025-23 – Deep Learning approach for explainable AI and Clinical Decision Support System

Prior to technological evolution, medical diagnosis focused on previous knowledge and behaviour of the disease. Sometimes, it is very difficult for a doctor to diagnose a disease based on his knowledge and experience. With the development of science and the help of various medical equipment and machines, it is easy to diagnose diseases at an early stage. Artificial intelligence (AI), machine learning, and deep learning are key areas for medical diagnosis. AI has been applied to solve several personalised medical problems.

Currently, artificial intelligence (AI) is used to collect medical data embedded in modern computing applications at home, in vehicles, and in the workplace. Modern machines are built to carry out specific activities or purposes, such as collecting medical data. These heterogeneous data were collected systematically but not processed accordingly. A clinical decision support system (CDSS) can convert raw medical data into meaningful data.

The increasing usage of CDSS in healthcare and the demand for software that enables medical professionals to make informed decisions are changing everyday clinical practise. However, as technology advances, not only are the benefits of technology growing, but so are the potential risks. A potential drawback of the CDSS is the over-reliance on the proposed decision, which leads to deskilling and rash decisions by medical experts. Thus, recognising health professionals’ requirements for specific disease-based diagnostic software and developing approaches to prevent technological over-reliance and improve trustworthiness are very important.

In this study, we report the results of the increasing usage of CDSS in healthcare and the associated risks of this utilisation.