Abstract:Objective - Medical image segmentation is essential for several clinical tasks, including diagnosis, surgical and treatment planning, and image-guided interventions. Deep Learning (DL) methods have become the state-of-the-art for several image segmentation scenarios. However, a large and well-annotated dataset is required to effectively train a DL model, which is usually difficult to obtain in clinical practice, especially for 3D images. Methods - In this paper, we proposed Deep-DM, a learning-guided deformable model framework for 3D medical imaging segmentation using limited training data. In the proposed method, an energy function is learned by a Convolutional Neural Network (CNN) and integrated into an explicit deformable model to drive the evolution of an initial surface towards the object to segment. Specifically, the learning-based energy function is iteratively retrieved from localized anatomical representations of the image containing the image information around the evolving surface at each iteration. By focusing on localized regions of interest, this representation excludes irrelevant image information, facilitating the learning process. Results and conclusion - The performance of the proposed method is demonstrated for the tasks of left ventricle and fetal head segmentation in ultrasound, left atrium segmentation in Magnetic Resonance, and bladder segmentation in Computed Tomography, using different numbers of training volumes in each study. The results obtained showed the feasibility of the proposed method to segment different anatomical structures in different imaging modalities. Moreover, the results also showed that the proposed approach is less dependent on the size of the training dataset in comparison with state-of-the-art DL-based segmentation methods, outperforming them for all tasks when a low number of samples is available. Significance - Overall, by offering a more robust and less data-intensive approach to accurately segmenting anatomical structures, the proposed method has the potential to enhance clinical tasks that require image segmentation strategies.

DDeep3M: Docker-powered Deep Learning for Biomedical Image Segmentation

A Lightweight Deep Network for 3D Medical Image Segmentation.

Deep-DM: Deep-driven deformable model for 3D image segmentation using limited data

DDeep3M-based Neuronal Cell Counting in 2D Large-scale Images

DeepS: a Web Server for Image Optical Sectioning and Super Resolution Microscopy Based on a Deep Learning Framework.

Breast Lesion Segmentation in Ultrasound Images by CDeep3M

DCSegNet: Deep Learning Framework Based on Divide-and-Conquer Method for Liver Segmentation

DeepBrainSeg: Automated Brain Region Segmentation for Micro-Optical Images with a Convolutional Neural Network

DDeep3M+: Adaptive Enhancement Powered Weakly Supervised Learning for Neuron Segmentation.

Automated 3D Fetal Brain Segmentation Using an Optimized Deep Learning Approach

DeepMRSeg: A convolutional deep neural network for anatomy and abnormality segmentation on MR images

Deep Learning Based Instance Segmentation in 3D Biomedical Images Using Weak Annotation

SAM-Med3D-MoE: Towards a Non-Forgetting Segment Anything Model via Mixture of Experts for 3D Medical Image Segmentation

Benchmarking of deep learning algorithms for 3D instance segmentation of confocal image datasets

MIDeepSeg: Minimally interactive segmentation of unseen objects from medical images using deep learning

Advanced Deep Networks for 3D Mitochondria Instance Segmentation

Deep Neural Network-Based Semantic Segmentation of Microvascular Decompression Images

Deep learning-based breast region extraction of mammographic images combining pre-processing methods and semantic segmentation supported by Deeplab v3+

SAM-Med3D: Towards General-purpose Segmentation Models for Volumetric Medical Images

SAM3D: Segment Anything Model in Volumetric Medical Images

DeepMapi: a Fully Automatic Registration Method for Mesoscopic Optical Brain Images Using Convolutional Neural Networks