Self-Supervised Learning for 3D Medical Image Analysis using 3D SimCLR and Monte Carlo Dropout
This work addresses data scarcity in medical imaging for clinicians and researchers, but it is incremental as it adapts existing methods to 3D data and adds Bayesian enhancements.
The paper tackled the problem of reducing labeled data needs in 3D medical image segmentation by proposing a 3D self-supervised method based on SimCLR and using Monte Carlo Dropout during inference, resulting in improved data-efficiency and performance on brain and pancreas tumor segmentation tasks.
Self-supervised learning methods can be used to learn meaningful representations from unlabeled data that can be transferred to supervised downstream tasks to reduce the need for labeled data. In this paper, we propose a 3D self-supervised method that is based on the contrastive (SimCLR) method. Additionally, we show that employing Bayesian neural networks (with Monte-Carlo Dropout) during the inference phase can further enhance the results on the downstream tasks. We showcase our models on two medical imaging segmentation tasks: i) Brain Tumor Segmentation from 3D MRI, ii) Pancreas Tumor Segmentation from 3D CT. Our experimental results demonstrate the benefits of our proposed methods in both downstream data-efficiency and performance.