Machine-Learning-Based Algorithms for the Detection of Residual Disease in Patients with Acute Myeloid Leukemia

For patients with acute myeloid leukemia (AML), multiparametric flow cytometry (MFC) of blood or bone marrow samples represents a rapid and efficient method for detecting minimal residual disease (MRD) and is increasingly being used alongside molecular genetic diagnostics for disease monitoring. The analysis of the resulting MFC datasets relies on the interpretation of two-dimensional projections with the manual identification of aberrant cell populations. This process is time-consuming, subjective, and limited to specialized expertise often found only at academic centers.

Computer-assisted statistical approaches, such as machine-learning techniques, offer numerous opportunities for achieving expert-independent, reliable, and reproducible diagnostics of MFC datasets. The goal of this project is to develop a software solution capable of performing risk assessment regarding the presence of MRD, thereby supporting subsequent clinical therapy decisions.

Scientists involved

• Prof Dr Ingmar Glauche
• Dr rer. nat. Friedemann Uschner
• Dipl. Inf. Lars Thielecke
• Dr rer. medic. Uta Oelschlägel
• Dr medic. Malte von Bonin
• Dr medic. Maximilian-Alexander Röhnert

Funding

Funded by the Wilhelm Sander Foundation.