Spine-GAN: Semantic segmentation of multiple spinal structures

A novel segmentation method for cervical vertebrae based on PointNet++ and converge segmentation

BACKGROUND

Cervical spine instability is the key pathogenic factor for cervical spondylosis, which may easily cause cervical spinal cord nerve compression, numbness, weakness, and even paralysis of the limbs. The reconstruction of the internal fixation of the cervical spine is of great therapeutic significance, but is a high-risk and difficult procedure that requires precise planning. The high similarities between vertebrae may interfere with automatic operation planning; therefore, the segmentation of vertebrae is of great significance.

METHODS

Our segmentation algorithm has 3 parts. Firstly, an adaptive threshold filter to segment the cervical vertebra tissue structure form CT images. Secondly, segmentation of single vertebrae based on PointNet++ is introduced to segmentation cervical spine. Finally, converge segmentation which is based on edge information is utilized to clearly distinguish the edges of the two vertebrae to enhance the accuracy segmentation result.

RESULTS

Our approach improved the accuracy of the system up to 96.15%, and achieved the highest reported average score based on this dataset. We compared the results of the CNN and PointNet methods on a separate dataset of 240 CT scans with 18 classes and achieved a significantly higher performance for any given vertebra. Our experiments illustrated the promise and robustness of recent PointNet++-based segmentation of medical images.

CONCLUSION

The proposed method has better classification performance for segmentation cervical spine images, which segment a three-dimensional vertebral body directly and effectively. Furthermore, the precise segmentation of a single vertebral body can be used in automatic biomechanical analysis, computer-aided diagnosis and other aspects, so as to improve the level of automation in the treatment of cervical spondylosis.

mustGAN: multi-stream Generative Adversarial Networks for MR Image Synthesis

Multi-contrast MRI protocols increase the level of morphological information available for diagnosis. Yet, the number and quality of contrasts are limited in practice by various factors including scan time and patient motion. Synthesis of missing or corrupted contrasts from other high-quality ones can alleviate this limitation. When a single target contrast is of interest, common approaches for multi-contrast MRI involve either one-to-one or many-to-one synthesis methods depending on their input. One-to-one methods take as input a single source contrast, and they learn a latent representation sensitive to unique features of the source. Meanwhile, many-to-one methods receive multiple distinct sources, and they learn a shared latent representation more sensitive to common features across sources. For enhanced image synthesis, we propose a multi-stream approach that aggregates information across multiple source images via a mixture of multiple one-to-one streams and a joint many-to-one stream. The complementary feature maps generated in the one-to-one streams and the shared feature maps generated in the many-to-one stream are combined with a fusion block. The location of the fusion block is adaptively modified to maximize task-specific performance. Quantitative and radiological assessments on T₁,- T₂-, PD-weighted, and FLAIR images clearly demonstrate the superior performance of the proposed method compared to previous state-of-the-art one-to-one and many-to-one methods.

The Clinical Value of CT Scans for the Early Diagnosis and Treatment of Spinal Fractures and Paraplegia

The early diagnosis and treatment of spinal fractures and paraplegia by CT scan is investigated in depth and its clinical value is discussed in this paper. In this paper, a novel circulatory generation adversarial network, Spine-GAN, is proposed for the diagnosis of various spinal diseases. The algorithmic model can fully automate the segmentation and classification of multiple spinal structures, such as intervertebral discs, vertebrae, and neuroforamina, simultaneously to intelligently generate a complete clinical diagnosis. The innovation of this method is that Spine-GAN not only overcomes the high variability and complexity of spinal structures in MRI images but also preserves the subtle differences between normal and abnormal spinal structures and dynamically learns obscure but important spatial pathological relationships between adjacent structures of the spine, thus overcoming the limitations of small datasets. Spine-GAN enables accurate segmentation, radiological classification, and pathological correlation representation of the three spinal diseases. Specifically, Spine-GAN achieves a pixel accuracy of 96.2% with a specificity and sensitivity distribution of 89.1% and 86%, respectively. The DMML-Net and Spine-GAN algorithm models have important applications and research values in the clinical diagnosis of spinal diseases and MRI image processing, as well as in the intelligent generation of medical image diagnostic reports, which are of great importance for the study of fine-grained image classification of pathological images. It also has a positive impact on the development of the software.

