Three-Dimensional Lumbosacral Reconstruction by An Artificial Intelligence-Based Automated MR Image Segmentation for Selecting the Approach of Percutaneous Endoscopic Lumbar Discectomy.

BACKGROUND: Assessing the 3-dimensional (3D) relationship between critical anatomical structures and the surgical channel can help select percutaneous endoscopic lumbar discectomy (PELD) approaches, especially at the L5/S1 level. However, previous evaluation methods for PELD were mainly assessed using 2-dimensional (2D) medical images, making the understanding of the 3D relationship of lumbosacral structures difficult. Artificial intelligence based on automated magnetic resonance (MR) image segmentation has the benefit of 3D reconstruction of medical images. OBJECTIVES: We developed and validated an artificial intelligence-based MR image segmentation method for constructing a 3D model of lumbosacral structures for selecting the appropriate approach of percutaneous endoscopic lumbar discectomy at the L5/S1 level. STUDY DESIGN: Three-dimensional reconstruction study using artificial intelligence based on MR image segmentation. SETTING: Spine and radiology center of a university hospital. METHODS: Fifty MR data samples were used to develop an artificial intelligence algorithm for automatic segmentation. Manual segmentation and labeling of vertebrae bone (L5 and S1 vertebrae bone), disc, lumbosacral nerve, iliac bone, and skin at the L5/S1 level by 3 experts were used as ground truth. Five-fold cross-validation was performed, and quantitative segmentation metrics were used to evaluate the performance of artificial intelligence based on the MR image segmentation method. The comparison analysis of quantitative measurements between the artificial intelligence-derived 3D (AI-3D) models and the ground truth-derived 3D (GT-3D) models was used to validate the feasibility of 3D lumbosacral structures reconstruction and preoperative assessment of PELD approaches. RESULTS: Artificial intelligence-based automated MR image segmentation achieved high mean Dice Scores of 0.921, 0.924, 0.885, 0.808, 0.886, and 0.816 for L5 vertebrae bone, S1 vertebrae bone, disc, lumbosacral nerves, iliac bone, and skin, respectively. There were no significant differences between AI-3D and GT-3D models in quantitative measurements. Comparative analysis of quantitative measures showed a high correlation and consistency. LIMITATIONS: Our method did not involve vessel segmentation in automated MR image segmentation. Our study's sample size was small, and the findings need to be validated in a prospective study with a large sample size. CONCLUSION: We developed an artificial intelligence-based automated MR image segmentation method, which effectively segmented lumbosacral structures (e.g., L5 vertebrae bone, S1 vertebrae bone, disc, lumbosacral nerve, iliac bone, and skin) simultaneously on MR images, and could be used to construct a 3D model of lumbosacral structures for choosing an appropriate approach of PELD at the L5/S1 level.

Cross-Domain Unpaired Learning for Low-Dose CT Imaging.

Supervised deep-learning techniques with paired training datasets have been widely studied for low-dose computed tomography (LDCT) imaging with excellent performance. However, the paired training datasets are usually difficult to obtain in clinical routine, which restricts the wide adoption of supervised deep-learning techniques in clinical practices. To address this issue, a general idea is to construct a pseudo paired training dataset based on the widely available unpaired data, after which, supervised deep-learning techniques can be adopted for improving the LDCT imaging performance by training on the pseudo paired training dataset. However, due to the complexity of noise properties in CT imaging, the LDCT data are difficult to generate in order to construct the pseudo paired training dataset. In this article, we propose a simple yet effective cross-domain unpaired learning framework for pseudo LDCT data generation and LDCT image reconstruction, which is denoted as CrossDuL. Specifically, a dedicated pseudo LDCT sinogram generative module is constructed based on a data-dependent noise model in the sinogram domain, and then instead of in the sinogram domain, a pseudo paired dataset is constructed in the image domain to train an LDCT image restoration module. To validate the effectiveness of the proposed framework, clinical datasets are adopted. Experimental results demonstrate that the CrossDuL framework can obtain promising LDCT imaging performance in both quantitative and qualitative measurements.

Deep Learning-Based Automated Magnetic Resonance Image Segmentation of the Lumbar Structure and Its Adjacent Structures at the L4/5 Level.

