A combine use of physicians and the new tools of AI, pet scan, Nanotechnology, etc. will give a more effective system.

Really interesting question. The fact of the matter is if the world medical specialists are able to input this into their workplace, it will be very fruitful for us.

Since MYCIN was developed in the early 70's by Edward Shortliffe and the fantastic boost that it gave in many application domains (e.g. PROSPECTOR, HYDROLAB, SATAN, etc.)  it remains frustrating that integrated expert-systems, able to directly gather data from where they are hosted or produced  in their various native forms and therefore running on a rather autonomous mode, have not as much as we would have hoped known the kind of industrial deployment that we would have expected.

It would deserve a careful analysis as to identify the many reasons that have somehow hampered the large scale usage of such systems.

I provide some references dealing with various systems we developed in several application domains.

Best,

Patrice.

Chapter Artificial Intelligence Tools and Techniques for Water-Resou...

Article HYDROLAB: an example of a new generation of compact expert systems

Chapter Artificial Intelligence Modelling of Complex Systems in Mode...

Technical Report Réalisation d'un Prototype de Simulateur Expert d'Aide au Co...

Conference Paper Une Nouvelle Classe de Simulateurs Destinée aux Aides Tactiq...

Dear Kamal

Healthcare means question of life and death. The instances abound in which AI fails to help take correct decision. I would prefer to visit an experienced physician/surgeon for diagnosis who does not take help of AI in diagnosis and prescription rather than to one who takes helps of AI. It is why I believe in precautionary principle in risk assessment.

A combine use of physicians and the new tools of AI, pet scan, Nanotechnology, etc. will give a more effective system.

Once the patient data collected, the diagnose is based on the stored medical knowledge. The data on symptoms or signs, special data of laboratory or tests or radiological images are process by the system using defined rules to obtain the possible diagnoses. Additional data such as the presence or absence of certain signs and symptoms help to make a final diagnose. The rules of these Expert systems include the diagnose criteria from world-wide Associations, as well as algorithms designed by members of the Laboratory of Intelligent Systems of the University of The Andes. The qualities of this system are:

1. It can diagnose of one or more diseases and suggests the appropriate therapy.

2. It diagnoses the absolute absence of anyone of these diseases.

3. It can find some symptoms or signs due to any exogenous cause (a differential diagnose).

4. It notifies to the doctor that the patient does not fill the minimum criteria for some of the diseases and in this case, suggests a new evaluation.

5. It suggests send the patient to a specialist.

The reasoning for establishing diagnoses or hypotheses of diagnoses is given as well as the plans for other examinations and for patient treatment. Also it is indicated when there are inexplicable signs, symptoms or laboratory data. They include the realization of a set of questions individualized for each subject and the selection of data that is going to be acquired answering the questions.

http://dl.acm.org/citation.cfm?id=1565279

From 13 to 20% of diagnosis established by doctors are wrong in developed countries, and up to 40-50% in developing World. I am afraid that the expert system applied in developing countries will not help to reduce cost of diagnostics without loss of  correct diagnosis rate.

http://www.ncpa.org/sub/dpd/index.php?Article_ID=23148

http://www.rightdiagnosis.com/mistakes/common.htm

Here are some more examples of the application of AI and ES in Medicine! 

http://www.medgadget.com/2007/04/estele_expert_s.html

http://www.ncbi.nlm.nih.gov/pubmed/10870619

Article An Approach to Develop Expert Systems in Medical Diagnosis U...

Article Knowledge Management in ESMDA: Expert System for Medical Dia...

(Although not exactly answering your question about expert systems), you could use Probabilistic Graphical Models for Medical Diagnosis and Decision Making:

https://class.coursera.org/pgm-003/lecture/6

http://pgm.stanford.edu/intro.pdf

http://www.igi-global.com/chapter/learning-probabilistic-graphical-models/67261

http://books.google.at/books/about/Medical_Diagnosis_Using_Artificial_Neura.html?id=4SqkoAEACAAJ&redir_esc=y

I think a lot of successful applications of ES in medicine has already been developed. It may be correct to say that application of ES has seen highest success in this field.  Other paradigms of AI has been used for other fields.  If some one intends to use CBR for application in Medicine field we may over come shortcomings of ES. It will have addition of learning module to keep the latest experiences in tact use these for better solutions.

Neural networks that use all the resources of mathematics and computer with newer methods of calculation and evaluation. They are systems that simulate the neural circuits of the brain and the way they manage and distribute information. They were designed as structures of nerve cells, which develop learning ability and are able to learn to perform the functions for which have not been directly programmed. There is talk of designing neural systems with 10 billion components, but it is clear that we are still far from the complexity of human brain. Born as functional model of biological neural networks, artificial neural networks have proved to be very useful tools within the classifications of objects in complex morphology, including the EEG, where the parameters used are varied and not always well determinable a priori, as related to criteria of similarity. Artificial neural networks, once trained by presenting to them the data paths of the measures taken as examples, are able to recognize and classify each chart patterns other tracks with speed, accuracy and stability much higher than those obtainable with the evaluation by eye , as it is experienced.

Starting from the aforementioned quantization of the EEG signal, you can get to its automatic classification, which consists in identifying the traits of track normal or altered and, within the latter, to a further division into several categories, similar to what is is usually the direct visual assessment.

However, an emergency decision may be taken only by human brain. 

Expert system which offers us  advice or make decisions that requires human intelligence for developing a system for curing diseases?? A very good idea and concept, still there is a "but" factor in this concept. When we talk of humans which is the most complex system in the world it just can not be replicated or exchanged to obtain 100% accurate results or to be more precise 100% satisfying results, the decisions we make according to the situations can only be done by us although ES could contain a knowledge base of what should be done if a particular situation is observed. Yes the level of precision in a particular operation could be best performed by an ES. There are many points which could be debated on this topic.

Again the ideology of ES is very nice but it should not be taken to the level where humans are said to be less effective than the SYSTEMS. I don't know how much relevant i am but this is my view about the topic.

it seems that all of the good ideas have not been universally implemented or applied yet!