Automated Pipeline to Generate Anatomically Accurate Patient-Specific Biomechanical Models of Healthy and Pathological FSUs

State-of-the-art preoperative biomechanical analysis for the planning of spinal surgery not only requires the generation of three-dimensional patient-specific models but also the accurate biomechanical representation of vertebral joints. The benefits offered by computational models suitable for such purposes are still outweighed by the time and effort required for their generation, thus compromising their applicability in a clinical environment. In this work, we aim to ease the integration of computerized methods into patient-specific planning of spinal surgery. We present the first pipeline combining deep learning and finite element methods that allows a completely automated model generation of functional spine units (FSUs) of the lumbar spine for patient-specific FE simulations (FEBio). The pipeline consists of three steps: (a) multiclass segmentation of cropped 3D CT images containing lumbar vertebrae using the DenseVNet network, (b) automatic landmark-based mesh fitting of statistical shape models onto 3D semantic segmented meshes of the vertebral models, and (c) automatic generation of patient-specific FE models of lumbar segments for the simulation of flexion-extension, lateral bending, and axial rotation movements. The automatic segmentation of FSUs was evaluated against the gold standard (manual segmentation) using 10-fold cross-validation. The obtained Dice coefficient was 93.7% on average, with a mean surface distance of 0.88 mm and a mean Hausdorff distance of 11.16 mm (N = 150). Automatic generation of finite element models to simulate the range of motion (ROM) was successfully performed for five healthy and five pathological FSUs. The results of the simulations were evaluated against the literature and showed comparable ROMs in both healthy and pathological cases, including the alteration of ROM typically observed in severely degenerated FSUs. The major intent of this work is to automate the creation of anatomically accurate patient-specific models by a single pipeline allowing functional modeling of spinal motion in healthy and pathological FSUs. Our approach reduces manual efforts to a minimum and the execution of the entire pipeline including simulations takes approximately 2 h. The automation, time-efficiency and robustness level of the pipeline represents a first step toward its clinical integration.

A Review of Deep-Learning-Based Medical Image Segmentation Methods

As an emerging biomedical image processing technology, medical image segmentation has made great contributions to sustainable medical care. Now it has become an important research direction in the field of computer vision. With the rapid development of deep learning, medical image processing based on deep convolutional neural networks has become a research hotspot. This paper focuses on the research of medical image segmentation based on deep learning. First, the basic ideas and characteristics of medical image segmentation based on deep learning are introduced. By explaining its research status and summarizing the three main methods of medical image segmentation and their own limitations, the future development direction is expanded. Based on the discussion of different pathological tissues and organs, the specificity between them and their classic segmentation algorithms are summarized. Despite the great achievements of medical image segmentation in recent years, medical image segmentation based on deep learning has still encountered difficulties in research. For example, the segmentation accuracy is not high, the number of medical images in the data set is small and the resolution is low. The inaccurate segmentation results are unable to meet the actual clinical requirements. Aiming at the above problems, a comprehensive review of current medical image segmentation methods based on deep learning is provided to help researchers solve existing problems.

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.

Self-channel-and-spatial-attention neural network for automated multi-organ segmentation on head and neck CT images

