Average Dice Coefficient Research Articles

Automated Segmentation and Diagnostic Measurement for the Evaluation of Cervical Spine Injuries Using X-Rays.

Accurate assessment of cervical spine X-ray images through diagnostic metrics plays a crucial role in determining appropriate treatment strategies for cervical injuries and evaluating surgical outcomes. Such assessment can be facilitated through the use of automatic methods such as machine learning and computer vision algorithms. A total of 852 cervical X-rays obtained from Gachon Medical Center were used for multiclass segmentation of the craniofacial bones (hard palate, basion, opisthion) and cervical spine (C1-C7), incorporating architectures such as EfficientNetB4, DenseNet201, and InceptionResNetV2. Diagnostic metrics automatically measured using computer vision algorithms were compared with manually measured metrics through Pearson's correlation coefficient and paired t-tests. The three models demonstrated high average dice coefficient values for the cervical spine (C1, 0.93; C2, 0.96; C3, 0.96; C4, 0.96; C5, 0.96; C6, 0.96; C7, 0.95) and lower values for the craniofacial bones (hard palate, 0.69; basion, 0.81; opisthion, 0.71). Comparison of manually measured metrics and automatically measured metrics showed high Pearson's correlation coefficients in McGregor's line (r = 0.89), space available cord (r = 0.94), cervical sagittal vertical axis (r = 0.99), cervical lordosis (r = 0.88), lower correlations in basion-dens interval (r = 0.65), basion-axial interval (r = 0.72), and Powers ratio (r = 0.62). No metric showed adjusted significant differences at P < 0.05 between manual and automatic metric measuring methods. These findings demonstrate the potential of multiclass segmentation in automating the measurement of diagnostic metrics for cervical spine injuries and showcase the clinical potential for diagnosing cervical spine injuries and evaluating cervical surgical outcomes.

MOTC: Abdominal Multi-objective Segmentation Model with Parallel Fusion of Global and Local Information.

Convolutional Neural Networks have been widely applied in medical image segmentation. However, the existence of local inductive bias in convolutional operations restricts the modeling of long-term dependencies. The introduction of Transformer enables the modeling of long-term dependencies and partially eliminates the local inductive bias in convolutional operations, thereby improving the accuracy of tasks such as segmentation and classification. Researchers have proposed various hybrid structures combining Transformer and Convolutional Neural Networks. One strategy is to stack Transformer blocks and convolutional blocks to concentrate on eliminating the accumulated local bias of convolutional operations. Another strategy is to nest convolutional blocks and Transformer blocks to eliminate bias within each nested block. However, due to the granularity of bias elimination operations, these two strategies cannot fully exploit the potential of Transformer. In this paper, a parallel hybrid model is proposed for segmentation, which includes a Transformer branch and a Convolutional Neural Network branch in encoder. After parallel feature extraction, inter-layer information fusion and exchange of complementary information are performed between the two branches, simultaneously extracting local and global features while eliminating the local bias generated by convolutional operations within the current layer. A pure convolutional operation is used in decoder to obtain final segmentation results. To validate the impact of the granularity of bias elimination operations on the effectiveness of local bias elimination, the experiments in this paper were conducted on Flare21 dataset and Amos22 dataset. The average Dice coefficient reached 92.65% on Flare21 dataset, and 91.61% on Amos22 dataset, surpassing comparative methods. The experimental results demonstrate that smaller granularity of bias elimination operations leads to better performance.

Multi-phase CT synthesis-assisted segmentation of abdominal organs

To propose a method for abdominal multi-organ segmentation assisted by multi-phase CT synthesis. Multi-phase CT synthesis for synthesizing high-quality CT images was used to increase the information details for image segmentation. A transformer block was introduced to help to capture long-range semantic information in cooperation with perceptual loss to minimize the differences between the real image and synthesized image. The model was trained using multi-phase CT dataset of 526 total cases from Nanfang Hospital. The mean maximum absolute error (MAE) of the synthesized non-contrast CT, venous phase contrast- enhanced CT (CECT), and delay phase CECT images from arterial phase CECT was 19.192±3.381, 20.140±2.676 and 22.538±2.874, respectively, which were better than those of images synthesized using other methods. Validation of the multi-phase CT synthesis-assisted abdominal multi-organ segmentation method showed an average dice coefficient of 0.847 for the internal validation set and 0.823 for the external validation set. The propose method is capable of synthesizing high-quality multi-phase CT images to effectively reduce the errors in registration between different phase CT images and improve the performance for segmentation of 13 abdominal organs.

