A Study of Domain Generalization on Ultrasound-based Multi-Class Segmentation of Arteries, Veins, Ligaments, and Nerves Using Transfer Learning
This work addresses the problem of limited generalizability of deep learning models for ultrasound-based anatomical landmark segmentation, which is crucial for robot-guided catheter insertion, by studying fine-tuning strategies to improve performance across different ultrasound scanners and settings.
This paper investigates the domain generalization capabilities of deep learning models for multi-class segmentation of anatomical structures (arteries, veins, ligaments, nerves) in ultrasound images. The authors study the effects of fine-tuning different contiguous blocks within a model on performance across diverse ultrasound data from various scanners and settings, and propose a method for predicting generalization on unseen datasets.
Identifying landmarks in the femoral area is crucial for ultrasound (US) -based robot-guided catheter insertion, and their presentation varies when imaged with different scanners. As such, the performance of past deep learning-based approaches is also narrowly limited to the training data distribution; this can be circumvented by fine-tuning all or part of the model, yet the effects of fine-tuning are seldom discussed. In this work, we study the US-based segmentation of multiple classes through transfer learning by fine-tuning different contiguous blocks within the model, and evaluating on a gamut of US data from different scanners and settings. We propose a simple method for predicting generalization on unseen datasets and observe statistically significant differences between the fine-tuning methods while working towards domain generalization.