Accurate segmentation of organs at risk (OARs) is necessary for adaptive head and neck (H&N) cancer treatment planning, but manual delineation is tedious, slow, and inconsistent. A self-channel-and-spatial-attention neural network (SCSA-Net) is developed for H&N OAR segmentation on CT images. To simultaneously ease the training and improve the segmentation performance, the proposed SCSA-Net utilizes the self-attention ability of the network. Spatial and channel-wise attention learning mechanisms are both employed to adaptively force the network to emphasize the meaningful features and weaken the irrelevant features simultaneously. The proposed network was first evaluated on a public dataset, which includes 48 patients, then on a separate serial CT dataset, which contains ten patients who received weekly diagnostic fan-beam CT scans. On the second dataset, the accuracy of using SCSA-Net to track the parotid and submandibular gland volume changes during radiotherapy treatment was quantified. The Dice similarity coefficient (DSC), positive predictive value (PPV), sensitivity (SEN), average surface distance (ASD), and 95% maximum surface distance (95SD) were calculated on the brainstem, optic chiasm, optic nerves, mandible, parotid glands, and submandibular glands to evaluate the proposed SCSA-Net. The proposed SCSA-Net consistently outperforms the state-of-the-art methods on the public dataset. Specifically, compared with Res-Net and SE-Net, which is constructed from squeeze-and-excitation block equipped residual blocks, the DSC of the optic nerves and submandibular glands is improved by 0.06, 0.03 and 0.05, 0.04 by the SCSA-Net. Moreover, the proposed method achieves statistically significant improvements in terms of DSC on all and eight of nine OARs over Res-Net and SE-Net, respectively. The trained network was able to achieve good segmentation results on the serial dataset, but the results were further improved after fine-tuning of the model using the simulation CT images. For the parotids and submandibular glands, the volume changes of individual patients are highly consistent between the automated and manual segmentation (Pearson's correlation 0.97-0.99). The proposed SCSA-Net is computationally efficient to perform segmentation (sim 2 s/CT).

Injury classification and level detection of the spinal cord based on the optimized recurrent neural network

Objectives

Spinal cord damage is one of the traumatic situations in persons that may cause the loss of sensation and proper functioning of the muscles either temporarily or permanently. Hence, steps to assure the recovery through the early functioning and precaution could safe-guard a proper interceptive. To ensure the recovery of spinal cord damage through optimized recurrent neural network.

Methods

The research on the spinal cord injury classification and level detection is done using the CT images, which is initially given to the segmentation that is done using the adaptive thresholding methodology. Once the segments are formed, the disc is localized using the sparse fuzzy C-means clustering approach. In the next step, the features are extracted from the localized disc and the features include the connectivity features, statistical features, image-level features, grid-level features, Histogram of Oriented Gradients (HOG), and Linear Gradient Pattern (LGP). Then, the injury detection is done based on the Crow search Rider Optimization algorithm-based Deep Convolutional Neural Network (CS-ROA-based DCNN). Once the result regarding the presence of the injury is obtained, the injury-level classification is done based on the proposed Deep Recurrent Neural Network (Deep RNN), and in case of the absence of injury, the process is terminated. Therefore, the injury detection classifier derives the level of the injury, such as normal, wedge, biconcavity, and crush.

Results

The experimentation is carried out using an Osteoporotic vertebral fractures database. The performance of the injury level detection based on the proposed model is evaluated based on accuracy, sensitivity, and specificity. The proposed model achieves the maximal accuracy of 0.895, maximal sensitivity of 0.871, and the maximal specificity of 0.933 with respect to K-Fold.

Conclusions

The experimental results show that the proposed model is better than the existing models in terms of accuracy, sensitivity, and specificity.

Automated left ventricular segmentation from cardiac magnetic resonance images via adversarial learning with multi-stage pose estimation network and co-discriminator

Left ventricular (LV) segmentation is essential for the early diagnosis of cardiovascular diseases, which has been reported as the leading cause of death all over the world. However, automated LV segmentation from cardiac magnetic resonance images (CMRI) using the traditional convolutional neural networks (CNNs) is still a challenging task due to the limited labeled CMRI data and low tolerances to irregular scales, shapes and deformations of LV. In this paper, we propose an automated LV segmentation method based on adversarial learning by integrating a multi-stage pose estimation network (MSPN) and a co-discrimination network. Different from existing CNNs, we use a MSPN with multi-scale dilated convolution (MDC) modules to enhance the ranges of receptive field for deep feature extraction. To fully utilize both labeled and unlabeled CMRI data, we propose a novel generative adversarial network (GAN) framework for LV segmentation by combining MSPN with co-discrimination networks. Specifically, the labeled CMRI are first used to initialize our segmentation network (MSPN) and co-discrimination network. Our GAN training includes two different kinds of epochs fed with both labeled and unlabeled CMRI data alternatively, which are different from the traditional CNNs only relied on the limited labeled samples to train the segmentation networks. As both ground truth and unlabeled samples are involved in guiding training, our method not only can converge faster but also obtain a better performance in LV segmentation. Our method is evaluated using MICCAI 2009 and 2017 challenge databases. Experimental results show that our method has obtained promising performance in LV segmentation, which also outperforms the state-of-the-art methods in terms of LV segmentation accuracy from the comparison results.