Using Artificial Intelligence for Rheumatic Heart Disease Detection by Echocardiography: Focus on Mitral Regurgitation.

Identification of children with latent rheumatic heart disease (RHD) by echocardiography, before onset of symptoms, provides an opportunity to initiate secondary prophylaxis and prevent disease progression. There have been limited artificial intelligence studies published assessing the potential of machine learning to detect and analyze mitral regurgitation or to detect the presence of RHD on standard portable echocardiograms. We used 511 echocardiograms in children, focusing on color Doppler images of the mitral valve. Echocardiograms were independently reviewed by an expert adjudication panel. Among 511 cases, 229 were normal, and 282 had RHD. Our automated method included harmonization of echocardiograms to localize the left atrium during systole using convolutional neural networks and RHD detection using mitral regurgitation jet analysis and deep learning models with an attention mechanism. We identified the correct view with an average accuracy of 0.99 and the correct systolic frame with an average accuracy of 0.94 (apical) and 0.93 (parasternal long axis). It localized the left atrium with an average Dice coefficient of 0.88 (apical) and 0.9 (parasternal long axis). Maximum mitral regurgitation jet measurements were similar to expert manual measurements (P value=0.83) and a 9-feature mitral regurgitation analysis showed an area under the receiver operating characteristics curve of 0.93, precision of 0.83, recall of 0.92, and F1 score of 0.87. Our deep learning model showed an area under the receiver operating characteristics curve of 0.84, precision of 0.78, recall of 0.98, and F1 score of 0.87. Artificial intelligence has the potential to detect RHD as accurately as expert cardiologists and to improve with more data. These innovative approaches hold promise to scale echocardiography screening for RHD.

Study on high-precision three-dimensional reconstruction of pulmonary lesions and surrounding blood vessels based on CT images.

The adoption of computerized tomography (CT) technology has significantly elevated the role of pulmonary CT imaging in diagnosing and treating pulmonary diseases. However, challenges persist due to the complex relationship between lesions within pulmonary tissue and the surrounding blood vessels. These challenges involve achieving precise three-dimensional reconstruction while maintaining accurate relative positioning of these elements. To effectively address this issue, this study employs a semi-automatic precise labeling process for the target region. This procedure ensures a high level of consistency in the relative positions of lesions and the surrounding blood vessels. Additionally, a morphological gradient interpolation algorithm, combined with Gaussian filtering, is applied to facilitate high-precision three-dimensional reconstruction of both lesions and blood vessels. Furthermore, this technique enables post-reconstruction slicing at any layer, facilitating intuitive exploration of the correlation between blood vessels and lesion layers. Moreover, the study utilizes physiological knowledge to simulate real-world blood vessel intersections, determining the range of blood vessel branch angles and achieving seamless continuity at internal blood vessel branch points. The experimental results achieved a satisfactory reconstruction with an average Hausdorff distance of 1.5 mm and an average Dice coefficient of 92%, obtained by comparing the reconstructed shape with the original shape,the approach also achieves a high level of accuracy in three-dimensional reconstruction and visualization. In conclusion, this study is a valuable source of technical support for the diagnosis and treatment of pulmonary diseases and holds promising potential for widespread adoption in clinical practice.

Multitask connected U-Net: automatic lung cancer segmentation from CT images using PET knowledge guidance.

Lung cancer is a predominant cause of cancer-related mortality worldwide, necessitating precise tumor segmentation of medical images for accurate diagnosis and treatment. However, the intrinsic complexity and variability of tumor morphology pose substantial challenges to segmentation tasks. To address this issue, we propose a multitask connected U-Net model with a teacher-student framework to enhance the effectiveness of lung tumor segmentation. The proposed model and framework integrate PET knowledge into the segmentation process, leveraging complementary information from both CT and PET modalities to improve segmentation performance. Additionally, we implemented a tumor area detection method to enhance tumor segmentation performance. In extensive experiments on four datasets, the average Dice coefficient of 0.56, obtained using our model, surpassed those of existing methods such as Segformer (0.51), Transformer (0.50), and UctransNet (0.43). These findings validate the efficacy of the proposed method in lung tumor segmentation tasks.

