Automated identification of cell populations in flow cytometry data with transformers. (May 2022)

• Record Type: Journal Article
• Authors: Wödlinger, Matthias; Reiter, Michael; Weijler, Lisa; Maurer-Granofszky, Margarita; Schumich, Angela; Sajaroff, Elisa O.; Groeneveld-Krentz, Stefanie; Rossi, Jorge G.; Karawajew, Leonid; Ratei, Richard; Dworzak, Michael N.
• Is Part Of: Computers in biology and medicine. Volume 144(2022)
• Publication Date: 2022-05
• Subjects: Multiparameter flow cytometry -- Automated gating -- Deep learning -- Self-attention; Medicine -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; 610.285
• DOI: 10.1016/j.compbiomed.2022.105314
• Languages: English
• ISSNs: 0010-4825

Abstract

Abstract: Acute Lymphoblastic Leukemia (ALL) is the most frequent hematologic malignancy in children and adolescents. A strong prognostic factor in ALL is given by the Minimal Residual Disease (MRD), which is a measure for the number of leukemic cells persistent in a patient. Manual MRD assessment from Multiparameter Flow Cytometry (FCM) data after treatment is time-consuming and subjective. In this work, we present an automated method to compute the MRD value directly from FCM data. We present a novel neural network approach based on the transformer architecture that learns to directly identify blast cells in a sample. We train our method in a supervised manner and evaluate it on publicly available ALL FCM data from three different clinical centers. Our method reaches a median F 1 score of ≈ 0.94 when evaluated on 519 B-ALL samples and shows better results than existing methods on 4 different datasets. Graphical abstract: Image 1 Highlights: Supervised learning of blast cell detection in ALL flow cytometry data. Novel attention based architecture for flow cytometry data that allows processing of up to 106 cells in a single forward pass. Light-weight neural network architecture that allows fast processing of samples. Comparative study of the performance of our method for data from different centers/fcm-machines.