IEEE ICTS4eHealth 2026 Special Session

In clinical applications, the presence of information technologies is becoming increasingly common, providing solutions that support physicians in their decision-making processes. Clinical decision support systems (CDSS) that incorporate artificial intelligence (AI) models are powerful tools for enhancing healthcare professionals' ability to deliver accurate, timely patient care. CDSSs are available for many tasks from diagnosis to prognosis: ROI detection, lesion classification, drug choice, treatment planning, and survival prediction.

Regardless of the job performed, it is mandatory to have a full understanding of the entire processing pipeline, from data acquisition to the CDSS outcome. What happens at each stage, which features are involved, and what role they play in the model/classifier's outcome are legitimate questions that need to be answered. Indeed, it is no longer sufficient to use computerized tools as black boxes, without providing information on how data is processed and quantified and, above all, on what types of data play a decisive role in the final decision. 

These needs motivate Explainable Artificial Intelligence (XAI). There are many advantages to implementing explainable models, including explanatory debugging, user acceptance and control, reliability and insight, and ethical/legal considerations.

In this special session, contributions proposing models/classifiers for clinical applications that address model explainability and feature interpretability are welcome.

The main topics are (but are not limited to): 

• explainable models for decision-making support in healthcare

• explainable machine learning algorithms

• interpretable features

• by design, ante-hoc, and post-hoc explainability

Submission Guidelines