Identification and visualization of fusion gene subtypes in APL using spatial attention mechanisms in vision models

TL;DR

This paper introduces a deep learning model with spatial attention to accurately identify and visualize APL fusion gene subtypes from bone marrow images.

Contribution

A novel CNN with spatial attention mechanism for high-precision APL subtype identification and visualization.

Findings

The model achieved 98.04% overall accuracy in classifying five APL fusion gene subtypes.

Attention maps improved model interpretability by highlighting key image regions.

The approach offers a rapid and accurate diagnostic alternative to traditional genetic testing.

Abstract

Acute promyelocytic leukemia (APL) features leukemic cell differentiation arrest at the promyelocytic stage, mainly due to the t (15;17), (q24; q21) translocation that forms the PML-RARA fusion protein. Variant RARα translocations, with distinct biological traits and all-trans retinoic acid (ATRA) responses, often cause misdiagnosis and lengthy genetic testing.

To solve these problems, we propose a spatial attention mechanism-enhanced convolutional neural network integrating ResNet Blocks and a spatial attention module (CNN with spatial attention), which can achieve high-precision identification of APL fusion gene subtypes and pixel-level visualization of key areas. Data collected from two hospitals and Kaggle, including bone marrow smear images of PML-RARA, TTMV-RARA, NPM1-RARA, STAT5B-RARA, and NUP98-RARG subtypes, were preprocessed to form a five-class dataset.

The model achieves an overall accuracy of 98.04% in five - class classification, with good performance in each category. The attention maps also enhance the model’s interpretability.

Such a novel and rapid diagnostic approach for APL subtypes, which achieves high - precision identification and pixel - level visualization, holds significant clinical value.

Figures