# Classifying Mitotic Figures in the MIDOG25 Challenge with Deep Ensemble Learning and Rule Based Refinement

## TL;DR

This paper presents a deep ensemble learning approach with rule-based refinement for classifying mitotic figures, achieving high accuracy but with trade-offs in sensitivity, highlighting the potential and challenges of automated tumor biomarker analysis.

## Contribution

Introduces a ConvNeXtBase ensemble with rule-based refinement for mitotic figure classification, demonstrating high accuracy and exploring the effects of RBR on performance.

## Key findings

- Ensemble achieved 84.02% balanced accuracy on MIDOG25.
- RBR increased specificity but decreased sensitivity.
- Deep ensembles are effective for atypical mitotic figure classification.

## Abstract

Mitotic figures (MFs) are relevant biomarkers in tumor grading. Differentiating atypical MFs (AMFs) from normal MFs (NMFs) remains difficult, as manual annotation is time-consuming and subjective. In this work an ensemble of ConvNeXtBase models was trained with AUCMEDI and extend with a rule-based refinement (RBR) module. On the MIDOG25 preliminary test set, the ensemble achieved a balanced accuracy of 84.02%. While the RBR increased specificity, it reduced sensitivity and overall performance. The results show that deep ensembles perform well for AMF classification. RBR can increase specific metrics but requires further research.