VisioRenalNet: Spatial Vision Transformer UNet for enhanced T2-Weighted Kidney MRI Segmentation

Chronic Kidney Disease (CKD) afflicts approximately 20% of the global population, underscoring the pressing need for precise detection and diagnosis. Accurate kidney segmentation in medical imaging is pivotal for capturing intricate details, facilitating effective evaluation and treatment planning. This work consists implementation of U-Net Variants along with enhanced variant for kidney segmentation in Magnetic Resonance Images (MRI). We systematically assess the performance of various U-Net variants, including U-Net, VGGUNet, DenseUNet, ResUNet, U-Net++, and AttentionUNet, utilizing a publicly available dataset of T2-weighted abdomen MRI scans from 100 subjects. Subsequently, we propose an enhanced U-Net framework that seamlessly integrates transformer blocks, spatial Attention Mechanism, and swish activation to enhance performance and robustness. Enhanced UNet provides improved feature learning, enhanced contextual information, selective focus and improved performance metrices for Kidney Segmentation. This significant advancement in medical image segmentation can contribute to enhanced diagnosis, more effective treatment planning, and improved disease management for individuals suffering from kidney diseases. Notably, our enhanced U-Net variant, after 50 epochs, achieved an average Dice coefficient of 89.1%, a Jaccard score of 80.4%, and an impressive accuracy of 96.2%, surpassing all baseline U-Net variants in all evaluation metrics, thereby affirming its superiority in kidney segmentation.

Blood clot and fibrin recognition method for serum images based on deep learning

BackgroundDetecting and identifying of clots and fibrins in serum is an important process in the analysis stage before laboratory analysis. Currently, visual examination is commonly employed in clinical laboratories for this purpose. However, this method is not only time-consuming but also highly subjective and may result in misjudgments. MethodA serum image blood clot and fibrin segmentation method based on improved UNeXt was proposed. The improved UNeXt segmentation network was used to train the self-built serum dataset, and the trained model was used for blood clot and fibrin segmentation in serum images. Whether the serum images contained blood clots and fibrins was identified according to the segmentation results. ResultsThe average Dice coefficient of the serum image segmentation network output was 0.8707, which realized more accurate segmentation of blood clots and fibrins in serum images. In 13,230 clinical serum samples, the sensitivity, specificity, and accuracy of blood clot and fibrins segmentation in serum images were 95.74%, 98.11% and 97.93%, respectively, which meet clinical test requirements. ConclusionsThe improved UNeXt segmentation network rapidly and accurately segmented and recognized blood clots and fibrins in serum images, which provided an accurate basis for the sampling height of the sampling needle in the automated biochemical and immunological assembly line.

Automated detection and segmentation of pleural effusion on ultrasound images using an Attention U-net.

Ultrasonic for detecting and evaluating pleural effusion is an essential part of the Extended Focused Assessment with Sonography in Trauma (E-FAST) in emergencies. Our study aimed to develop an Artificial Intelligence (AI) diagnostic model that automatically identifies and segments pleural effusion areas on ultrasonography. An Attention U-net and a U-net model were used to detect and segment pleural effusion on ultrasound images of 848 subjects through fully supervised learning. Sensitivity, specificity, precision, accuracy, F1 score, the receiver operating characteristic (ROC) curve, and the area under the curve (AUC) were used to assess the model's effectiveness in classifying the data. The dice coefficient was used to evaluate the segmentation performance of the model. In 10 random tests, the Attention U-net and U-net 's average sensitivity of 97% demonstrated that the pleural effusion was well detectable. The Attention U-net performed better at identifying negative images than the U-net, which had an average specificity of 91% compared to 86% for the U-net. Additionally, the Attention U-net was more accurate in predicting the pleural effusion region because its average dice coefficient was 0.86 as opposed to the U-net's average dice coefficient of 0.82. The Attention U-net showed excellent performance in detecting and segmenting pleural effusion on ultrasonic images, which is expected to enhance the operation and application of E-FAST in clinical work.