Unifying neural learning and symbolic reasoning for spinal medical report generation

Automated medical report generation in spine radiology, i.e., given spinal medical images and directly create radiologist-level diagnosis reports to support clinical decision making, is a novel yet fundamental study in the domain of artificial intelligence in healthcare. However, it is incredibly challenging because it is an extremely complicated task that involves visual perception and high-level reasoning processes. In this paper, we propose the neural-symbolic learning (NSL) framework that performs human-like learning by unifying deep neural learning and symbolic logical reasoning for the spinal medical report generation. Generally speaking, the NSL framework firstly employs deep neural learning to imitate human visual perception for detecting abnormalities of target spinal structures. Concretely, we design an adversarial graph network that interpolates a symbolic graph reasoning module into a generative adversarial network through embedding prior domain knowledge, achieving semantic segmentation of spinal structures with high complexity and variability. NSL secondly conducts human-like symbolic logical reasoning that realizes unsupervised causal effect analysis of detected entities of abnormalities through meta-interpretive learning. NSL finally fills these discoveries of target diseases into a unified template, successfully achieving a comprehensive medical report generation. When employed in a real-world clinical dataset, a series of empirical studies demonstrate its capacity on spinal medical report generation and show that our algorithm remarkably exceeds existing methods in the detection of spinal structures. These indicate its potential as a clinical tool that contributes to computer-aided diagnosis.

Sequential conditional reinforcement learning for simultaneous vertebral body detection and segmentation with modeling the spine anatomy

Accurate vertebral body (VB) detection and segmentation are critical for spine disease identification and diagnosis. Existing automatic VB detection and segmentation methods may cause false-positive results to the background tissue or inaccurate results to the desirable VB. Because they usually cannot take both the global spine pattern and the local VB appearance into consideration concurrently. In this paper, we propose a Sequential Conditional Reinforcement Learning network (SCRL) to tackle the simultaneous detection and segmentation of VBs from MR spine images. The SCRL, for the first time, applies deep reinforcement learning into VB detection and segmentation. It innovatively models the spatial correlation between VBs from top to bottom as sequential dynamic-interaction processes, thereby globally focusing detection and segmentation on each VB. Simultaneously, SCRL also perceives the local appearance feature of each desirable VB comprehensively, thereby achieving accurate detection and segmentation result. Particularly, SCRL seamlessly combines three parts: 1) Anatomy-Modeling Reinforcement Learning Network dynamically interacts with the image and focuses an attention-region on the VB; 2) Fully-Connected Residual Neural Network learns rich global context information of the VB including both the detailed low-level features and the abstracted high-level features to detect the accurate bounding-box of the VB based on the attention-region; 3) Y-shaped Network learns comprehensive detailed texture information of VB including multi-scale, coarse-to-fine features to segment the boundary of VB from the attention-region. On 240 subjects, SCRL achieves accurate detection and segmentation results, where on average the detection IoU is 92.3%, segmentation Dice is 92.6%, and classification mean accuracy is 96.4%. These excellent results demonstrate that SCRL can be an efficient aided-diagnostic tool to assist clinicians when diagnosing spinal diseases.

