High Segmentation Accuracy Research Articles

Feasibility study of adaptive radiotherapy with Ethos for breast cancer.

The aim of this study was to assess the feasibility of online adaptive radiotherapy with Ethos for breast cancer. This retrospective study included 20 breast cancer patients previously treated with TrueBeam. All had undergone breast surgery for different indications (right/left, lumpectomy/mastectomy) and were evenly divided between these four cases, with five extended cone beam computed tomography (CBCT) scans per patient. The dataset was used in an Ethos emulator to test the full adaptive workflow. The contours generated by artificial intelligence (AI) for the influencers (left and right breasts and lungs, heart) and elastic or rigid propagation for the target volumes (internal mammary chain (IMC) and clavicular lymph nodes (CLNs)) were compared to the initial contours delineated by the physician using two metrics: Dice similarity coefficient (DICE) and Hausdorff 95% distance (HD95). The repeatability of influencer generation was investigated. The times taken by the emulator to generate contours, optimize plans, and calculate doses were recorded. The quality of the scheduled and adapted plans generated by Ethos was assessed using planning target volume (PTV) coverage, homogeneity indices (HIs), and doses to organs at risk (OARs) via dose-volume histogram (DVH) metrics. Quality assurance (QA) of the treatment plans was performed using an independent portal dosimetry tool (EpiQA) and gamma index. On average, the DICE for the influencers was greater than 0.9. Contours resulting from rigid propagation had a higher DICE and a lower HD95 than those resulting from elastic deformation but remained below the values obtained for the influencers: DICE values were 0.79 ± 0.11 and 0.46 ± 0.17 for the CLN and IMC, respectively. Regarding the repeatability of the influencer segmentation, the DICE was close to 1, and the mean HD95 was strictly less than 0.15mm. The mean time was 73 ± 4 s for contour generation per AI and 80 ± 9 s for propagations. The average time was 53 ± 3 s for dose calculation and 125 ± 9 s for plan optimization. A dosimetric comparison of scheduled and adapted plans showed a significant difference in PTV coverage: dose received by 95% of the volume (D95%) values were higher and closer to the prescribed doses for adapted plans. Doses to organs at risk were similar. The average gamma index for quality assurance of adapted plans was 99.93 ± 0.38 for a 3%/3mm criterion. This study comprehensively evaluated the Ethos® adaptive workflow for breast cancer and its potential technical limitations. Although the results demonstrated the high accuracy of AI segmentation and the superiority of adapted plans in terms of target volume coverage, a medical assessment is still required.

Detection of breakage and impurity ratios for raw sugarcane based on estimation model and MDSC-DeepLabv3.

Broken cane and impurities such as top, leaf in harvested raw sugarcane significantly influence the yield of the sugar manufacturing process. It is crucial to determine the breakage and impurity ratios for assessing the quality and price of raw sugarcane in sugar refineries. However, the traditional manual sampling approach for detecting breakage and impurity ratios suffers from subjectivity, low efficiency, and result discrepancies. To address this problem, a novel approach combining an estimation model and semantic segmentation method for breakage and impurity ratios detection was developed. A machine vision-based image acquisition platform was designed, and custom image and mass datasets of cane, broken cane, top, and leaf were created. For cane, broken cane, top, and leaf, normal fitting of mean surface densities based on pixel information and measured mass was conducted. An estimation model for the mass of each class and the breakage and impurity ratios was established using the mean surface density and pixels. Furthermore, the MDSC-DeepLabv3+ model was developed to accurately and efficiently segment pixels of the four classes of objects. This model integrates improved MobileNetv2, atrous spatial pyramid pooling with deepwise separable convolution and strip pooling module, and coordinate attention mechanism to achieve high segmentation accuracy, deployability, and efficiency simultaneously. Experimental results based on the custom image and mass datasets showed that the estimation model achieved high accuracy for breakage and impurity ratios between estimated and measured value with R2 values of 0.976 and 0.968, respectively. MDSC-DeepLabv3+ outperformed the compared models with mPA and mIoU of 97.55% and 94.84%, respectively. Compared to the baseline DeepLabv3+, MDSC-DeepLabv3+ demonstrated significant improvements in mPA and mIoU and reduced Params, FLOPs, and inference time, making it suitable for deployment on edge devices and real-time inference. The average relative errors of breakage and impurity ratios between estimated and measured values were 11.3% and 6.5%, respectively. Overall, this novel approach enables high-precision, efficient, and intelligent detection of breakage and impurity ratios for raw sugarcane.