Prediction of preoperative microvascular invasion by dynamic radiomic analysis based on contrast-enhanced computed tomography.

Microvascular invasion (MVI) is a common complication of hepatocellular carcinoma (HCC) surgery, which is an important predictor of reduced surgical prognosis. This study aimed to develop a fully automated diagnostic model to predict pre-surgical MVI based on four-phase dynamic CT images. A total of 140 patients with HCC from two centers were retrospectively included (training set, n = 98; testing set, n = 42). All CT phases were aligned to the portal venous phase, and were then used to train a deep-learning model for liver tumor segmentation. Radiomics features were extracted from the tumor areas of original CT phases and pairwise subtraction images, as well as peritumoral features. Lastly, linear discriminant analysis (LDA) models were trained based on clinical features, radiomics features, and hybrid features, respectively. Models were evaluated by area under curve (AUC), accuracy, sensitivity, specificity, positive and negative predictive values (PPV and NPV). Overall, 86 and 54 patients with MVI- (age, 55.92 ± 9.62 years; 68 men) and MVI+ (age, 53.59 ± 11.47 years; 43 men) were included. Average dice coefficients of liver tumor segmentation were 0.89 and 0.82 in training and testing sets, respectively. The model based on radiomics (AUC = 0.865, 95% CI: 0.725-0.951) showed slightly better performance than that based on clinical features (AUC = 0.841, 95% CI: 0.696-0.936). The classification model based on hybrid features achieved better performance in both training (AUC = 0.955, 95% CI: 0.893-0.987) and testing sets (AUC = 0.913, 95% CI: 0.785-0.978), compared with models based on clinical and radiomics features (p-value < 0.05). Moreover, the hybrid model also provided the best accuracy (0.857), sensitivity (0.875), and NPV (0.917). The classification model based on multimodal intra- and peri-tumoral radiomics features can well predict HCC patients with MVI.

Tumor Segmentation in Colorectal Ultrasound Images Using an Ensemble Transfer Learning Model: Towards Intra-Operative Margin Assessment

Tumor boundary identification during colorectal cancer surgery can be challenging, and incomplete tumor removal occurs in approximately 10% of the patients operated for advanced rectal cancer. In this paper, a deep learning framework for automatic tumor segmentation in colorectal ultrasound images was developed, to provide real-time guidance on resection margins using intra-operative ultrasound. A colorectal ultrasound dataset was acquired consisting of 179 images from 74 patients, with ground truth tumor annotations based on histopathology results. To address data scarcity, transfer learning techniques were used to optimize models pre-trained on breast ultrasound data for colorectal ultrasound data. A new custom gradient-based loss function (GWDice) was developed, which emphasizes the clinically relevant top margin of the tumor while training the networks. Lastly, ensemble learning methods were applied to combine tumor segmentation predictions of multiple individual models and further improve the overall tumor segmentation performance. Transfer learning outperformed training from scratch, with an average Dice coefficient over all individual networks of 0.78 compared to 0.68. The new GWDice loss function clearly decreased the average tumor margin prediction error from 1.08 mm to 0.92 mm, without compromising the segmentation of the overall tumor contour. Ensemble learning further improved the Dice coefficient to 0.84 and the tumor margin prediction error to 0.67 mm. Using transfer and ensemble learning strategies, good tumor segmentation performance was achieved despite the relatively small dataset. The developed US segmentation model may contribute to more accurate colorectal tumor resections by providing real-time intra-operative feedback on tumor margins.

Cross-layer connected network with adaptive attention mechanism for 3D multi-organ and tumor segmentations from CT