FocusNetv2: Imbalanced large and small organ segmentation with adversarial shape constraint for head and neck CT images

Radiotherapy is a treatment where radiation is used to eliminate cancer cells. The delineation of organs-at-risk (OARs) is a vital step in radiotherapy treatment planning to avoid damage to healthy organs. For nasopharyngeal cancer, more than 20 OARs are needed to be precisely segmented in advance. The challenge of this task lies in complex anatomical structure, low-contrast organ contours, and the extremely imbalanced size between large and small organs. Common segmentation methods that treat them equally would generally lead to inaccurate small-organ labeling. We propose a novel two-stage deep neural network, FocusNetv2, to solve this challenging problem by automatically locating, ROI-pooling, and segmenting small organs with specifically designed small-organ localization and segmentation sub-networks while maintaining the accuracy of large organ segmentation. In addition to our original FocusNet, we employ a novel adversarial shape constraint on small organs to ensure the consistency between estimated small-organ shapes and organ shape prior knowledge. Our proposed framework is extensively tested on both self-collected dataset of 1,164 CT scans and the MICCAI Head and Neck Auto Segmentation Challenge 2015 dataset, which shows superior performance compared with state-of-the-art head and neck OAR segmentation methods.

Recovering from missing data in population imaging - Cardiac MR image imputation via conditional generative adversarial nets

Accurate ventricular volume measurements are the primary indicators of normal/abnor- mal cardiac function and are dependent on the Cardiac Magnetic Resonance (CMR) volumes being complete. However, missing or unusable slices owing to the presence of image artefacts such as respiratory or motion ghosting, aliasing, ringing and signal loss in CMR sequences, significantly hinder accuracy of anatomical and functional cardiac quantification, and recovering from those is insufficiently addressed in population imaging. In this work, we propose a new robust approach, coined Image Imputation Generative Adversarial Network (I2-GAN), to learn key features of cardiac short axis (SAX) slices near missing information, and use them as conditional variables to infer missing slices in the query volumes. In I2-GAN, the slices are first mapped to latent vectors with position features through a regression net. The latent vector corresponding to the desired position is then projected onto the slice manifold, conditioned on intensity features through a generator net. The generator comprises residual blocks with normalisation layers that are modulated with auxiliary slice information, enabling propagation of fine details through the network. In addition, a multi-scale discriminator was implemented, along with a discriminator-based feature matching loss, to further enhance performance and encourage the synthesis of visually realistic slices. Experimental results show that our method achieves significant improvements over the state-of-the-art, in missing slice imputation for CMR, with an average SSIM of 0.872. Linear regression analysis yields good agreement between reference and imputed CMR images for all cardiac measurements, with correlation coefficients of 0.991 for left ventricular volume, 0.977 for left ventricular mass and 0.961 for right ventricular volume.

MB-FSGAN: Joint segmentation and quantification of kidney tumor on CT by the multi-branch feature sharing generative adversarial network

The segmentation of the kidney tumor and the quantification of its tumor indices (i.e., the center point coordinates, diameter, circumference, and cross-sectional area of the tumor) are important steps in tumor therapy. These quantifies the tumor morphometrical details to monitor disease progression and accurately compare decisions regarding the kidney tumor treatment. However, manual segmentation and quantification is a challenging and time-consuming process in practice and exhibit a high degree of variability within and between operators. In this paper, MB-FSGAN (multi-branch feature sharing generative adversarial network) is proposed for simultaneous segmentation and quantification of kidney tumor on CT. MB-FSGAN consists of multi-scale feature extractor (MSFE), locator of the area of interest (LROI), and feature sharing generative adversarial network (FSGAN). MSFE makes strong semantic information on different scale feature maps, which is particularly effective in detecting small tumor targets. The LROI extracts the region of interest of the tumor, greatly reducing the time complexity of the network. FSGAN correctly segments and quantifies kidney tumors through joint learning and adversarial learning, which effectively exploited the commonalities and differences between the two related tasks. Experiments are performed on CT of 113 kidney tumor patients. For segmentation, MB-FSGAN achieves a pixel accuracy of 95.7%. For the quantification of five tumor indices, the R² coefficient of tumor circumference is 0.9465. The results show that the network has reliable performance and shows its effectiveness and potential as a clinical tool.