Deep learning-based workflow for hip joint morphometric parameter measurement from CT images

Objective. Precise hip joint morphometry measurement from CT images is crucial for successful preoperative arthroplasty planning and biomechanical simulations. Although deep learning approaches have been applied to clinical bone surgery planning, there is still a lack of relevant research on quantifying hip joint morphometric parameters from CT images. Approach. This paper proposes a deep learning workflow for CT-based hip morphometry measurement. For the first step, a coarse-to-fine deep learning model is designed for accurate reconstruction of the hip geometry (3D bone models and key landmark points). Based on the geometric models, a robust measurement method is developed to calculate a full set of morphometric parameters, including the acetabular anteversion and inclination, the femoral neck shaft angle and the inclination, etc. Our methods were validated on two datasets with different imaging protocol parameters and further compared with the conventional 2D x-ray-based measurement method. Main results. The proposed method yields high bone segmentation accuracies (Dice coefficients of 98.18% and 97.85%, respectively) and low landmark prediction errors (1.55 mm and 1.65 mm) on both datasets. The automated measurements agree well with the radiologists’ manual measurements (Pearson correlation coefficients between 0.47 and 0.99 and intraclass correlation coefficients between 0.46 and 0.98). This method provides more accurate measurements than the conventional 2D x-ray-based measurement method, reducing the error of acetabular cup size from over 2 mm to less than 1 mm. Moreover, our morphometry measurement method is robust against the error of the previous bone segmentation step. As we tested different deep learning methods for the prerequisite bone segmentation, our method produced consistent final measurement results, with only a 0.37 mm maximum inter-method difference in the cup size. Significance. This study proposes a deep learning approach with improved robustness and accuracy for pelvis arthroplasty planning.

U2-MNet: An Improved Neural Network for Breast Tumor Segmentation

In recent years, UNet and its variants have achieved excellent performance. However, since the convolution kernel in UNet focuses solely on local pixels, these models struggle to model long-range dependencies. This issue has been addressed by recently proposed segmentation models based on transformer architecture. The internal self-attention mechanisms included in these systems capture global contextual information to improve segmentation effects. However, the results are often not ideal without pre-training on a large-scale dataset. Therefore, we designed a model (U2-MNet) to overcome the limitations of convolutional kernels, enabling the model to achieve high-accuracy segmentation without pre-training. This approach adopts a multi-layer framework, which can effectively describe multi-level channel information, employing a window-based channel MLP (WCM) block. It utilizes a sliding window and an MLP to capture information about details in local features. In addition, multi-level channel cross mixing (MCCM) block is included in each skip connection to reduce noise after aggregation of low-level and high-level features. The proposed model was trained from scratch and tested on Breast UltraSound Images (BUSI) dataset. Our model achieved 82.76%, 73.17%, and 86.24% on the Dice, IOU, and Precision indicators.

A novel methodology for water-sensitive papers analysis focusing on the segmentation of overlapping droplets to better characterize deposition pattern

Water-sensitive papers (WSPs), as an artificial target, have been generally used to evaluate spray quality in pesticide applications. Though many tools were available to analyze WSPs, they were not effective in processing the presented overlapping droplets, which prevented accurate quantification of spray quality. Here, a novel methodology was proposed to analyze WSPs, focusing on the segmentation of overlapping droplets based on concave point detection and ellipse fitting. Up to 553 WSPs obtained from field trials in apple trees were used to validate this methodology. A high overall segmentation accuracy of 77.8% was achieved for the WSPs with coverage below 25%, which allowed to precisely characterize the corresponding deposition pattern. A universal linear relationship was observed between the droplet density and coverage, independent of the sprayer and canopy characteristics. In addition, the droplet size distributions for all spray applications showed a similar trend. As the proposed methodology was not effective to segment the complex overlapping spots on the WSP with high coverage (>25%), simulated WSPs were generated to estimate the deposition pattern of high coverage based on the droplet size distribution deposited on the WSPs with low coverage. A linear relationship with a much higher slope was observed between droplet density and coverage for the high deposition, which enabled an approximate estimation of the corresponding droplet density. In general, the proposed approaches allowed to obtain more accurate spray quality indicators from the WSP collectors and could be widely used for spray application evaluation to promote precision spraying.

