Synergizing Deep Learning and Biological Heuristics for Extreme Long-Tail White Blood Cell Classification

TL;DR

This paper introduces a hybrid deep learning framework combining generative restoration, transformer-based representation, and biological priors to improve rare white blood cell classification under class imbalance and domain shift.

Contribution

It presents a novel integration of generative, transformer, and biologically-inspired modules for enhanced rare-class WBC classification.

Findings

Achieved Macro-F1 score of 0.77139 on WBCBench 2026 challenge.

Demonstrated robustness under severe class imbalance and domain shift.

Highlighted the effectiveness of biological priors in deep learning models.

Abstract

Automated white blood cell (WBC) classification is essential for leukemia screening but remains challenged by extreme class imbalance, long-tail distributions, and domain shift, leading deep models to overfit dominant classes and fail on rare subtypes. We propose a hybrid framework for rare-class generalization that integrates a generative Pix2Pix-based restoration module for artifact removal, a Swin Transformer ensemble with MedSigLIP contrastive embeddings for robust representation learning, and a biologically-inspired refinement step using geometric spikiness and Mahalanobis-based morphological constraints to recover out-of-distribution predictions. Evaluated on the WBCBench 2026 challenge, our method achieves a Macro-F1 of 0.77139 on the private leaderboard, demonstrating strong performance under severe imbalance and highlighting the value of incorporating biological priors into deep learning for hematological image analysis. The code is available at https://github.com/trongduc-nguyen/WBCBench2026