Adversarial convolutional network for esophageal tissue segmentation on OCT images

Automatic segmentation is important for esophageal OCT image processing, which is able to provide tissue characteristics such as shape and thickness for disease diagnosis. Existing automatical segmentation methods based on deep convolutional networks may not generate accurate segmentation results due to limited training set and various layer shapes. This study proposed a novel adversarial convolutional network (ACN) to segment esophageal OCT images using a convolutional network trained by adversarial learning. The proposed framework includes a generator and a discriminator, both with U-Net alike fully convolutional architecture. The discriminator is a hybrid network that discriminates whether the generated results are real and implements pixel classification at the same time. Leveraging on the adversarial training, the discriminator becomes more powerful. In addition, the adversarial loss is able to encode high order relationships of pixels, thus eliminating the requirements of post-processing. Experiments on segmenting esophageal OCT images from guinea pigs confirmed that the ACN outperforms several deep learning frameworks in pixel classification accuracy and improves the segmentation result. The potential clinical application of ACN for detecting eosinophilic esophagitis (EoE), an esophageal disease, is also presented in the experiment.

Adversarial convolutional network for esophageal tissue segmentation on OCT images

Automatic segmentation is important for esophageal OCT image processing, which is able to provide tissue characteristics such as shape and thickness for disease diagnosis. Existing automatical segmentation methods based on deep convolutional networks may not generate accurate segmentation results due to limited training set and various layer shapes. This study proposed a novel adversarial convolutional network (ACN) to segment esophageal OCT images using a convolutional network trained by adversarial learning. The proposed framework includes a generator and a discriminator, both with U-Net alike fully convolutional architecture. The discriminator is a hybrid network that discriminates whether the generated results are real and implements pixel classification at the same time. Leveraging on the adversarial training, the discriminator becomes more powerful. In addition, the adversarial loss is able to encode high order relationships of pixels, thus eliminating the requirements of post-processing. Experiments on segmenting esophageal OCT images from guinea pigs confirmed that the ACN outperforms several deep learning frameworks in pixel classification accuracy and improves the segmentation result. The potential clinical application of ACN for detecting eosinophilic esophagitis (EoE), an esophageal disease, is also presented in the experiment.

A Review on the Use of Artificial Intelligence in Spinal Diseases

Artificial neural networks (ANNs) have been used in a wide variety of real-world applications and it emerges as a promising field across various branches of medicine. This review aims to identify the role of ANNs in spinal diseases. Literature were searched from electronic databases of Scopus and Medline from 1993 to 2020 with English publications reported on the application of ANNs in spinal diseases. The search strategy was set as the combinations of the following keywords: “artificial neural networks,” “spine,” “back pain,” “prognosis,” “grading,” “classification,” “prediction,” “segmentation,” “biomechanics,” “deep learning,” and “imaging.” The main findings of the included studies were summarized, with an emphasis on the recent advances in spinal diseases and its application in the diagnostic and prognostic procedures. According to the search strategy, a set of 3,653 articles were retrieved from Medline and Scopus databases. After careful evaluation of the abstracts, the full texts of 89 eligible papers were further examined, of which 79 articles satisfied the inclusion criteria of this review. Our review indicates several applications of ANNs in the management of spinal diseases including (1) diagnosis and assessment of spinal disease progression in the patients with low back pain, perioperative complications, and readmission rate following spine surgery; (2) enhancement of the clinically relevant information extracted from radiographic images to predict Pfirrmann grades, Modic changes, and spinal stenosis grades on magnetic resonance images automatically; (3) prediction of outcomes in lumbar spinal stenosis, lumbar disc herniation and patient-reported outcomes in lumbar fusion surgery, and preoperative planning and intraoperative assistance; and (4) its application in the biomechanical assessment of spinal diseases. The evidence suggests that ANNs can be successfully used for optimizing the diagnosis, prognosis and outcome prediction in spinal diseases. Therefore, incorporation of ANNs into spine clinical practice may improve clinical decision making.