Revolutionizing law enforcement: The role of artificial intelligence in license plate recognition

This paper aims to offer a comprehensive review of the current state of the art in artificial intelligence (AI) as applied to license plate recognition. With the rapidly evolving nature of AI technology, deep learning approaches have gained popularity in license plate recognition, as exemplified by the success of AlphaGo. The diversity of AI in license plate recognition is notable, with numerous studies proposing systems that have achieved high accuracy in segmentation and recognition. The process of reading license plates is complex and involves several stages, including image capture, pre-processing, license plate identification, character segmentation, and recognition. Law enforcement widely uses automatic license plate recognition (ALPR) technology for detecting and preventing criminal activities, tracking stolen vehicles, and identifying suspects. Additionally, ALPR technology can monitor travel time on significant roadways, which can provide the Department of Transportation with useful data for efficient traffic management. Overall, this paper highlights the importance of AI in license plate recognition and its potential to revolutionize the field.

Estimation of right lobe graft weight for living donor liver transplantation using deep learning-based fully automatic computed tomographic volumetry

This study aimed at developing a fully automatic technique for right lobe graft weight estimation using deep learning algorithms. The proposed method consists of segmentation of the full liver region from computed tomography (CT) images, classification of the entire liver region into the right and left lobes, and estimation of the right lobe graft weight from the CT-measured right lobe graft volume using a volume-to-weight conversion formula. The first two steps were performed with a transformer-based deep learning model. To train and evaluate the model, a total of 248 CT datasets (188 for training, 40 for validation, and 20 for testing and clinical evaluation) were used. The Dice similarity coefficient (DSC), mean surface distance (MSD), and the 95th percentile Hausdorff distance (HD95) were used for evaluating the segmentation accuracy of the full liver region and the right liver lobe. The correlation coefficient (CC), percentage error (PE), and percentage absolute error (PAE) were used for the clinical evaluation of the estimated right lobe graft weight. The proposed method achieved high accuracy in segmentation for DSC, MSD, and HD95 (95.9% ± 1.0%, 1.2 ± 0.4 mm, and 5.2 ± 1.9 mm for the entire liver region; 92.4% ± 2.7%, 2.0 ± 0.7 mm, and 8.8 ± 2.9 mm for the right lobe) and in clinical evaluation for CC, PE, and PAE (0.859, − 1.8% ± 9.6%, and 8.6% ± 4.7%). For the right lobe graft weight estimation, the present study underestimated the graft weight by − 1.8% on average. A mean difference of − 21.3 g (95% confidence interval: − 55.7 to 13.1, p = 0.211) between the estimated graft weight and the actual graft weight was achieved in this study. The proposed method is effective for clinical application.

Deep Learning-Based Semantic Segmentation of Urban Areas Using Heterogeneous Unmanned Aerial Vehicle Datasets

Deep learning techniques have recently shown remarkable efficacy in the semantic segmentation of natural and remote sensing (RS) images. However, these techniques heavily rely on the size of the training data, and obtaining large RS imagery datasets is difficult (compared to RGB images), primarily due to environmental factors such as atmospheric conditions and relief displacement. Unmanned aerial vehicle (UAV) imagery presents unique challenges, such as variations in object appearance due to UAV flight altitude and shadows in urban areas. This study analyzed the combined segmentation network (CSN) designed to train heterogeneous UAV datasets effectively for their segmentation performance across different data types. Results confirmed that CSN yielded high segmentation accuracy on specific classes and can be used on diverse data sources for UAV image segmentation. The main contributions of this study include analyzing the impact of CSN on segmentation accuracy, experimenting with structures with shared encoding layers to enhance segmentation accuracy, and investigating the influence of data types on segmentation accuracy.

