Segmentation of Multiple Myeloma Plasma Cells in Microscopy Images with Noisy Labels

TL;DR

This paper presents a robust segmentation method for multiple myeloma plasma cells in microscopy images, effectively handling noisy labels through extensive augmentation and ensemble strategies, achieving state-of-the-art accuracy.

Contribution

The paper introduces a novel combination of augmentation, ensemble, and advanced segmentation architectures to improve accuracy on noisy microscopy image labels.

Findings

Achieved a mean IoU of 0.9389 on the SegPC-2021 test set.

Demonstrated robustness to noisy labels with extensive augmentation.

Outperformed previous methods in the competition.

Abstract

A key component towards an improved and fast cancer diagnosis is the development of computer-assisted tools. In this article, we present the solution that won the SegPC-2021 competition for the segmentation of multiple myeloma plasma cells in microscopy images. The labels used in the competition dataset were generated semi-automatically and presented noise. To deal with it, a heavy image augmentation procedure was carried out and predictions from several models were combined using a custom ensemble strategy. State-of-the-art feature extractors and instance segmentation architectures were used, resulting in a mean Intersection-over-Union of 0.9389 on the SegPC-2021 final test set.