3D medical image segmentation is one of the most important and challenging tasks for medical image analysis. To enhance the feature extraction ability of 3D convolutional neural networks (CNN) and improve the segmentation accuracy on small tumors, this paper presents a cross-layer connected network with an adaptive attention mechanism to achieve efficient segmentations for 3D multi-organ and tumor from CT. A cross layer connection is proposed to utilize features more effectively by reusing multiple layers of semantic information while recovering spatial information. The pre-activation Residual Block and Squeeze-and-Excitation (SE) block are introduced to enhance the feature extraction ability. Due to the serious class imbalance problem caused by small tumors, the training process of the 3D CNN model is unstable and small targets are difficult to segment. To relieve the training instability problem, Anti-Aliased Pooling (AAP) is introduced to replace the original maxpooling. Moreover, a region of interest (ROI) adaptive attention mechanism based on the auxiliary loss function is developed to improve the segmentation accuracy on small target tumors by combining the target detection method with semantic segmentation models. Extensive experiments show that the dice coefficient of 3D liver tumor segmentation on the LiTS dataset reaches 0.668 by using our designed network, and the experiments on the BRATS dataset for 3D brain tumor segmentation yield average dice coefficient, sensitivity, and Hausdorff distance of 0.706, 0.730, and 5.972 mm, respectively, which are obviously superior to some existing methods.

Research on Pectoral Muscle Segmentation Algorithm of CT Image Based on Deep Learning.

Chronic obstructive pulmonary disease (COPD) is a common respiratory disease, which seriously endangers human health and is also one of the important causes of death. The death rate of COPD in China is the highest in the world, and the problem of under-diagnosis of the disease is very serious. The gold standard for the diagnosis of COPD is lung function examination, and clinical studies have shown that CT and other imaging methods can be included in the auxiliary diagnosis of COPD. CT images can be used to assess pectoral muscle area, which is associated with COPD severity. Patients with lower pectoral muscle area often have more severe expiratory airflow obstruction and other problems. Therefore, the key of the research is to accurately segment the pectoral muscle in CT images. The medical image segmentation method based on deep learning can dig out more abundant internal information of data, so it has gradually become the preferred method in the aspect of medical image segmentation. In this paper, a pectoral muscle segmentation algorithm based on U-Net and its variant U-Net++ is proposed, which is of great significance for evaluating the severity of disease in patients with COPD. The network is composed of symmetrical encoders and decoders, which can effectively learn from very little labelled data by using appropriate data enhancement methods, and is therefore very suitable for medical image segmentation. The experimental results on the data set provided by Jiangsu Province Hospital show that the average Dice coefficient of the proposed algorithm is more than 94% and the average accuracy rate is 91%. The algorithm can accurately segment the pectoral muscle in CT images and has good segmentation performance.

BTSC-TNAS: A neural architecture search-based transformer for brain tumor segmentation and classification

Glioblastoma (GBM), isolated brain metastasis (SBM), and primary central nervous system lymphoma (PCNSL) possess a high level of similarity in histomorphology and clinical manifestations on multimodal MRI. Such similarities have led to challenges in the clinical diagnosis of these three malignant tumors. However, many existing models solely focus on either the task of segmentation or classification, which limits the application of computer-aided diagnosis in clinical diagnosis and treatment. To solve this problem, we propose a multi-task learning transformer with neural architecture search (NAS) for brain tumor segmentation and classification (BTSC-TNAS). In the segmentation stage, we use a nested transformer U-shape network (NTU-NAS) with NAS to directly predict brain tumor masks from multi-modal MRI images. In the tumor classification stage, we use the multiscale features obtained from the encoder of NTU-NAS as the input features of the classification network (MSC-NET), which are integrated and corrected by the classification feature correction enhancement (CFCE) block to improve the accuracy of classification. The proposed BTSC-TNAS achieves an average Dice coefficient of 80.86% and 87.12% for the segmentation of tumor region and the maximum abnormal region in clinical data respectively. The model achieves a classification accuracy of 0.941. The experiments performed on the BraTS 2019 dataset show that the proposed BTSC-TNAS has excellent generalizability and can provide support for some challenging tasks in the diagnosis and treatment of brain tumors.