Reciprocal cross-modal guidance for liver lesion segmentation from multiple phases under incomplete overlap

Objective:Liver lesion segmentation is beneficial for computer-aided disease diagnosis, which can be further enhanced by incorporating multi-phase images. Nevertheless, existing segmentation models still need improvement as they build on the unsolid assumption that lesions are well-aligned between different phases. Our goal was to develop a robust model that achieves high accuracy in lesion segmentation, even in cases where incomplete overlap occurs across multiple phases. Methods:To overcome the strong assumption of precise alignment between phases, we introduce a novel deep neural network featuring multi-scale feature fusion and cross-phase feature guidance. The former enables the retrieval of effective information from each phase that is helpful in lesion segmentation, while the latter ensures exchanging of guidance maps for interphase segmentation. Results:Comprehensive experiments were carried out on an in-house multi-phase dataset having liver lesions, as well as a multi-modal dataset from the CHAOS challenge. Results show that our model surpasses existing state-of-the-art methods in segmenting liver lesions and abdominal organs from multi-phase images, even if the registration between phases is inaccurate. Conclusion:A competent solution was proposed for accurate liver lesion segmentation from multi-phase images, considering the imperfect registration between phases. The model was proved to be effective in extracting pertinent information from each phase and sharing useful information between phases. Significance:The accommodation of incomplete overlap in multi-phase images provides a new paradigm of accurate and robust lesion segmentation, which is particularly helpful in real-world scenarios where registration accuracy may be compromised.

MHorUNet: High-order spatial interaction UNet for skin lesion segmentation

In recent years, dermoscopy, as a noninvasive means of detection, has been increasingly used in the auxiliary diagnosis of skin disease, especially for skin cancer, such as malignant melanoma. And the automatic segment is a key step to improve accuracy of diagnosis. Generally, UNet models and its alternative schemes have occupied the vast majority of segmentation tasks in medical image processing. However, many of the current models are not perfect, the ordinary convolution in the UNet model cannot exhibit spatial dependence and remote interaction, while the use of Transformers as a convolution alternative is gradually becoming mainstream, but there are problems such as large data volume requirements as well as high computational effort in dealing with medical clinical problems. Therefore, we propose a HorUNet model with higher-order spatial interaction based on recursive gate convolution, and add a multi-stage dimensional fusion mechanism to the skip connection part to form the MHorUNet model architecture. The higher-order interaction mechanism with recursive gate convolution not only has the key factors for the success of Transformers, but also retains the excellent characteristics of convolution itself. We first performed comparative experiments by performing in two typical public skin lesion datasets (ISIC2017 and ISIC2018) and then used the PH2 dataset and our own dataset as external validation. The experimental results show that our method performs best in several metrics. This confirms that our model has a better generalization capability in terms of medically accurate segmentation results with high segmentation accuracy. The code can be obtained from https://github.com/wurenkai/MHorUNet.

Unsupervised Flame Segmentation Method Based on GK-RGB in Complex Background

Fires are disastrous events with significant negative impacts on both people and the environment. Thus, timely and accurate fire detection and firefighting operations are crucial for social development and ecological protection. In order to segment the flame accurately, this paper proposes the GK-RGB unsupervised flame segmentation method. In this method, RGB segmentation is used as the central algorithm to extract flame features. Additionally, a Gaussian filtering method is applied to remove noise interference from the image. Moreover, K-means mean clustering is employed to address incomplete flame segmentation caused by flame colours falling outside the fixed threshold. The experimental results show that the proposed method achieves excellent results on four flame images with different backgrounds at different time periods: Accuracy: 97.71%, IOU: 81.34%, and F1-score: 89.61%. Compared with other methods, GK-RGB has higher segmentation accuracy and is more suitable for the detection of fire. Therefore, the method proposed in this paper helps the application of firefighting and provides a new reference value for the detection and identification of fires.

Hybrid attention mechanism of feature fusion for medical image segmentation

AbstractTraditional convolution neural networks (CNN) have achieved good performance in multi‐organ segmentation of medical images. Due to the lack of ability to model long‐range dependencies and correlations between image pixels, CNN usually ignores the information of channel dimension. To further improve the performance of multi‐organ segmentation, a hybrid attention mechanism model is proposed. First, a CNN was used to extract multi‐scale feature maps and fed into the Channel Attention Enhancement Module (CAEM) to selectively pay attention to target organs in medical images, and the Transformer encoded tokenized image patches from CNN feature maps as the input sequence to model long‐range dependencies. Second, the decoder upsampled the output from Transformer and fused with the CAEM features in multi‐scale through skip connections. Finally, we introduced a Refinement Module (RM) after the decoder to improve feature correlations of the same organ and the feature discriminability between different organs. The model outperformed on dice coefficient (%) and hd95 on both the synapse multi‐organ segmentation and cardiac diagnosis challenge datasets. The hybrid attention mechanisms exhibited high efficiency and high segmentation accuracy in medical images.

