INESC TEC developed artificial intelligence algorithms to support medical diagnosis

The TAMI project (Transparent Artificial Medical Intelligence), led by First Solutions and with the participation of INESC TEC through the Centre for Telecommunications and Multimedia (CTM) and the Centre for Biomedical Engineering Research (C-BER), aims to make medical diagnosis supported by Artificial Intelligence (AI) clearer and more reliable.

There are many examples of artificial intelligence application in the healthcare sector, with benefits in terms of data processing, support for medical diagnosis and identification of top treatments. As the introduction of AI into healthcare increases, so does the need to make this technology more transparent and accessible to doctors and patients, thus increasing the reliability of medical diagnosis.

In this sense, the TAMI project will develop a set of tools to support medical decision, based on AI algorithms, which explain to the professionals and the patients the diagnosis of a certain disease and its cause.

These explanations can be visual (identifying the areas of the image that allow drawing conclusions or diagnosing a specific disease) or textual (through a concept or phrase that makes sense to humans).

According to First Solutions, “these tools have several advantages: on the one hand, they will support the doctors in making a decision. On the other hand, they will contribute to a better recognition of said suggestions and solutions by the clinical teams. In addition, they will help identifying potential flaws and improving the diagnosis support system”.

Understanding the reasons “why” to trust and create knowledge

This system can also help determining disease patterns, thus generating new knowledge. By explaining how it obtained a certain diagnosis, the system generates greater reliability for the doctor and the patient.

“The project will explore new quantitative methods to objectively assess and compare different explanations of automatic decisions, and to generate better explanations, thus providing different explanations, while adapting said explanations to those who will use them and explaining multi-modal decisions. Moreover, it will focus on new visualisation solutions for the interpretation of decisions based on imaging data”, explained Jaime Cardoso, CTM coordinator.

First Solutions – Sistemas de Informação SA, leads the project, in partnership with INESC TEC, Fraunhofer Portugal Research Association, Regional Health Administration of the North I.P. and Carnegie Mellon University (U.S.A.).

The project was selected as part of the “Go Portugal – Global Science and Technology Partnerships Portugal” initiative, within the scope of the CMU Portugal Programme. The TAMI received €1.791M from the North Portugal Regional Operational Programme, COMPETE2020, the Foundation for Science and Technology (FCT), partner companies and the Carnegie Mellon University, through the CMU Portugal Programme.

The INESC TEC researcher mentioned in this news piece is associated with UP-FEUP.