Cyto‐Knet: An instance segmentation approach for multiple myeloma plasma cells using conditional kernels

AbstractMultiple myeloma disrupts normal blood cell production, requiring early detection due to the increased risk of bone metastases. Although various artificial intelligence (AI) methods have been developed to assist pathologists, they often lack comprehensive metrics to measure both detection and segmentation errors. This study presents a hybrid framework that combines deep learning and heuristic techniques to achieve accurate instance segmentation of individual plasma cells in Giemsa‐stained cytology images. Our proposed network, called cyto‐Knet, incorporates an innovative color‐balancing method as a preprocessing step to standardize the appearance of cytological images. Our network leverages a 4‐class segmentation strategy with conditional kernels to enhance segmentation performance and accuracy. Additionally, a marker‐based watershed algorithm is employed in the postprocessing step to address the challenge of merged objects. Extensive validation at both pixel and object‐based levels demonstrates superior performance compared with state‐of‐the‐art techniques. Our method achieves pixel‐based metrics (precision, recall, and F1‐score) of approximately 0.90. The object‐based evaluation reveals an average Dice coefficient of 0.9130 and an aggregated Jaccard index of 0.8237. Importantly, our solution is designed for integration into an end‐to‐end system for diagnosis support and can be easily extended to other applications.

An attention-guided network for bilateral ventricular segmentation in pediatric echocardiography

Accurate segmentation of pediatric echocardiograms is a challenging task, because significant heart-size changes with age and faster heart rate lead to more blurred boundaries on cardiac ultrasound images compared with adults. To address these problems, a dual decoder network model combining channel attention and scale attention is proposed in this paper. Firstly, an attention-guided decoder with deep supervision strategy is used to obtain attention maps for the ventricular regions. Then, the generated ventricular attention is fed back to multiple layers of the network through skip connections to adjust the feature weights generated by the encoder and highlight the left and right ventricular areas. Finally, a scale attention module and a channel attention module are utilized to enhance the edge features of the left and right ventricles. The experimental results demonstrate that the proposed method in this paper achieves an average Dice coefficient of 90.63% in acquired bilateral ventricular segmentation dataset, which is better than some conventional and state-of-the-art methods in the field of medical image segmentation. More importantly, the method has a more accurate effect in segmenting the edge of the ventricle. The results of this paper can provide a new solution for pediatric echocardiographic bilateral ventricular segmentation and subsequent auxiliary diagnosis of congenital heart disease.

Multiencoder‐based federated intelligent deep learning model for brain tumor segmentation

AbstractGlioma, a primary tumor derived from brain glial cells, is around 45% of all intracranial tumors. Magnetic resonance imaging's (MRI's) precise glioma segmentation is crucial for clinical purposes. This article presents a novel automatic brain tumor segmentation approach based on a multi‐encoder‐based federated intelligent deep learning framework. The suggested method uses a U‐shaped network design that multiplies the single contraction path into several paths to explore semantic information modalities deeply. The basic convolutional layer uses an Inception module and dilated convolutions to extract multi‐scale features from the images using artificial intelligent. To emphasize segmentation‐related information while ignoring redundant channel dimension information and improving the accuracy of network segmentation, lightweight channel attention efficient channel attention (ECA) modules are inserted into the bottleneck layer and decoder. The collection of data for the 2018 Brain Tumor Segmentation Challenge (BraTS 2018) is used to test the effectiveness of the suggested structure, and the findings indicate that the growth core, for the entire tumor and the augmented tumor regions, respectively, the average Dice coefficients are 0.880, 0.784, and 0.757. These findings support the proposed algorithm's ability to accurately and successfully segregate multimodal MRI brain tumors.

Thermal damage map prediction during irreversible electroporation with U-Net

ABSTRACT Recent developments in cancer treatment with irreversible electroporation (IRE) have led to a renewed interest in developing a treatment planning system based on Deep-Learning methods. This paper will give an account of U-Net, as a Deep-Learning architecture usage for predicting thermal damage area during IRE. In this study, an irregular shape of the liver tumor with MIMICS and 3-Matic software was created from Magnetic Resonance Imaging (MRI) images. To create electric field distribution and thermal damage maps in IRE, COMSOL Multiphysics 5.3 finite element analysis was performed. It was decided to use the pair needle, single bipolar, and multi-tine electrodes with different geometrical parameters as electrodes. The U-Net was designed as a Deep-Learning network to train and predict the thermal damage area from electric field distribution in the IRE. The average DICE coefficient and accuracy of trained U-Net for predicting thermal damage area on test data sets were 0.96 and 0.98, respectively, for the dataset consisting of all electrode type electric field intensity images. This is the first time that U-Net has been used to predict thermal damage area. The results of this research support the idea that the U-Net can be used for predicting thermal damage areas during IRE as a treatment planning system.