(1) Background: This study aims to develop a deep learning model based on a 3D Deeplab V3+ network to automatically segment multiple structures from magnetic resonance (MR) images at the L4/5 level. (2) Methods: After data preprocessing, the modified 3D Deeplab V3+ network of the deep learning model was used for the automatic segmentation of multiple structures from MR images at the L4/5 level. We performed five-fold cross-validation to evaluate the performance of the deep learning model. Subsequently, the Dice Similarity Coefficient (DSC), precision, and recall were also used to assess the deep learning model's performance. Pearson's correlation coefficient analysis and the Wilcoxon signed-rank test were employed to compare the morphometric measurements of 3D reconstruction models generated by manual and automatic segmentation. (3) Results: The deep learning model obtained an overall average DSC of 0.886, an average precision of 0.899, and an average recall of 0.881 on the test sets. Furthermore, all morphometry-related measurements of 3D reconstruction models revealed no significant difference between ground truth and automatic segmentation. Strong linear relationships and correlations were also obtained in the morphometry-related measurements of 3D reconstruction models between ground truth and automated segmentation. (4) Conclusions: We found it feasible to perform automated segmentation of multiple structures from MR images, which would facilitate lumbar surgical evaluation by establishing 3D reconstruction models at the L4/5 level.

SpineRegNet: Spine Registration Network for volumetric MR and CT image by the joint estimation of an affine-elastic deformation field.

Spine registration for volumetric magnetic resonance (MR) and computed tomography (CT) images plays a significant role in surgical planning and surgical navigation system for the radiofrequency ablation of spine intervertebral discs. The affine transformation of each vertebra and elastic deformation of the intervertebral disc exist at the same time. This situation is a major challenge in spine registration. Existing spinal image registration methods failed to solve the optimal affine-elastic deformation field (AEDF) simultaneously, only consider the overall rigid or elastic alignment with the help of a manual spine mask, and encounter difficulty in meeting the accuracy requirements of clinical registration application. In this study, we propose a novel affine-elastic registration framework named SpineRegNet. The SpineRegNet consists of a Multiple Affine Matrices Estimation (MAME) Module for multiple vertebrae alignment, an Affine-Elastic Fusion (AEF) Module for joint estimation of the overall AEDF, and a Local Rigidity Constraint (LRC) Module for preserving the rigidity of each vertebra. Experiments on T2-weighted volumetric MR and CT images show that the proposed approach achieves impressive performance with mean Dice similarity coefficients of 91.36%, 81.60%, and 83.08% for the mask of the vertebrae on Datasets A-C, respectively. The proposed technique does not require a mask or manual participation during the tests and provides a useful tool for clinical spinal disease surgical planning and surgical navigation systems.

Automated measurement of spine indices on axial MR images for lumbar spinal stenosis diagnosis using segmentation-guided regression network.

PURPOSE: Automated measurement of spine indices on axial magnetic resonance (MR) images plays a significant role in lumbar spinal stenosis diagnosis. Existing direct spine indices measurement approaches fail to explicitly focus on the task-specific region or feature channel with the additional information for guiding. We aim to achieve accurate spine indices measurement by introducing the guidance of the segmentation task. METHODS: In this paper, we propose a segmentation-guided regression network (SGRNet) to achieve automated spine indices measurement. SGRNet consists of a segmentation path for generating the spine segmentation prediction and a regression path for producing spine indices estimation. The segmentation path is a U-Net-like network which includes a segmentation encoder and a decoder which generates multilevel segmentation features and segmentation prediction. The proposed segmentation-guided attention module (SGAM) in the regression encoder extracts the attention-aware regression feature under the guidance of the segmentation feature. Based on the attention-aware regression feature, a fully connected layer is utilized to output the accurate spine indices estimation. RESULTS: Experiments on the open-access Lumbar Spine MRI data set show that SGRNet achieves state-of-the-art performance with a mean absolute error of 0.49 mm and mean Pearson correlation coefficient of 0.956 for four indices estimation. CONCLUSIONS: The proposed SGAM in SGRNet is capable of improving the performance of spine indices measurement by focusing on the task-specific region and feature channel under the guidance of the segmentation task.