MRLN: Multi-Task Relational Learning Network for MRI Vertebral Localization, Identification, and Segmentation

Magnetic resonance imaging (MRI) vertebral localization, identification, and segmentation are important steps in the automatic analysis of spines. Due to the similar appearances of vertebrae, the accurate segmentation, localization, and identification of vertebrae remain challenging. Previous methods solved the three tasks independently, ignoring the intrinsic correlation among them. In this paper, we propose a multi-task relational learning network (MRLN) that utilizes both the relationships between vertebrae and the relevance of the three tasks. A dilation convolution group is used to expand the receptive field, and LSTM(Long Short-Term Memory) to learn the prior knowledge of the order relationship between the vertebral bodies. We introduce a co-attention module to learn the correlation information, localization-guided segmentation attention(LGSA) and segmentation-guided localization attention(SGLA), in the decoder stage of segmentation and localization tasks. Learning two tasks simultaneously as well as the correlation between tasks can not only avoid the overfitting of a single task but also correct each other. To avoids the cumbersome weight adjustment for different tasks loss functions, we formulated a novel XOR loss that provides a direct evaluation criterion for the localization relationship of the semantic location regression and semantic segmentation. This method was evaluated on a dataset which includes multiple MRI modalities (T1 and T2), various fields of view. Experimental results demonstrate that both of the co-attention and XOR loss work outperforms the most recent state of art.

Toward Automated 3D Spine Reconstruction from Biplanar Radiographs Using CNN for Statistical Spine Model Fitting

To date, 3D spine reconstruction from biplanar radiographs involves intensive user supervision and semi-automated methods that are time-consuming and not effective in clinical routine. This paper proposes a new, fast, and automated 3D spine reconstruction method through which a realistic statistical shape model of the spine is fitted to images using convolutional neural networks (CNN). The CNNs automatically detect the anatomical landmarks controlling the spine model deformation through a hierarchical and gradual iterative process. The performance assessment used a set of 68 biplanar radiographs, composed of both asymptomatic subjects and adolescent idiopathic scoliosis patients, in order to compare automated reconstructions with ground truths build using multiple experts-supervised reconstructions. The mean (SD) errors of landmark locations (3D Euclidean distances) were 1.6 (1.3) mm, 1.8 (1.3) mm, and 2.3 (1.4) mm for the vertebral body center, endplate centers, and pedicle centers, respectively. The clinical parameters extracted from the automated 3D reconstruction (reconstruction time is less than one minute) presented an absolute mean error between 2.8° and 4.7° for the main spinal parameters and between 1° and 2.1° for pelvic parameters. Automated and expert's agreement analysis reported that, on average, 89% of automated measurements were inside the expert's confidence intervals. The proposed automated 3D spine reconstruction method provides an important step that should help the dissemination and adoption of 3D measurements in clinical routine.

Active Appearance Model Induced Generative Adversarial Network for Controlled Data Augmentation

Data augmentation is an important strategy for enlarging training datasets in deep learning-based medical image analysis. This is because large, annotated medical datasets are not only difficult and costly to generate, but also quickly become obsolete due to rapid advances in imaging technology. Image-to-image conditional generative adversarial networks (C-GAN) provide a potential solution for data augmentation. However, annotations used as inputs to C-GAN are typically based only on shape information, which can result in undesirable intensity distributions in the resulting artificially-created images. In this paper, we introduce an active cell appearance model (ACAM) that can measure statistical distributions of shape and intensity and use this ACAM model to guide C-GAN to generate more realistic images, which we call A-GAN. A-GAN provides an effective means for conveying anisotropic intensity information to C-GAN. A-GAN incorporates a statistical model (ACAM) to determine how transformations are applied for data augmentation. Traditional approaches for data augmentation that are based on arbitrary transformations might lead to unrealistic shape variations in an augmented dataset that are not representative of real data. A-GAN is designed to ameliorate this. To validate the effectiveness of using A-GAN for data augmentation, we assessed its performance on cell analysis in adaptive optics retinal imaging, which is a rapidly-changing medical imaging modality. Compared to C-GAN, A-GAN achieved stability in fewer iterations. The cell detection and segmentation accuracy when assisted by A-GAN augmentation was higher than that achieved with C-GAN. These findings demonstrate the potential for A-GAN to substantially improve existing data augmentation methods in medical image analysis.