Evaluation of Automated Coverage and Distance Mapping Selections to Improve Reliability and Clinical Utility of 3D Weightbearing CT Assessments

Category: Other; Hindfoot Introduction/Purpose: Progressive collapsing foot deformity (PCFD) is a complex three-dimensional (3D) deformity where adjacent structures may adopt subtle differences in positioning that result in increased contact or subluxation. Recent studies have highlighted the need for and utility of 3D analyses in PCFD using weightbearing CT (WBCT) and bone segmentation. Beyond the limitations of triplanar imaging, 3D distance and coverage mapping analyses have further highlighted key regional differences like sinus tarsi narrowing ahead of impingement and early middle facet uncoverage ahead of collapse. However, these analyses rely upon manual identification of subregions hindering the utility of 3D mapping clinically. The objective of this study was to compare an automated selection process with manual selections in the context of subtalar regional distance and coverage maps in PCFD. Methods: In this IRB-approved retrospective study, WBCT data of 20 consecutive patients with flexible PCFD and 10 controls were analyzed. Subregions of the peritalar surface (middle and posterior facets of the calcaneus and talus; sinus tarsi area) were manually selected by two experts on manually generated bone surfaces of all 30 feet. An automated algorithm for selecting coverage area was applied to identify the same regions on the semi-automatically generated bones (Figure). A 3D distance mapping (DM) technique was used to create coverage maps (CMs) across the entire peritalar surface where areas with distances less than 4mm were defined as covered. DM and CM percentages were compared using intra-class correlations and t-tests between PCFD and control groups. The Sørensen–Dice index, or Dice coefficient, was used for comparisons of selections on the semi-automated surfaces to evaluate reproducibility of expert selections. Results: The automated process produced identical selections resulting in perfect intra-method ICCs of 1.00 for all regions and Dice coefficients of 1.00. The average Dice coefficient for all manual selections was 0.903 (range: 0.865-0.935) indicating that observers were able to reliably select the same regions with 90% overlap. When assessing reliability of manual selections, intra- observer ICCs ranged from 0.41-0.92 while inter-observer ICCs ranged from 0.47-0.99 were found. Despite strong significant correlations, average coverage was significantly lower in the sinus tarsi region of the automated selections vs the manual selections (34.3±16.8% vs 23.1±12.7%, p&lt; 0.005). However, mean distances in each region were not significantly different in the middle facet or the sinus tarsi regions (p=0.323, p=0.095, respectively). Conclusion: Understanding of the complex 3D deformities that constitute PCFD requires sensitive and reproducible measures. Fully automated 3D assessments of coverage and bone relations can help improve understanding these deformities aiding in diagnosis, staging, and objective evaluation of treatment effects. Prior work with this method has specifically identified the middle facet and sinus tarsi regions as being of particular importance. Compared to manual selections, these regions were well identified by the automated process. This represents a major step toward viable use of fully automated 3D coverage and distance mapping when evaluating PCFD patients.

U-Net transfer learning backbones for lesions segmentation in breast ultrasound images

&lt;p&gt;Breast ultrasound images are highly valuable for the early detection of breast cancer. However, the drawback of these images is low-quality resolution and the presence of speckle noise, which affects their interpretability and makes them radiologists’ expertise-dependent. As medical images, breast ultrasound datasets are scarce and imbalanced, and annotating them is tedious and time-consuming. Transfer learning, as a deep learning technique, can be used to overcome the dataset deficiency in available images. This paper presents the implementation of transfer learning U-Net backbones for the automatic segmentation of breast ultrasound lesions and implements a threshold selection mechanism to deliver optimal generalized segmentation results of breast tumors. The work uses the public breast ultrasound images (BUSI) dataset and implements ten state-of-theart candidate models as U-Net backbones. We have trained these models with a five-fold cross-validation technique on 630 images with benign and malignant cases. Five out of ten models showed good results, and the best U-Net backbone was found to be DenseNet121. It achieved an average Dice coefficient of 0.7370 and a sensitivity of 0.7255. The model’s robustness was also evaluated against normal cases, and the model accurately detected 72 out of 113 images, which is higher than the four best models.&lt;/p&gt;