PCAN: Pixel-wise classification and attention network for thoracic disease classification and weakly supervised localization.

Automatic chest X-ray (CXR) disease classification has drawn increasing public attention as CXR is widely used in thoracic disease diagnosis. Existing classification networks typically employ a global average pooling layer to produce the final feature for the subsequent classifier. This limits the classification performance owing to the characteristics of lesions in CXR images, including small relative sizes, varied absolute sizes, and different occurrence locations. In this study, we propose a pixel-wise classification and attention network (PCAN) to simultaneously perform disease classification and weakly supervised localization, which provides interpretability for disease classification. The PCAN comprises a backbone network for extracting mid-level features, a pixel-wise classification branch (pc-branch) for generating pixel-wise diagnoses, and a pixel-wise attention branch (pa-branch) for producing pixel-wise weights. The pc-branch is capable of explicitly detecting small lesions, and the pa-branch is capable of adaptively focusing on different regions when classifying different thoracic diseases. Then, the pixel-wise diagnoses are multiplied with the pixel-wise weights to obtain the disease localization map, which provides the sizes and locations of lesions in a manner of weakly supervised learning. The final image-wise diagnosis is obtained by summing up the disease localization map at the spatial dimension. Comprehensive experiments conducted on the ChestX-ray14 and CheXpert datasets demonstrate the effectiveness of the proposed PCAN, which has great potential for thoracic disease diagnosis and treatment. The source codes are available at https://github.com/fzfs/PCAN.

Automated Magnetic Resonance Image Segmentation of Spinal Structures at the L4-5 Level with Deep Learning: 3D Reconstruction of Lumbar Intervertebral Foramen.

OBJECTIVE: 3D reconstruction of lumbar intervertebral foramen (LIVF) has been beneficial in evaluating surgical trajectory. Still, the current methods of reconstructing the 3D LIVF model are mainly based on manual segmentation, which is laborious and time-consuming. This study aims to explore the feasibility of automatically segmenting lumbar spinal structures and increasing the speed and accuracy of 3D lumbar intervertebral foramen (LIVF) reconstruction on magnetic resonance image (MRI) at the L4-5 level. METHODS: A total of 100 participants (mean age: 42.2 ± 14.0 years; 52 males and 48 females; mean body mass index, 22.7 ± 3.2 kg/m2 ), were enrolled in this prospective study between March and July 2020. All participants were scanned on L4-5 level with a 3T MR unit using 3D T2-weighted sampling perfection with application-optimized contrast with various flip-angle evolutions (SPACE) sequences. The lumbar spine's vertebra bone structures (VBS) and intervertebral discs (IVD) were manually segmented by skilled surgeons according to their anatomical outlines from MRI. Then all manual segmentation were saved and used for training. An automated segmentation method based on a 3D U-shaped architecture network (3D-UNet) was introduced for the automated segmentation of lumbar spinal structures. A number of quantitative metrics, including dice similarity coefficient (DSC), precision, and recall, were used to evaluate the performance of the automated segmentation method on MRI. Wilcoxon signed-rank test was applied to compare morphometric parameters, including foraminal area, height and width of 3D LIVF models between automatic and manual segmentation. The intra-class correlation coefficient was used to assess the test-retest reliability and inter-observer reliability of multiple measurements for these morphometric parameters of 3D LIVF models. RESULTS: The automatic segmentation performance of all spinal structures (VBS and IVD) was found to be 0.918 (healthy levels: 0.922; unhealthy levels: 0.916) for the mean DSC, 0.922 (healthy levels: 0.927; unhealthy levels: 0.920) for the mean precision, and 0.917 (healthy levels: 0.918; unhealthy levels: 0.917) for the mean recall in the test dataset. It took approximately 2.5 s to achieve each automated segmentation, far less than the 240 min for manual segmentation. Furthermore, no significant differences were observed in the foraminal area, height and width of the 3D LIVF models between manual and automatic segmentation images (P > 0.05). CONCLUSION: A method of automated MRI segmentation based on deep learning algorithms was capable of rapidly generating accurate segmentation of spinal structures and can be used to construct 3D LIVF models from MRI at the L4-5 level.

