabulary, and need to store the case details in a formalized way. In Project N, we use a controlled vocabulary and generate formalized cases, and if this be integrated into an EMR it can work wonders for researchers. Taking this even further, we can identify patterns of unexplained case findings in a disease and even research which treatment modality works better.


Now that we have dreamed the dream, its time to make living the dream possible. Lets remove our imagination hats, and have a dose of reality -- How in the blue hell are we going to build such a system? Why couldn't such a system be build so far? If the top guys at Artificial intelligence lab are still on it, how do you expect us commoners to ever dream to make one? Who will agree to spend his/her valuable time on building a software which for now appears to be a figment of imagination? What resources do we have? What does one get in return by contributing to this project? Who is responsible if the system makes a mistake? What about ethical issues? Words dont speak, actions do - what is the product of our actions, what did we build so far? Even if we build the basic MES for rural areas, how does one implement it? The hardware and hardwired people, what about them?

Your rating: None Average: 5 (2 votes)
Premium Drupal Themes by Adaptivethemes