Generative adversarial network in medical imaging: A review

Generative adversarial networks have gained a lot of attention in the computer vision community due to their capability of data generation without explicitly modelling the probability density function. The adversarial loss brought by the discriminator provides a clever way of incorporating unlabeled samples into training and imposing higher order consistency. This has proven to be useful in many cases, such as domain adaptation, data augmentation, and image-to-image translation. These properties have attracted researchers in the medical imaging community, and we have seen rapid adoption in many traditional and novel applications, such as image reconstruction, segmentation, detection, classification, and cross-modality synthesis. Based on our observations, this trend will continue and we therefore conducted a review of recent advances in medical imaging using the adversarial training scheme with the hope of benefiting researchers interested in this technique.

Shape constrained fully convolutional DenseNet with adversarial training for multiorgan segmentation on head and neck CT and low-field MR images

PURPOSE

Image-guided radiotherapy provides images not only for patient positioning but also for online adaptive radiotherapy. Accurate delineation of organs-at-risk (OARs) on Head and Neck (H&N) CT and MR images is valuable to both initial treatment planning and adaptive planning, but manual contouring is laborious and inconsistent. A novel method based on the generative adversarial network (GAN) with shape constraint (SC-GAN) is developed for fully automated H&N OARs segmentation on CT and low-field MRI.

METHODS AND MATERIAL

A deep supervised fully convolutional DenseNet is employed as the segmentation network for voxel-wise prediction. A convolutional neural network (CNN)-based discriminator network is then utilized to correct predicted errors and image-level inconsistency between the prediction and ground truth. An additional shape representation loss between the prediction and ground truth in the latent shape space is integrated into the segmentation and adversarial loss functions to reduce false positivity and constrain the predicted shapes. The proposed segmentation method was first benchmarked on a public H&N CT database including 32 patients, and then on 25 0.35T MR images obtained from an MR-guided radiotherapy system. The OARs include brainstem, optical chiasm, larynx (MR only), mandible, pharynx (MR only), parotid glands (both left and right), optical nerves (both left and right), and submandibular glands (both left and right, CT only). The performance of the proposed SC-GAN was compared with GAN alone and GAN with the shape constraint (SC) but without the DenseNet (SC-GAN-ResNet) to quantify the contributions of shape constraint and DenseNet in the deep neural network segmentation.

RESULTS

The proposed SC-GAN slightly but consistently improve the segmentation accuracy on the benchmark H&N CT images compared with our previous deep segmentation network, which outperformed other published methods on the same or similar CT H&N dataset. On the low-field MR dataset, the following average Dice's indices were obtained using improved SC-GAN: 0.916 (brainstem), 0.589 (optical chiasm), 0.816 (mandible), 0.703 (optical nerves), 0.799 (larynx), 0.706 (pharynx), and 0.845 (parotid glands). The average surface distances ranged from 0.68 mm (brainstem) to 1.70 mm (larynx). The 95% surface distance ranged from 1.48 mm (left optical nerve) to 3.92 mm (larynx). Compared with CT, using 95% surface distance evaluation, the automated segmentation accuracy is higher on MR for the brainstem, optical chiasm, optical nerves and parotids, and lower for the mandible. The SC-GAN performance is superior to SC-GAN-ResNet, which is more accurate than GAN alone on both the CT and MR datasets. The segmentation time for one patient is 14 seconds using a single GPU.

CONCLUSION

The performance of our previous shape constrained fully CNNs for H&N segmentation is further improved by incorporating GAN and DenseNet. With the novel segmentation method, we showed that the low-field MR images acquired on a MR-guided radiation radiotherapy system can support accurate and fully automated segmentation of both bony and soft tissue OARs for adaptive radiotherapy.

Deep Learning in the Biomedical Applications: Recent and Future Status

Deep neural networks represent, nowadays, the most effective machine learning technology in biomedical domain. In this domain, the different areas of interest concern the Omics (study of the genome—genomics—and proteins—transcriptomics, proteomics, and metabolomics), bioimaging (study of biological cell and tissue), medical imaging (study of the human organs by creating visual representations), BBMI (study of the brain and body machine interface) and public and medical health management (PmHM). This paper reviews the major deep learning concepts pertinent to such biomedical applications. Concise overviews are provided for the Omics and the BBMI. We end our analysis with a critical discussion, interpretation and relevant open challenges.