Efficient multi-objective evolutionary neural architecture search for U-Nets with diamond atrous convolution and Transformer for medical image segmentation

Deep encoder–decoder neural networks like U-Nets have made significant contributions to the development of computer vision applications such as image segmentation. Neural architecture search (NAS) has the potential to further automatically adjust the architectures of U-Nets for various medical image segmentation tasks. Most of the NAS techniques focus on optimizing segmentation accuracies of network architectures. In real-world medical image segmentation scenarios, two main challenges are poor medical image quality and diverse deployment devices with different computing capabilities. A large architecture designed only for the high segmentation accuracy is difficult to run on various deployment devices. To address these challenges, this paper proposes a multi-objective evolutionary neural architecture search method (CTU-NAS) for U-Nets with diamond atrous convolution and Transformer for medical image segmentation. A hybrid U-Net architecture (CTU-Net) with diamond atrous convolution and Transformer modules is designed as the supernet of CTU-NAS. Then a channel search strategy based on sorting and selection is applied to speed up the search for subnets by precisely selecting and training the most important channels more times. In addition, CTU-NAS employs a dual acceleration mechanism based on weight sharing and surrogate model to lower the cost of evaluations of subnets. CTU-NAS applies a multi-objective evolutionary algorithm to balance between the segmentation accuracy and the number of parameters. Experimental results on two medical segmentation datasets show that CTU-NAS is capable of quickly generating a group of excellent network architectures with different sizes and their performances also outperform or come close to those of the manually designed networks.

Based on cross-scale fusion attention mechanism network for semantic segmentation for street scenes.

Semantic segmentation, which is a fundamental task in computer vision. Every pixel will have a specific semantic class assigned to it through semantic segmentation methods. Embedded systems and mobile devices are difficult to deploy high-accuracy segmentation algorithms. Despite the rapid development of semantic segmentation, the balance between speed and accuracy must be improved. As a solution to the above problems, we created a cross-scale fusion attention mechanism network called CFANet, which fuses feature maps from different scales. We first design a novel efficient residual module (ERM), which applies both dilation convolution and factorized convolution. Our CFANet is mainly constructed from ERM. Subsequently, we designed a new multi-branch channel attention mechanism (MCAM) to refine the feature maps at different levels. Experiment results show that CFANet achieved 70.6% mean intersection over union (mIoU) and 67.7% mIoU on Cityscapes and CamVid datasets, respectively, with inference speeds of 118 FPS and 105 FPS on NVIDIA RTX2080Ti GPU cards with 0.84M parameters.

Target Soybean Leaf Segmentation Model Based on Leaf Localization and Guided Segmentation

The phenotypic characteristics of soybean leaves are of great significance for studying the growth status, physiological traits, and response to the environment of soybeans. The segmentation model for soybean leaves plays a crucial role in morphological analysis. However, current baseline segmentation models are unable to accurately segment leaves in soybean leaf images due to issues like leaf overlap. In this paper, we propose a target leaf segmentation model based on leaf localization and guided segmentation. The segmentation model adopts a two-stage segmentation framework. The first stage involves leaf detection and target leaf localization. Based on the idea that a target leaf is close to the center of the image and has a relatively large area, we propose a target leaf localization algorithm. We also design an experimental scheme to provide optimal localization parameters to ensure precise target leaf localization. The second stage utilizes the target leaf localization information obtained from the first stage to guide the segmentation of the target leaf. To reduce the dependency of the segmentation results on the localization information, we propose a solution called guidance offset strategy to improve segmentation accuracy. We design multiple guided model experiments and select the one with the highest segmentation accuracy. Experimental results demonstrate that the proposed model exhibits strong segmentation capabilities, with the highest average precision (AP) and average recall (AR) reaching 0.976 and 0.981, respectively. We also compare our segmentation results with current baseline segmentation models, and multiple quantitative indicators and qualitative analysis indicate that our segmentation results are better.