Automatic Grading of Disc Herniation, Central Canal Stenosis and Nerve Roots Compression in Lumbar Magnetic Resonance Image Diagnosis.

Aim: Accurate severity grading of lumbar spine disease by magnetic resonance images (MRIs) plays an important role in selecting appropriate treatment for the disease. However, interpreting these complex MRIs is a repetitive and time-consuming workload for clinicians, especially radiologists. Here, we aim to develop a multi-task classification model based on artificial intelligence for automated grading of lumbar disc herniation (LDH), lumbar central canal stenosis (LCCS) and lumbar nerve roots compression (LNRC) at lumbar axial MRIs. Methods: Total 15254 lumbar axial T2W MRIs as the internal dataset obtained from the Fifth Affiliated Hospital of Sun Yat-sen University from January 2015 to May 2019 and 1273 axial T2W MRIs as the external test dataset obtained from the Third Affiliated Hospital of Southern Medical University from June 2016 to December 2017 were analyzed in this retrospective study. Two clinicians annotated and graded all MRIs using the three international classification systems. In agreement, these results served as the reference standard; In disagreement, outcomes were adjudicated by an expert surgeon to establish the reference standard. The internal dataset was randomly split into an internal training set (70%), validation set (15%) and test set (15%). The multi-task classification model based on ResNet-50 consists of a backbone network for feature extraction and three fully-connected (FC) networks for classification and performs the classification tasks of LDH, LCCS, and LNRC at lumbar MRIs. Precision, accuracy, sensitivity, specificity, F1 scores, confusion matrices, receiver-operating characteristics and interrater agreement (Gwet k) were utilized to assess the model's performance on the internal test dataset and external test datasets. Results: A total of 1115 patients, including 1015 patients from the internal dataset and 100 patients from the external test dataset [mean age, 49 years ± 15 (standard deviation); 543 women], were evaluated in this study. The overall accuracies of grading for LDH, LCCS and LNRC were 84.17% (74.16%), 86.99% (79.65%) and 81.21% (74.16%) respectively on the internal (external) test dataset. Internal and external testing of three spinal diseases showed substantial to the almost perfect agreement (k, 0.67 - 0.85) for the multi-task classification model. Conclusion: The multi-task classification model has achieved promising performance in the automated grading of LDH, LCCS and LNRC at lumbar axial T2W MRIs.

Improving sensitivity and connectivity of retinal vessel segmentation via error discrimination network.

PURPOSE: Automated retinal vessel segmentation is crucial to the early diagnosis and treatment of ophthalmological diseases. Many deep-learning-based methods have shown exceptional success in this task. However, current approaches are still inadequate in challenging vessels (e.g., thin vessels) and rarely focus on the connectivity of vessel segmentation. METHODS: We propose using an error discrimination network (D) to distinguish whether the vessel pixel predictions of the segmentation network (S) are correct, and S is trained to obtain fewer error predictions of D. Our method is similar to, but not the same as, the generative adversarial network. Three types of vessel samples and corresponding error masks are used to train D, as follows: (1) vessel ground truth; (2) vessel segmented by S; (3) artificial thin vessel error samples that further improve the sensitivity of D to wrong small vessels. As an auxiliary loss function of S, D strengthens the supervision of difficult vessels. Optionally, we can use the errors predicted by D to correct the segmentation result of S. RESULTS: Compared with state-of-the-art methods, our method achieves the highest scores in sensitivity (86.19%, 86.26%, and 86.53%) and G-Mean (91.94%, 91.30%, and 92.76%) on three public datasets, namely, STARE, DRIVE, and HRF. Our method also maintains a competitive level in other metrics. On the STARE dataset, the F1-score and area under the receiver operating characteristic curve (AUC) of our method rank second and first, respectively, reaching 84.51% and 98.97%. The top scores of the three topology-relevant metrics (Conn, Inf, and Cor) demonstrate that the vessels extracted by our method have excellent connectivity. We also validate the effectiveness of error discrimination supervision and artificial error sample training through ablation experiments. CONCLUSIONS: The proposed method provides an accurate and robust solution for difficult vessel segmentation.

