On the importance of local and global feature learning for automated measurable residual disease detection in flow cytometry data

TL;DR

This paper assesses deep learning approaches for detecting residual disease in flow cytometry data, emphasizing the roles of local and global feature learning, and proposes improvements to the existing SOTA model for better accuracy and generalization.

Contribution

It introduces two modifications to the current SOTA model, enhancing performance and generalization in MRD detection from flow cytometry data.

Findings

Enhanced SOTA model outperforms previous methods on public datasets.

Improved model generalization across different laboratories.

Insights provided for future deep learning architecture design in FCM analysis.

Abstract

This paper evaluates various deep learning methods for measurable residual disease (MRD) detection in flow cytometry (FCM) data, addressing questions regarding the benefits of modeling long-range dependencies, methods of obtaining global information, and the importance of learning local features. Based on our findings, we propose two adaptations to the current state-of-the-art (SOTA) model. Our contributions include an enhanced SOTA model, demonstrating superior performance on publicly available datasets and improved generalization across laboratories, as well as valuable insights for the FCM community, guiding future DL architecture designs for FCM data analysis. The code is available at \url{https://github.com/lisaweijler/flowNetworks}.