CAIDSA: Computer-aided intelligent diagnostic system for bronchial asthma

The objective of this paper is to design an interactive computer-aided intelligent diagnostic system for bronchial asthma (CAIDSA) based on the clinico-epidemiological information. Primarily, a number of disease-related subjective questions pertaining to the epidemiological features are considered to be developed as an intelligent probing system. In this evaluation system, a popping factor, whose values are assigned by medical experts, is defined in such a way that a question is popped subsequently depending upon a patient’s response to the previous question. And finally, based on the subjective evaluation, the specific type of asthma is identified with its degree of severity in terms of probability. This disease probability is estimated using neural network where the decision rules are set to recommend whether a patient should go for confirmatory tests (clinical tests) or not. If a patient is suspected or even detected, she/he is sent for confirmatory tests on the basis of which the disease is confirmed, and its disease probability is again revised to increase the diagnostic accuracy. In this proposed CAIDSA tool, there is also a scope for learning the diagnostic features for better diagnostic improvement. In fact, the entire schematic diagram of CAIDSA and the outline of the software implementation are pointed out in this work.