# A Knowledge Distillation framework for Multi-Organ Segmentation of Medaka Fish in Tomographic Image

## TL;DR

This paper introduces a self-training knowledge distillation framework for multi-organ segmentation in Medaka fish tomographic images, reducing annotation effort while improving segmentation accuracy.

## Contribution

It presents a novel self-training approach with a quality classifier and pixel-wise knowledge distillation for efficient multi-organ segmentation.

## Key findings

- Improves mean IoU by 5.9% on the full dataset.
- Reduces annotation effort by using three times less markup.
- Enhances segmentation performance with pseudo-label refinement.

## Abstract

Morphological atlases are an important tool in organismal studies, and modern high-throughput Computed Tomography (CT) facilities can produce hundreds of full-body high-resolution volumetric images of organisms. However, creating an atlas from these volumes requires accurate organ segmentation. In the last decade, machine learning approaches have achieved incredible results in image segmentation tasks, but they require large amounts of annotated data for training. In this paper, we propose a self-training framework for multi-organ segmentation in tomographic images of Medaka fish. We utilize the pseudo-labeled data from a pretrained Teacher model and adopt a Quality Classifier to refine the pseudo-labeled data. Then, we introduce a pixel-wise knowledge distillation method to prevent overfitting to the pseudo-labeled data and improve the segmentation performance. The experimental results demonstrate that our method improves mean Intersection over Union (IoU) by 5.9% on the full dataset and enables keeping the quality while using three times less markup.