TAMI

Transparent Artificial Medical Intelligence

The project will develop a set of tools to support medical decision, based on artificial intelligence algorithms. The project will work on a platform for commercial, scientific and academic use that will provide “consumers” access to results and explanations of diagnostic orders, filtered data sets access for investigators or scientists, and a knowledge base for academic purposes.

Keywords: Artificial intelligence · Computer-aided decision · Healthcare ethically in their context.

Leading company in Portugal
FIRST SOLUTIONS - SISTEMAS DE INFORMAÇÃO S.A.

Project Start Date
01/04/2020

Project End Date
01/03/2023

TAMI, explaining medical diagnosis through AI tools

The project main mission is to help make medical diagnosis clearer and more reliable, supported by artificial intelligence. TAMI will develop a set of tools to support medical decision based on AI algorithms, that will explain to both clinician and researchers the diagnosis of a specific disease and its causes, focusing on cervical cancer, lung diseases and eye diseases.

The system will explain the information visually through images or text, by presenting information through a concept or sentence that makes sense to the user.

What to expect

The tools developed in this project will potentially help to detect possible flaws and introduce improvements in the diagnosis support systems while helping to determine disease patterns and offering new sources of knowledge.

Meet the partners

Promoter:
FIRST SOLUTIONS – Tiago Oliveira

Academic Co-promoters:
FRAUNHOFER PORTUGAL – Luís Rosado
INESC TEC – Ana Maria Mendonça and Jaime Cardoso
Administração Regional de Saúde do Norte (ARS Norte), I.P. – Pedro Sousa

CMU:
Electrical and Computer Engineering Department – Asim Smailagic

Goals

The project will be based on the following specific objectives involving the development of research in the following areas:

Quantitative methods to objectively assess and compare different explanations of the automatic decisions;
Methods to generate better explanations, providing variety in the explanations, adapting it to who will consume them and explaining multimodal decisions;
Novel visualization solutions for interpretations of decisions based on imagiological data.

In order to accomplish that, TAMI will use clinical data, Including textual and image data, in order to design and validate interpretable machine learning models. During the project, different multimodal settings will be tested to enable a better understanding of the AI-based decisions. Moreover, the algorithms will be designed to generate self-explanatory AI-based decisions, minimize bias, and act ethically in their context.