Three-dimensional reconstruction of Kambin's triangle based on automated magnetic resonance image segmentation.

The three-dimensional (3D) anatomy of Kambin's triangle is crucial for surgical planning in minimally invasive spine surgery via the transforaminal approach. Few pieces of research have, however, used image segmentation to explore the 3D reconstruction of Kambin's triangle. This study aimed to develop a new method of 3D reconstruction of Kambin's triangle based on automated magnetic resonance image (MRI) segmentation of the lumbar spinal structures. An experienced (>5 years) "ground truth" spinal pain physician meticulously segmented and labeled spinal structures (e.g., bones, dura mater, discs, and nerve roots) on MRI. Subsequently, a 3D U-Net algorithm was developed for automatically segmenting lumbar spinal structures for the 3D reconstruction of Kambin's triangle. The Dice similarity coefficient (DSC), precision, recall, and the area of Kambin's triangle were used to assess anatomical performance. The automatic segmentation of all spinal structures at the L4/L5 levels and L5/S1 levels resulted in good performance: DSC = 0.878/0.883, precision = 0.889/0.890, recall = 0.873/0.882. Furthermore, the area measurements of Kambin's triangle revealed no significant difference between ground truth and automatic segmentation (p = 0.333 at the L4/L5 level, p = 0.302 at the L5/S1 level). The 3D U-Net model used in this study performed well in terms of simultaneous segmentation of multi-class spinal structures (including bones, dura mater, discs, and nerve roots) on MRI, allowing for accurate 3D reconstruction of Kambin's triangle.

Segmentation of the True Lumen of Aorta Dissection via Morphology-Constrained Stepwise Deep Mesh Regression.

The lumen of aortic dissection (AD) has important clinical value for preoperative diagnosis, interoperative intervention, and post-operative evaluation of AD diseases. AD segmentation is challenging because (i) fitting its irregular profile by using traditional models is difficult, and (ii) the size of the AD image is usually so big that many algorithms have to perform down-sampling to reduce the computational burden, thereby reducing the resolution of the result. In this paper, an automatic AD segmentation algorithm, in which a 3D mesh is gradually moved to the surface of AD based on the offset estimated by a deep mesh deformation module, is presented. AD morphology is used to constrain the initial mesh and guide the deformation, which improves the efficiency of the deep network and avoids down-sampling. Moreover, a stepwise regression strategy is introduced to solve the mesh folding problem and improve the uniformity of the mesh points. On an AD database that involves 35 images, the proposed method obtains the mean Dice of 94.12% and symmetric 95% Hausdorff distance of 2.85 mm, which outperforms five state-of-the-art AD segmentation methods. The average processing time is 16.6 s, and the memory used to train the network is only 0.36 GB, indicating that this method is easy to apply in clinical practice.

Computerized Characterization of Spinal Structures on MRI and Clinical Significance of 3D Reconstruction of Lumbosacral Intervertebral Foramen.

