Automated Identification of Cell Populations in Flow Cytometry Data with Transformers

TL;DR

This paper introduces a transformer-based neural network method for automated detection of blast cells in flow cytometry data to assess minimal residual disease in leukemia, improving accuracy and efficiency over manual methods.

Contribution

It presents a novel transformer architecture tailored for flow cytometry data to automatically identify leukemic cells, outperforming existing approaches.

Findings

Median F1 score of ~0.94 on B-ALL samples.

Outperforms existing methods on four datasets.

Effective across data from three clinical centers.

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 F1 score of ~0.94 when evaluated on 519 B-ALL samples and shows better results than existing methods on 4 different datasets