Abstract:Accurate segmentation of multiple targets, such as ribs, clavicles, heart, and lung fields, from chest X-ray images is crucial for diagnosing various lung diseases. Currently, mainstream deep learning methods heavily rely on fully annotated large-scale datasets. However, annotating all objects in chest X-ray images is a labor-intensive and time-consuming task. The publicly available partially annotated chest X-ray datasets have varying annotation objects and standards, and they are seldom utilized comprehensively in existing studies. To address these challenges, we propose A Multi-objective Segmentation Method for chest X-rays based on Collaborative Learning from Multiple Partially Annotated Datasets (called: MSM-CLMPAD). Our approach first utilizes an encoder constructed with densely connected blocks to extract multi-scale features from multiple partially annotated datasets. Then, leveraging the overlapping relationships among segmentation targets, we design a dual decoder guided by the attention mechanism. The novel attention-guided decoder effectively disentangles the features corresponding to various targets. Importantly, we propose an alternating training strategy for different datasets, facilitating collaborative learning of the same network model across datasets with diverse annotation targets. The experimental results performing on four public datasets demonstrate that our method achieves superior Dice and Jaccard coefficients compared to other popular methods, particularly for overlapping targets and unclear regions. We also explore the mutual influence of different targets in chest X-rays, offering a solution for interaction among partial datasets and further alleviating the difficulties of data annotation for multi-organ segmentation tasks.

Adaptive Annotation Correlation Based Multi-Annotation Learning for Calibrated Medical Image Segmentation

ACTION++: Improving Semi-supervised Medical Image Segmentation with Adaptive Anatomical Contrast

Transformer-based Annotation Bias-aware Medical Image Segmentation

Bootstrapping Semi-supervised Medical Image Segmentation with Anatomical-aware Contrastive Distillation

AHU-MultiNet: Adaptive loss balancing based on homoscedastic uncertainty in multi-task medical image segmentation network

Modeling Annotator Variation and Annotator Preference for Multiple Annotations Medical Image Segmentation.

Semi-supervised Medical Image Segmentation Based on Multi-scale Knowledge Discovery and Multi-task Ensemble.

Joint Learning Annotator Calibration and Annotator Preference for Multiple Annotations Optic Disc and Cup Segmentation.

A sparse annotation strategy based on attention-guided active learning for 3D medical image segmentation

Interactive medical image segmentation with self-adaptive confidence calibration

Predictive Accuracy-Based Active Learning for Medical Image Segmentation

Improving Segment Anything on the Fly: Auxiliary Online Learning and Adaptive Fusion for Medical Image Segmentation

Correlation-Aware Mutual Learning for Semi-supervised Medical Image Segmentation

Learning from AI-Generated Annotations for Medical Image Segmentation

Annotation-Efficient Task Guidance for Medical Segment Anything

Modeling Annotator Preference and Stochastic Annotation Error for Medical Image Segmentation

Combining Contrastive Learning and Shape Awareness for Semi-Supervised Medical Image Segmentation

Reliable Mutual Distillation for Medical Image Segmentation under Imperfect Annotations

A Multi-Objective Segmentation Method for Chest X-rays Based on Collaborative Learning from Multiple Partially Annotated Datasets

Learning Robust Medical Image Segmentation from Multi-source Annotations

3D Medical Image Segmentation with Sparse Annotation via Cross-Teaching between 3D and 2D Networks