BACKGROUND: Segmentation of spinal structures is important in medical imaging analysis, which facilitates surgeons to plan a preoperative trajectory for the transforaminal approach. However, manual segmentation of spinal structures is time-consuming, and studies have not explored automatic segmentation of spinal structures at the L5/S1 level. OBJECTIVES: This study sought to develop a new method based on a deep learning algorithm for automatic segmentation of spinal structures. The resulting algorithm may be used to rapidly generate a precise 3D lumbosacral intervertebral foramen model to assist physicians in planning an ideal trajectory in L5/S1 lumbar transforaminal radiofrequency ablation (LTRFA). STUDY DESIGN: This was an observational study for developing a new technique on spinal structures segmentation. STUDY SITE: The study was carried out at the department of radiology and spine surgery at our hospital. METHODS: A total of 100 L5/S1 level data samples from 100 study patients were used in this study. Masks of vertebral bone structures (VBSs) and intervertebral discs (IVDs) for all data samples were segmented manually by a skilled surgeon and served as the "ground truth." After data preprocessing, a 3D-UNet model based on deep learning was used for automated segmentation of lumbar spine structures at L5/S1 level magnetic resonance imaging (MRI). Segmentation performances and morphometric measurement were used for 3D lumbosacral intervertebral foramen (LIVF) reconstruction  generated by either manual segmentation and automatic segmentation. RESULTS: The 3D-UNet model showed high performance in automatic segmentation of lumbar spinal structures (VBSs and IVDs). The corresponding mean Dice similarity coefficient (DSC) of 5-fold cross-validation scores for L5 vertebrae, IVDs, S1 vertebrae, and all L5/S1 level spinal structures were 93.46 ± 2.93%, 90.39 ± 6.22%, 93.32 ± 1.51%, and 92.39 ± 2.82%, respectively. Notably, the analysis showed no associated difference in morphometric measurements between the manual and automatic segmentation at the L5/S1 level. LIMITATIONS: Semantic segmentation of multiple spinal structures (such as VBSs, IVDs, blood vessels, muscles, and ligaments) was simultaneously not integrated into the deep-learning method in this study. In addition, large clinical experiments are needed to evaluate the clinical efficacy of the model. CONCLUSION: The 3D-UNet model developed in this study based on deep learning can effectively and simultaneously segment VBSs and IVDs at L5/S1 level formMR images, thereby enabling rapid and accurate 3D reconstruction of LIVF models. The method can be used to segment VBSs and IVDs of spinal structures on MR images within near-human expert performance; therefore, it is reliable for reconstructing LIVF for L5/S1 LTRFA.

DGMSNet: Spine segmentation for MR image by a detection-guided mixed-supervised segmentation network.

Spine segmentation for magnetic resonance (MR) images is important for various spinal diseases diagnosis and treatment, yet is still a challenge due to the inter-class similarity, i.e., shape and appearance similarities appear in neighboring spinal structures. To reduce inter-class similarity, existing approaches focus on enhancing the semantic information of spinal structures in the supervised segmentation network, whose generalization is limited by the size of pixel-level annotated dataset. In this paper, we propose a novel detection-guided mixed-supervised segmentation network (DGMSNet) to achieve automated spine segmentation. DGMSNet consists of a segmentation path for generating the spine segmentation prediction and a detection path (i.e., regression network) for producing heatmaps prediction of keypoints. A detection-guided learner in the detection path is introduced to generate a dynamic parameter, which is employed to produce a semantic feature map for segmentation path by adaptive convolution. A mixed-supervised loss comprised of a weighted combination of segmentation loss and detection loss is utilized to train DGMSNet with a pixel-level annotated dataset and a keypoints-detection annotated dataset. During training, a series of models are trained with various loss weights. In inference, a detection-guided label fusion approach is proposed to integrate the segmentation predictions generated by those trained models according to the consistency of predictions from the segmentation path and detection path. Experiments on T2-weighted MR images show that DGMSNet achieves the state-of-the-art performance with mean Dice similarity coefficients of 94.39% and 87.21% for segmentations of 5 vertebral bodies and 5 intervertebral discs on the in-house and public datasets respectively.

Quantifying Axial Spine Images Using Object-Specific Bi-Path Network.

Automatic estimation of indices from medical images is the main goal of computer-aided quantification (CADq), which speeds up diagnosis and lightens the workload of radiologists. Deep learning technique is a good choice for implementing CADq. Usually, to acquire high-accuracy quantification, specific network architecture needs to be designed for a given CADq task. In this study, considering that the target organs are the intervertebral disc and the dural sac, we propose an object-specific bi-path network (OSBP-Net) for axial spine image quantification. Each path of the OSBP-Net comprises a shallow feature extraction layer (SFE) and a deep feature extraction sub-network (DFE). The SFEs use different convolution strides because the two target organs have different anatomical sizes. The DFEs use average pooling for downsampling based on the observation that the target organs have lower intensity than the background. In addition, an inter-path dissimilarity constraint is proposed and applied to the output of the SFEs, taking into account that the activated regions in the feature maps of two paths should be different theoretically. An inter-index correlation regularization is introduced and applied to the output of the DFEs based on the observation that the diameter and area of the same object express an approximately linear relation. The prediction results of OSBP-Net are compared to several state-of-the-art machine learning-based CADq methods. The comparison reveals that the proposed methods precede other competing methods extensively, indicating its great potential for spine CADq.

