3D Deep Learning for Anatomical Structure Segmentation in Multiple Imaging Modalities

Abstract

Accurate, automated quantitative segmentation of anatomical structures in radiological scans, such as Magnetic Resonance Imaging (MRI) and Computer Tomography (CT), can produce significant biomarkers and can be integrated into computer-aided diagnosis (CADx) systems to support the interpretation of medical images from multi-protocol scanners. However, there are serious challenges towards developing robust automated segmentation techniques, including high variations in anatomical structure and size, varying image spatial resolutions resulting from different scanner protocols, and the presence of blurring artefacts. This paper presents a novel computing approach for automated organ and muscle segmentation in medical images from multiple modalities by harnessing the advantages of deep learning techniques in a two-part process. (1) a 3D encoder-decoder, Rb-UNet, builds a localisation model and a 3D Tiramisu network generates a boundary-preserving segmentation model for each target structure; (2) the fully trained Rb-UNet predicts a 3D bounding box encapsulating the target structure of interest, after which the fully trained Tiramisu model performs segmentation to reveal organ or muscle boundaries for every protrusion and indentation. The proposed approach is evaluated on six different datasets, including MRI, Dynamic Contrast Enhanced (DCE) MRI and CT scans targeting the pancreas, liver, kidneys and iliopsoas muscles. We achieve quantitative measures of mean Dice similarity coefficient (DSC) that surpasses or are comparable with the state-of-the-art and demonstrate statistical stability. A qualitative evaluation performed by two independent experts in radiology and radiography verified the preservation of detailed organ and muscle boundaries.

AI Driven IoT Web-Based Application for Automatic Segmentation and Reconstruction of Abdominal Organs from Medical Images

“…The main aim is to improve convergence through this bypass with identity connections for convolutional blocks at each scale. Empirically tested, the 3D Rb-UNet model performed significantly better than the standard 3D U-Net for organ localisation [11].…”

“…The algorithms used to perform automatic organ segmentation in medical volumes are based on 3D deep learning techniques that employ volumetric information instead of 2D pixel information, presented and evaluated in [11]. The developed framework has a two-part process: the first part develops a localisation model known as 3D Rb-UNet to "capture" the target organ of interest, and the second part performs detailed organ segmentation through a 3D Tiramisu network.…”

“…Empirically tested, this upsampling path built from dense blocks performs better than an upsampling path with conventional operations in DenseNet or U-Net. Further details about the network architecture can be found in [11].…”

“…2. Overview of the proposed automated organ segmentation approach [11]. The figure shows two different stages.…”

“…Fig.4. Overview of the proposed automated organ segmentation approach[11]. A single 2D slice in a medical volume is displayed in three different views (axial, sagittal and coronal) with over-imposed segmentation results.…”