ATLAS

Liver cancer is the second leading cause of cancer-related death. Diagnosis and treatment are time-critical and require highly patient-specific diagnostic and treatment pathways. Medical decision-making is based on a multitude of interdependent factors related to various medical disciplines, past experiences and clinical guidelines. The consideration of all decision factors in combination with the possible therapy approaches is a major challenge for the physicians and often cannot be solved in an optimal way even in the interdisciplinary tumor board. In this project, we develop ATLAS, a decision support tool, which will significantly assist clinicians with this challenge. Based on AI methods, ATLAS processes all relevant patient data from databases, systems medicine and continuum biomechanical in silico prognoses modeling data as well as individual patient data. The tool will be developed in a co-design approach by experts in surgical oncology, mathematical modeling, and machine learning. Chosen technologies integrate the automated understanding of a highly complex patient situation through the simulation of liver functions with expert knowledge and ontology-guided learning with knowledge graphs from retrospective cases of liver tumors. ATLAS will be based on detailed historical data cohort from over 6,000 patients with liver tumors and will be evaluated on case studies at the Jena University Hospital. The integration of medical expert knowledge, mathematical modeling and artificial intelligence constitutes a highly original and most promising approach for high-quality diagnoses and treatments of liver tumors resulting in patient-specific prognosis improvement. Scientific insights from this projects will offer exploitation possibilities for the transfer to malignancies in other organs, such as lungs, kidneys or the brain. Tool and demonstrator development will provide for sustainable exploitation pathways for future commercial applications.

Clinical integration of AI and Simulation for Tumor Liver ASsessment (ATLAS) in the treatment of liver tumors. As the central decision-making element, the tumor board will empowered by ATLAS in the diagnostics and treatment decisions of liver tumors. Currently, the treatment decision is based primarily on the patient’s clinical data, image data, histological reports and the current state of scientific evidence. In the future, ATLAS will support the physicians as a decision support in the clinical process by additionally using In Silico Data and AI. Already in the outpatient clinic, ATLAS will provide recommendations on the necessary diagnostics with regard to the specific tumor and the patient’s condition. In the tumor board, ATLAS will then support the physicians by predicting the outcome of the therapy. ATLAS also contributes to save resources in the structured oncological follow-up by recommending patient-specific examination intervals and diagnostics. As a clinical decision support tool, ATLAS will personalize the therapy and improve patient specific outcome.

We had an awesome time with the ATLAS team at our first in-person meeting in Stuttgart! We covered a lot - checking our progress, brainstorming how to blend AI, simulation, and clinical data, and enjoying swabian dinner together. It's exciting to see how we're all working together towards an AI- and simulation-based clinical decision supporting tool for liver tumor assessment. 

Tim Ricken
Institute of Structural Mechanics and Dynamics in Aerospace Structures
University of Stuttgart, Pfaffenwaldring 27, 70569 Stuttgart, Germany
tim.ricken@isd.uni-stuttgart.de, +40 711 685 63612
https://www.isd.uni-stuttgart.de/

Steffen Staab
Institute for Artificial Intelligence (KI) – Analytic Computing
University of Stuttgart, Universitätsstraße 32, 70569 Stuttgart, Germany 
steffen.staab@ipvs.uni-stuttgart.de, +49 711 685 88100
https://www.ki.uni-stuttgart.de/departments/ac/

Matthias König
Institute for Theoretical Biology (ITB)
Humboldt-University Berlin, Phillipstraße 13, 10115 Berlin, Germany 
koenigmx@hu-berlin.de, +49 30 2093 98435
https://livermetabolism.com/

Hans-Michael Tautenhahn
Department of Visceral, Transplantation, Thoracic and Vascular Surgery
University Hospital Leipzig, Leipzig, Liebigstr. 20, 04103 Leipzig 
hans-michael.tautenhahn@medizin.uni-leipzig.de, +49 341 9719966
https://www.uniklinikum-leipzig.de/einrichtungen/vttg/%C3%A4rztliche-mitarbeiter-innen

H. M. Tautenhahn, T. Ricken, U. Dahmen, et al. SimLivA–Modeling ischemia-reperfusion injury in the liver: A first step towards a clinical decision support tool, GAMM-Mitteilungen. (2024), e202370003. https://doi.org/10.1002/gamm.202370003 

L. Lambers, N. Waschinsky, J. Schleicher, et al. Quantifying fat zonation in liver lobules: an integrated multiscale in silico model combining disturbed microperfusion and fat metabolism via a continuum biomechanical bi-scale, tri-phasic approach. Biomech Model Mechanobiol 23, 631–653 (2024). https://doi.org/10.1007/s10237-023-01797-0

S. Gerhäusser, L. Lambers, L. Mandl, et al. Simulation of zonation-function relationships in the liver using coupled multiscale models: Application to drug-induced liver injury. bioRxiv 2024.03.26.586870 (2024); doi: https://doi.org/10.1101/2024.03.26.586870