SpineParseNet: Spine Parsing for Volumetric MR Image by a Two-Stage Segmentation Framework With Semantic Image Representation.

Spine parsing (i.e., multi-class segmentation of vertebrae and intervertebral discs (IVDs)) for volumetric magnetic resonance (MR) image plays a significant role in various spinal disease diagnoses and treatments of spine disorders, yet is still a challenge due to the inter-class similarity and intra-class variation of spine images. Existing fully convolutional network based methods failed to explicitly exploit the dependencies between different spinal structures. In this article, we propose a novel two-stage framework named SpineParseNet to achieve automated spine parsing for volumetric MR images. The SpineParseNet consists of a 3D graph convolutional segmentation network (GCSN) for 3D coarse segmentation and a 2D residual U-Net (ResUNet) for 2D segmentation refinement. In 3D GCSN, region pooling is employed to project the image representation to graph representation, in which each node representation denotes a specific spinal structure. The adjacency matrix of the graph is designed according to the connection of spinal structures. The graph representation is evolved by graph convolutions. Subsequently, the proposed region unpooling module re-projects the evolved graph representation to a semantic image representation, which facilitates the 3D GCSN to generate reliable coarse segmentation. Finally, the 2D ResUNet refines the segmentation. Experiments on T2-weighted volumetric MR images of 215 subjects show that SpineParseNet achieves impressive performance with mean Dice similarity coefficients of 87.32 ± 4.75%, 87.78 ± 4.64%, and 87.49 ± 3.81% for the segmentations of 10 vertebrae, 9 IVDs, and all 19 spinal structures respectively. The proposed method has great potential in clinical spinal disease diagnoses and treatments.

Multiple Axial Spine Indices Estimation via Dense Enhancing Network With Cross-Space Distance-Preserving Regularization.

Automatic estimation of axial spine indices is clinically desired for various spine computer aided procedures, such as disease diagnosis, therapeutic evaluation, pathophysiological understanding, risk assessment, and biomechanical modeling. Currently, the spine indices are manually measured by physicians, which is time-consuming and laborious. Even worse, the tedious manual procedure might result in inaccurate measurement. To deal with this problem, in this paper, we aim at developing an automatic method to estimate multiple indices from axial spine images. Inspired by the success of deep learning for regression problems and the densely connected network for image classification, we propose a dense enhancing network (DE-Net) which uses the dense enhancing blocks (DEBs) as its main body, where a feature enhancing layer is added to each of the bypass in a dense block. The DEB is designed to enhance discriminative feature embedding from the intervertebral disc and the dural sac areas. In addition, the cross-space distance-preserving regularization (CSDPR), which enforces consistent inter-sample distances between the output and the label spaces, is proposed to regularize the loss function of the DE-Net. To train and validate the proposed method, we collected 895 axial spine MRI images from 143 subjects and manually measured the indices as the ground truth. The results show that all deep learning models obtain very small prediction errors, and the proposed DE-Net with CSDPR acquires the smallest error among all methods, indicating that our method has great potential for spine computer aided procedures.

Direct automated quantitative measurement of spine by cascade amplifier regression network with manifold regularization.

Automated quantitative measurement of the spine (i.e., multiple indices estimation of heights, widths, areas, and so on for the vertebral body and disc) plays a significant role in clinical spinal disease diagnoses and assessments, such as osteoporosis, intervertebral disc degeneration, and lumbar disc herniation, yet still an unprecedented challenge due to the variety of spine structure and the high dimensionality of indices to be estimated. In this paper, we propose a novel cascade amplifier regression network (CARN) with manifold regularization including local structure-preserved manifold regularization (LSPMR) and adaptive local shape-constrained manifold regularization (ALSCMR), to achieve accurate direct automated multiple indices estimation. The CARN architecture is composed of a cascade amplifier network (CAN) for expressive feature embedding and a linear regression model for multiple indices estimation. The CAN produces an expressive feature embedding by cascade amplifier units (AUs), which are used for selective feature reuse by stimulating effective feature and suppressing redundant feature during propagating feature map between adjacent layers. During training, the LSPMR is employed to obtain discriminative feature embedding by preserving the local geometric structure of the latent feature space similar to the target output manifold. The ALSCMR is utilized to alleviate overfitting and generate realistic estimation by learning the multiple indices distribution. Experiments on T1-weighted MR images of 215 subjects and T2-weighted MR images of 20 subjects show that the proposed approach achieves impressive performance with mean absolute errors of 1.22 ±â€¯1.04 mm and 1.24 ±â€¯1.07 mm for the 30 lumbar spinal indices estimation of the T1-weighted and T2-weighted spinal MR images respectively. The proposed method has great potential in clinical spinal disease diagnoses and assessments.

