System und Verfahren zur Unterstützung der Multidisziplinären Entscheidungsfindung für Behandlungsempfehlungen

Details

Veröffentlichungs-Nr.

EP4787384

Anmeldetag

31. Januar 2025

Veröffentlichung

5. August 2026

Rechtsraum

EP

IPC

G16H50/20

Abstract

A system and a method for assisting multidisciplinary decision-making for treatment recommendations are described. A two-tier inference system with a first-tier trained model and at least one second-tier trained model is provided. The method includes accessing patient features derived from a clinical record of a patient, performing a high-level classification with the first-tier trained model using the patient features to identify a treatment category for the patient, performing a low-level classification with the at least one second-tier trained model using the patient features and the identified treatment category to identify one or more treatment options for the patient within the identified treatment category, determining for at least some of the patient features contribution scores, each contribution score indicating an impact of a corresponding patient feature on the identified treatment category or the identified one or more treatment options, and rendering at least one of the identified treatment category, the identified one or more treatment options, and one or more indications of the contribution scores on an interactive dashboard.

Anmelder

Firmen

Universitätsmedizin der Johannes Gutenberg-Universität Mainz
Deutsches Forschungszentrum für Künstliche Intelligenz GmbH (DFKI)

Land

🇩🇪 Deutschland

Fachgebiete