AdInfer is an AI system specially designed to administer diagnostic interviews and conduct epidemiological studies in the general population. It has the capability to remotely screen large numbers of individuals from the general population and provide data for the calculation of prevalence and incidence of disorders based on a representative sample of the population.
AdInfer also allows to screen patients and their families in order to identify symptoms or patterns of complaints that may not have been seen or identified as related to the disorder otherwise and generate a sharper knowledge of the elements belonging to a diagnosis. This can be done by calculating the relative weight of those elements in the diagnosis of a given pathology. Quality, frequency, intensity are all elements that provide indication on the strength of the associative links between the different criteria and the diagnosis. This will contribute to gain a better understanding of rare disorders and potentially facilitate earlier diagnosis.
With its capability to screen large numbers of individuals in the general populations and clinical populations and to collect large amounts of information, AdInfer can also contribute to the screening and recruitment of participants for clinical trials and generate new insights for the development of treatments.
In 1983, AdInfer (©M Ohayon, 1983) was first created as a level 0+ expert system for the assessment of psychiatric disorders. While some computerized tools use predetermined diagnostic trees in which the software only goes from one node to another without attempting to look for other paths, expert systems are making their decision during the interview, looking for the optimal way to reach their conclusions: to make a diagnosis. From 1983 to 1991, AdInfer went through several changes to increase its diagnostic abilities and diversify its fields of knowledge. In 1990, Sleep-EVAL (©M Ohayon, 1990), a level-2 expert system endowed with a causal reasoning mode, was created for the diagnosis of sleep and mental disorders in general and clinical populations. Neural Network and Fuzzy logic capabilities were developed and increased potentiality and functionality of the system. Contrary to most computerized tools that are unable to analyze temporal information, for example, to determine which symptom appeared first, Sleep-EVAL includes a mathematical preprocessor that allows it to make this type of analysis. Since the beginning of the System, Deep Learning was the heart of the System and gave us an increased ability to develop powerful knowledge bases. The AdInfer/Sleep-EVAL System uses its own data collections (based on epidemiological and clinical research) to learn and improve its knowledge.
AdInfer stores its expertise in its knowledge base. This knowledge base is the result of a deep learning process built on AdInfer data bases. The AdInfer System has been collecting data for 30 years and uses them in a self-training to improve its interactions with the users. The knowledge bases are the results of this deep learning which uses the resources of probabilistic methods (Bayesian theorem) combined with genetic algorithms and fuzzy logic. The result is a fully adaptative system.
The expert knowledge base can be customized for any medical specialty. The combination of deep learning and fuzzy reasoning is the source of the power of our hybrid AI system.
The AdInfer/EVAL system has been used to generate over 200 publications in numerous fields, spanning public health, sleep medicine, mental health, epidemiology, mathematics, and gastroenterology.
