Generalizing Nucleus Recognition Model in Multi-source Images via Pruning

TL;DR

This paper introduces a novel pruning-based domain generalization method to improve multiclass nucleus recognition in multi-source Ki67 IHC images, effectively handling domain shifts and class mismatches.

Contribution

The proposed approach searches for a domain-agnostic subnetwork through iterative pruning and fine-tuning, enhancing generalization across diverse Ki67 image sources.

Findings

Outperforms existing DG methods on Ki67 nucleus recognition.

Achieves superior accuracy especially in challenging lost category cases.

Demonstrates competitive results on public datasets.

Abstract

Ki67 is a significant biomarker in the diagnosis and prognosis of cancer, whose index can be evaluated by quantifying its expression in Ki67 immunohistochemistry (IHC) stained images. However, quantitative analysis on multi-source Ki67 images is yet a challenging task in practice due to cross-domain distribution differences, which result from imaging variation, staining styles, and lesion types. Many recent studies have made some efforts on domain generalization (DG), whereas there are still some noteworthy limitations. Specifically in the case of Ki67 images, learning invariant representation is at the mercy of the insufficient number of domains and the cell categories mismatching in different domains. In this paper, we propose a novel method to improve DG by searching the domain-agnostic subnetwork in a domain merging scenario. Partial model parameters are iteratively pruned according to the domain gap, which is caused by the data converting from a single domain into merged domains during training. In addition, the model is optimized by fine-tuning on merged domains to eliminate the interference of class mismatching among various domains. Furthermore, an appropriate implementation is attained by applying the pruning method to different parts of the framework. Compared with known DG methods, our method yields excellent performance in multiclass nucleus recognition of Ki67 IHC images, especially in the lost category cases. Moreover, our competitive results are also evaluated on the public dataset over the state-of-the-art DG methods.