Augmenting progress monitoring in soil-foundation construction utilizing SOLOv2-based instance segmentation and visual BIM representation

Efficient management of construction progress requires regular progress tracking, but monitoring progress on a daily basis can be both time-consuming and labor-intensive since the meticulous manual data processing is involved. Soil-foundation construction entails multiple uncertain underground safety risks, the elimination of which requires significant time and effort, thereby increasing the likelihood of schedule overruns. Automation and augmentation of progress monitoring are necessary for soil-foundation construction. There are two primary research challenges in the context of soil-foundation construction: the precise differentiation of objects with similar apparent characteristics, and the accurate detection of partially obscured objects. The existing approaches primarily address such issues by image post-processing rather than augmenting image feature inference. To enhance the efficiency of progress monitoring and facilitate unmanned inspection, an integrated computer vision-based framework for the tracking of soil-foundation construction progress was proposed. An improved SOLOv2 was utilized to qualitatively and quantitatively recognize the construction progress. Subsequently, the actual and differential progress could be integrated into streamlined BIM based on a self-adaptive grid-based mapping method. The approach was applied to a campus building project in China, resulting in a high segmentation accuracy (mAP = 90.9%). The improved SOLOv2 was found to surpass other state-of-the-art segmentation algorithms. Further, the impact of grid size on the mapping accuracy of construction progress was explored. The present study promotes automated schedule tracking and provides a feasible approach for developing a digital twin of soil-foundation construction.

Instance Segmentation of Lotus Pods and Stalks in Unstructured Planting Environment Based on Improved YOLOv5

Accurate segmentation of lotus pods and stalks with pose variability is a prerequisite for realizing the robotic harvesting of lotus pods. However, the complex growth environment of lotus pods causes great difficulties in conducting the above task. In this study, an instance segmentation model, LPSS-YOLOv5, for lotus pods and stalks based on the latest YOLOv5 v7.0 instance segmentation model was proposed. The CBAM attention mechanism was integrated into the network to improve the model’s feature extraction ability. The scale distribution of the multi-scale feature layer was adjusted, a 160 × 160 small-scale detection layer was added, and the original 20 × 20 large-scale detection layer was removed, which improved the model’s segmentation accuracy for small-scale lotus stalks and reduced the model size. On the medium-large scale test set, LPSS-YOLOv5 achieved a mask mAP0.5 of 99.3% for all classes. On the small-scale test set, the mAP0.5 for all classes and AP0.5 for stalks were 88.8% and 83.3%, which were 2.6% and 5.0% higher than the baseline, respectively. Compared with the mainstream Mask R-CNN and YOLACT models, LPSS-YOLOv5 showed a much higher segmentation accuracy, speed, and smaller size. The 2D and 3D localization tests verified that LPSS-YOLOv5 could effectively support the picking point localization and the pod–stalk affiliation confirmation.

GMDNet: An Irregular Pavement Crack Segmentation Method Based on Multi-Scale Convolutional Attention Aggregation

Pavement cracks are the primary type of distress that cause road damage, and deep-learning-based pavement crack segmentation is a critical technology for current pavement maintenance and management. To address the issues of segmentation discontinuity and poor performance in the segmentation of irregular cracks faced by current semantic segmentation models, this paper proposes an irregular pavement crack segmentation method based on multi-scale convolutional attention aggregation. In this approach, GhostNet is first introduced as the model backbone network for reducing parameter count, with dynamic convolution enhancing GhostNet’s feature extraction capability. Next, a multi-scale convolutional attention aggregation module is proposed to cause the model to focus more on crack features and thus improve the segmentation effect on irregular cracks. Finally, a progressive up-sampling structure is used to enrich the feature information by gradually fusing feature maps of different depths to enhance the continuity of segmentation results. The experimental results on the HGCrack dataset show that GMDNet has a lighter model structure and higher segmentation accuracy than the mainstream semantic segmentation algorithms, achieving 75.16% of MIoU and 84.43% of F1 score, with only 7.67 M parameters. Therefore, the GMDNet proposed in this paper can accurately and efficiently segment irregular cracks on pavements that are more suitable for pavement crack segmentation scenarios in practical applications.

A Fully Automatic Method to Segment Choroid Plexuses in Multiple Sclerosis Using Conventional MRI Sequences.