Hippocampus Segmentation Based on Iterative Local Linear Mapping With Representative and Local Structure-Preserved Feature Embedding.

Hippocampus segmentation plays a significant role in mental disease diagnoses, such as Alzheimer's disease, epilepsy, and so on. Patch-based multi-atlas segmentation (PBMAS) approach is a popular method for hippocampus segmentation and has achieved a promising result. However, the PBMAS approach needs high computation cost due to registration and the segmentation accuracy is subject to the registration accuracy. In this paper, we propose a novel method based on iterative local linear mapping (ILLM) with the representative and local structure-preserved feature embedding to achieve accurate and robust hippocampus segmentation with no need for registration. In the proposed approach, semi-supervised deep autoencoder (SSDA) exploits unsupervised deep autoencoder and local structure-preserved manifold regularization to nonlinearly transform the extracted magnetic resonance (MR) patch to embedded feature manifold, whose adjacent relationship is similar to the signed distance map (SDM) patch manifold. Local linear mapping is used to preliminarily predict SDM patch corresponding to the MR patch. Subsequently, threshold segmentation generates a preliminary segmentation. The ILLM refines the segmentation result iteratively by ensuring the local constraints of embedded feature manifold and SDM patch manifold using a space-constrained dictionary update. Thus, a refined segmentation is obtained with no need for registration. The experiments on 135 subjects from ADNI dataset show that the proposed approach is superior to the state-of-the-art PBMAS and classification-based approaches with mean Dice similarity coefficients of 0.8852±0.0203 and 0.8783 ± 0.0251 for bilateral hippocampus segmentation of 1.5T and 3.0T datasets, respectively.

[Segmentation of the prostate on magnetic resonance images using an ellipsoidal shape prior constraint algorithm].

We propose a novel strategy for multi-atlas-based image segmentation of the prostate on magnetic resonance (MR) images using an ellipsoidal shape prior constraint algorithm. An ellipsoidal shape prior constraint was incorporated into the process of multi-atlas based segmentation to restrict the regions of interest on the prostate images and avoid the interference by the surrounding tissues and organs in atlas selection. In the subsequent process of atlas fusion, the ellipsoidal shape prior constraint calibrated and compensated for the shape prior obtained by the registration technique to avoid incorrect segmentation caused by registration errors. Evaluation of this proposed method on prostate images from 50 subjects showed that this algorithm was effective and yielded a mean Dice similarity coefficients of 0.8812, suggesting its high accuracy and robustness to segment the prostate on MR images.

Hippocampus Segmentation Based on Local Linear Mapping.

We propose local linear mapping (LLM), a novel fusion framework for distance field (DF) to perform automatic hippocampus segmentation. A k-means cluster method is propose for constructing magnetic resonance (MR) and DF dictionaries. In LLM, we assume that the MR and DF samples are located on two nonlinear manifolds and the mapping from the MR manifold to the DF manifold is differentiable and locally linear. We combine the MR dictionary using local linear representation to present the test sample, and combine the DF dictionary using the corresponding coefficients derived from local linear representation procedure to predict the DF of the test sample. We then merge the overlapped predicted DF patch to obtain the DF value of each point in the test image via a confidence-based weighted average method. This approach enabled us to estimate the label of the test image according to the predicted DF. The proposed method was evaluated on brain images of 35 subjects obtained from SATA dataset. Results indicate the effectiveness of the proposed method, which yields mean Dice similarity coefficients of 0.8697, 0.8770 and 0.8734 for the left, right and bi-lateral hippocampus, respectively.