Abstract:Self-supervised learning methods have witnessed a recent surge of interest after proving successful in multiple application fields. In this work, we leverage these techniques, and we propose 3D versions for five different self-supervised methods, in the form of proxy tasks. Our methods facilitate neural network feature learning from unlabeled 3D images, aiming to reduce the required cost for expert annotation. The developed algorithms are 3D Contrastive Predictive Coding, 3D Rotation prediction, 3D Jigsaw puzzles, Relative 3D patch location, and 3D Exemplar networks. Our experiments show that pretraining models with our 3D tasks yields more powerful semantic representations, and enables solving downstream tasks more accurately and efficiently, compared to training the models from scratch and to pretraining them on 2D slices. We demonstrate the effectiveness of our methods on three downstream tasks from the medical imaging domain: i) Brain Tumor Segmentation from 3D MRI, ii) Pancreas Tumor Segmentation from 3D CT, and iii) Diabetic Retinopathy Detection from 2D Fundus images. In each task, we assess the gains in data-efficiency, performance, and speed of convergence. Interestingly, we also find gains when transferring the learned representations, by our methods, from a large unlabeled 3D corpus to a small downstream-specific dataset. We achieve results competitive to state-of-the-art solutions at a fraction of the computational expense. We publish our implementations for the developed algorithms (both 3D and 2D versions) as an open-source library, in an effort to allow other researchers to apply and extend our methods on their datasets.

FocalMix: Semi-Supervised Learning for 3D Medical Image Detection

Focalmix: Semi-supervised learning for 3d medical image detection

Semi-supervised medical image classification based on CamMix

Detection of High-Low Risk Lung Tumors Using Semi-Supervised and Selective Labeling Techniques

Lesionmix Data Enhancement and Entropy Minimization for Semi-Supervised Lesion Segmentation of Lung Cancer

Confidence-Enhanced Semi-supervised Learning for Mediastinal Neoplasm Segmentation

3D Self-Supervised Methods for Medical Imaging

Semi-Supervised CT Lesion Segmentation Using Uncertainty-Based Data Pairing and SwapMix

Self-supervised Learning for Few-shot Medical Image Segmentation

Robust and accurate pulmonary nodule detection with self-supervised feature learning on domain adaptation

SM2C: Boost the Semi-supervised Segmentation for Medical Image by using Meta Pseudo Labels and Mixed Images

Medical Instrument Segmentation in 3D US by Hybrid Constrained Semi-Supervised Learning

Semi-Supervised Medical Image Classification Combined with Unsupervised Deep Clustering

Dual Uncertainty-Guided Mixing Consistency for Semi-Supervised 3D Medical Image Segmentation

Self-supervised learning via inter-modal reconstruction and feature projection networks for label-efficient 3D-to-2D segmentation

Semi-supervised learning for ABUS tumor detection using deep learning method

Semi-MedSeq: Semi-supervised Semantic Segmentation for Medical Image Sequences.

High-Fidelity Image Synthesis from Pulmonary Nodule Lesion Maps using Semantic Diffusion Model

Beyond Low-Dimensional Features: Enhancing Semi-Supervised Medical Image Semantic Segmentation with Advanced Consistency Learning Techniques

Low-Quality Sensor Data-Based Semi-Supervised Learning for Medical Image Segmentation

An Automatic Ultrasonic Segmentation Method by Two-stage Semi-supervised Learning Strategy