Choroid plexus (CP) volume has been recently proposed as a proxy for brain neuroinflammation in multiple sclerosis (MS). To develop and validate a fast automatic method to segment CP using routinely acquired brain T1-weighted and FLAIR MRI. Retrospective. Fifty-five MS patients (33 relapsing-remitting, 22 progressive; mean age = 46.8 ± 10.2 years; 31 women) and 60 healthy controls (HC; mean age = 36.1 ± 12.6 years, 33 women). 3D T2-weighted FLAIR and 3D T1-weighted gradient echo sequences at 3.0 T. Brain tissues were segmented on T1-weighted sequences and a Gaussian Mixture Model (GMM) was fitted to FLAIR image intensities obtained from the ventricle masks of the SIENAX. A second GMM was then applied on the thresholded and filtered ventricle mask. CP volumes were automatically determined and compared with those from manual segmentation by two raters (with 3 and 10 years' experience; reference standard). CP volumes from previously published automatic segmentation methods (freely available Freesurfer [FS] and FS-GMM) were also compared with reference standard. Expanded Disability Status Scale (EDSS) score was assessed within 3 days of MRI. Computational time was assessed for each automatic technique and manual segmentation. Comparisons of CP volumes with reference standard were evaluated with Bland Altman analysis. Dice similarity coefficients (DSC) were computed to assess automatic CP segmentations. Volume differences between MS and HC for each method were assessed with t-tests and correlations of CP volumes with EDSS were assessed with Pearson's correlation coefficients (R). A P value <0.05 was considered statistically significant. Compared to manual segmentation, the proposed method had the highest segmentation accuracy (mean DSC = 0.65 ± 0.06) compared to FS (mean DSC = 0.37 ± 0.08) and FS-GMM (0.58 ± 0.06). The percentage CP volume differences relative to manual segmentation were -0.1% ± 0.23, 4.6% ± 2.5, and -0.48% ± 2 for the proposed method, FS, and FS-GMM, respectively. The Pearson's correlations between automatically obtained CP volumes and the manually obtained volumes were 0.70, 0.54, and 0.56 for the proposed method, FS, and FS-GMM, respectively. A significant correlation between CP volume and EDSS was found for the proposed automatic pipeline (R = 0.2), for FS-GMM (R = 0.3) and for manual segmentation (R = 0.4). Computational time for the proposed method (32 ± 2 minutes) was similar to the manual segmentation (20 ± 5 minutes) but <25% of the FS (120 ± 15 minutes) and FS-GMM (125 ± 15 minutes) methods. This study developed an accurate and easily implementable method for automatic CP segmentation in MS using T1-weighted and FLAIR MRI. 1 TECHNICAL EFFICACY: Stage 4.

Using Artificial Intelligence to Diagnose Osteoporotic Vertebral Fractures on Plain Radiographs.

Osteoporotic vertebral fracture (OVF) is a risk factor for morbidity and mortality in elderly population, and accurate diagnosis is important for improving treatment outcomes. OVF diagnosis suffers from high misdiagnosis and underdiagnosis rates, as well as high workload. Deep learning methods applied to plain radiographs, a simple, fast, and inexpensive examination, might solve this problem. We developed and validated a deep-learning-based vertebral fracture diagnostic system using area loss ratio, which assisted a multitasking network to perform skeletal position detection and segmentation and identify and grade vertebral fractures. As the training set and internal validation set, we used 11,397 plain radiographs from six community centers in Shanghai. For the external validation set, 1276 participants were recruited from the outpatient clinic of the Shanghai Sixth People's Hospital (1276 plain radiographs). Radiologists performed all X-ray images and used the Genant semiquantitative tool for fracture diagnosis and grading as the ground truth data. Accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were used to evaluate diagnostic performance. The AI_OVF_SH system demonstrated high accuracy and computational speed in skeletal position detection and segmentation. In the internal validation set, the accuracy, sensitivity, and specificity with the AI_OVF_SH model were 97.41%, 84.08%, and 97.25%, respectively, for all fractures. The sensitivity and specificity for moderate fractures were 88.55% and 99.74%, respectively, and for severe fractures, they were 92.30% and 99.92%. In the external validation set, the accuracy, sensitivity, and specificity for all fractures were 96.85%, 83.35%, and 94.70%, respectively. For moderate fractures, the sensitivity and specificity were 85.61% and 99.85%, respectively, and 93.46% and 99.92% for severe fractures. Therefore, the AI_OVF_SH system is an efficient tool to assist radiologists and clinicians to improve the diagnosing of vertebral fractures. © 2023 The Authors. Journal of Bone and Mineral Research published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research (ASBMR).