Welcome to AI-MED

Where Artificial Intelligence meets Medicine to leverage Smarter Medical Decisions

AI-MED introduces PAIRS ( Phyisican Assistant Artificial Intelligence System) An Ingenious & State of the Art Clinical Decision Support System (CDSS)

Clinical Domains

Multiple. General Practice

Diagnosis of difficult cases, Technologies used: Decision support systems, Bayesian probabilistic belief networks, variational methods for efficient inference in large scale probabilistic models. Physician Assistant Artificial Intelligence Reference Systsem (PAIRS) is designed to help doctors diagnose difficult cases.

Medical Diagnosis

PAIRS

Physician Assistant Artificial Intelligence Reference System (PAIRS) is a patented product. It is useful for diagnosing difficult to diagnose cases. For any given case it suggests some tests and a differential diagnosis. The suggestions along with doctor’s intuition helps in arriving at a diagnosis.

Similar Products

Isabelle, QMR-DT and DxPlain

There is interest in application of AI to Medicine since last several decades. Earliest product was DxPlain followed by QMR-DT. Usually some logic or algorithm is applied to a database for identification of diagnosis given patient data. DxPlain and QMR-DT are no longer being developed, Isabelle has world wide market.

PAIRS ADVANTAGE

AI and Natural Language Processing

Doctors have option of entering feature data as they choose and a SNOMED-CT based NLP helps find relevant features in database. Combined with this capability, there is a diagnostic algorithm that suggests tests and possible diagnosis. This process helps doctor acquire knowledge and also treat their patients.

0%

Diagnostic Data Support

How is Pairs relevant to you ?

Difficult Cases

Some internal medicine cases in domain of connective tissue disorders or autoimmune diseases have multiple organ involvement. These cases are difficult to diagnose.

Identify Patient Data

User choice for easy identification and entry of patient data.
(1). Automated dropdown listing.
(2). Natural Language Processor identifies patient data by using a SNOMED-CT algorithm.

Diagnostic Algorithms

Bayesian probabilistic belief networks are commonly used in artificial intelligence. Direct application of these for medical diagnosis is difficult. Alternative solution is an approx method which is implemented in PAIRS.

Mis or Missed Diagnosis

Errors in medical practice are common and are catastrophic to both patient and doctor. This can be prevented by use of Clinical Decision Support Systems (CDSS).

Patient Data Entry

Completing patient data entry involves identification of corresponding features in PAIRS database. This involves use of NLP for similar or related features, acronyms. Good case building helps diagnosis.

Diagnostic Probability

User identifies either infection, neoplasia, autoimmune or others as expected disease process. Based on this PAIRS gives diagnosis from top 25 possibilities. It also suggests investigations for them

PAIRS Advantage

AI-MED helps you in validating your diagnosis though various factors

Featured on India Innovation Growth Programme 2006

Database

Totally 30 000 disease-feature links for about 476 diseases and 2000 features.

Artificial Intelligence

Natural Language Processing based on SNOMED-CT data and algorithm. Diagnostic algorithm includes approximation method of Bayesian probabilistic belief networks as developed by Jaakkola & Jordan and published in Journal of Artificial Intelligence Research (Variational probabilistic inference and QMR-DT 10 (1999) 291-322.).