Lesion Boundary Research Articles

AUTOENCODER PERFORMANCE ANALYSIS OF SKIN LESION DETECTION

Melanoma is a rapidly pervasive and deathly type of skin cancer that is responsible for most deaths from this kind of disease. It can quickly prevail in other organs if not handled early. Fortunately, the symptoms of skin cancer become visible to the sick, which creates a chance to detect it at an early stage. Because people know so little about their specific symptoms and because of a shortage of expert doctors, automated skin cancer detection has become an important public health issue. Many computer-aided diagnostic methods have been suggested so far. Besides traditional techniques based on image processing, researchers have recently used deep learning successfully for many different purposes. Deep neural networks are widely used in segmentation, classification, detection, etc. In this paper, we check the applicability of deep learning approaches to the segmentation of skin lesions to detect lesion boundaries by evaluating five architectures: U-NET, RESU-NET, VGG16UNET, DENSENET121, and EfficientNet-B0 by presenting a comparative view of those approaches. The five architectures were trained on three different data sets: ISIC 2016, ISIC 2018, and PH2, each set consisting of skin lesion images and the ground truth for their segmentation, and then used pre-processed on three datasets. Quantitative evaluation metrics were used for evaluating the performance of the studied architectures. Among these five architectures, the DENSENET121 architecture showed the best precision rate in all training datasets. We obtained an above 95% precision score in the PH2 dataset. In addition, the pre-processing steps were beneficial for the results.

Multimodal Imaging of Torpedo Maculopathy With Fluorescence Adaptive Optics Imaging of Individual Retinal Pigmented Epithelial Cells.

Torpedo maculopathy (TM) is a rare congenital defect of the retinal pigment epithelium (RPE). The RPE is often evaluated clinically using fundus autofluorescence (AF), a technique that visualizes RPE structure at the tissue level from the intrinsic AF of RPE fluorophores. TM lesions typically emit little or no AF, but this macroscopic assessment is unable to resolve the RPE cells, leaving the organization of the RPE cell mosaic in TM unknown. We used fluorescence adaptive optics scanning laser ophthalmoscopy (AOSLO) to show here for the first time the microscopic cellular-level structural alterations to the RPE cell mosaic in TM that underlie the tissue-level changes seen in conventional clinical imaging. We evaluated two patients with TM using conventional clinical imaging techniques and adaptive optics (AO) infrared autofluorescence (IRAF) in AOSLO. Confocal AOSLO revealed relatively normal cones outside the TM lesion but altered cone appearance within it and along its margins in both patients. We quantified cone topography and RPE cell morphometry from the fovea to the margin of the lesion in case 1 and found cone density to be within the normal range across the locations imaged. However, RPE morphometric analysis revealed disrupted RPE cells outside the margin of the lesion; the mean RPE cell area was greater than two standard deviations above the normative range up to approximately 1.5 mm from the lesion margin. Similar morphometric changes were seen to individual RPE cells in case 2. Multi-modal imaging with AOSLO reveals that RPE cells are abnormal in TM well beyond the margins of the characteristic TM lesion boundary defined with conventional clinical imaging. Since the TM fovea appears to be fully formed, with normal cone packing, it is possible that the congenital RPE defect in TM occurs relatively late in retinal development. This work demonstrates how cellular level imaging of the RPE can provide new insight into RPE pathologies, particularly for rare conditions such as TM.

Lesion distribution pattern of parenchymal Neuro-Behçet's disease using probability mapping

Background and PurposeBehçet's disease (BD) is an inflammatory disorder with multisystemic involvement. The most disabling aspect of BD is Neuro-Behçet's disease (NBD). In NBD, parenchymal lesions tend to occur in the mesodiencephalic region and brainstem, as reported in large series of NBD. Our study aimed to generate probability maps of parenchymal lesions to compare patient subgroups with different clinical and laboratory features. MethodWe included 66 non-standardized acute relapse MRIs of 55 patients with parenchymal NBD (p-NBD). We used T2-weighted axial images to digitalize the lesions using the CAD software. Boundaries of lesions were determined as polygons and converted into high-definition raster datasets. Then, digitalized lesion maps were transferred into the ICBM-152 brain template to perform spatial analyses. Finally, we created subtraction maps to compare the patient subgroups. ResultsWe used a total of 66 MRIs of 55 patients to generate the probability maps. The most frequently affected parenchymal structures were the rostral pons, mesencephalon, and diencephalic region. Interestingly, the brainstem was more commonly affected in females than males (p<0.01). In the late-onset disease, lesions were localized in the corticospinal tracts and caudal brainstem (p<0.01). Progressive disease and severe disability at the end of the follow-up period were associated with corticospinal tract lesions during relapses (p<0.01). Patients with positive pathergy tests were more likely to present right hemisphere involvement (p<0.01). Additionally, cyclosporine-induced lesions tend to be in atypical locations such as hemispheric white matter. ConclusionsIn the published studies, lesions in NBD were localized according to coarse anatomical regions. Our study uses visual maps to offer accurate lesion localizations using non-standardized brain MRIs, allowing comparisons across different NBD subgroups. By using this technique, we investigated the relationship of the clinical and laboratory features with the lesion locations. We found that the late age of onset was associated with a poor prognosis. Additionally, corticospinal lesions may predict severe and progressive disease course, requiring aggressive treatment. Interestingly, females had more brainstem lesions and lesion lateralization might be influenced by the pathergy test status.

Effect of Combination of Traditional Chinese Medicine with Western Medicine on Endometrial Carcinoma and Its Influence on Ultrasound, MRI, Tumor Markers HE4 and CA125.

Objective To study the clinical efficacy of integrated traditional Chinese medicine (TCM) and Western medicine (WM) in treating endometrial cancer and the influence on ultrasound, magnetic resonance imaging (MRI), tumor markers, human epididymis protein 4 (HE4) and carbohydrate antigen 125 (CA125). Method A total of 152 cases of patients with endometrial carcinoma were randomly divided into two groups: the TCM + WM group and the WM group. The WM group was treated with megestrol acetate tablets, and the TCM + WM group was treated with Radix Astragali injection on the basis of the control group. The levels of inflammatory factors, HE4 and CA125 in serum, were detected using enzyme-linked immunosorbent assay (ELISA) or radioimmunoassay. The characteristics of ultrasound images and MRI images were observed and recorded. Toxicity, side effects, and the 3-year cumulative survival rate after treatment were assessed. Results After treatment, the levels of interleukin-4 (IL-4), tumor necrosis factor-alpha (TNF-α), and high-sensitivity C-reactive protein (hs-CRP) in both groups decreased, and the decrease in the TCM + WM group was more obvious than that in the WM group. There were statistically significant differences between the two groups in lesion shape, boundary, blood flow signal, lesion diameter, resistance index (RI), echo, intima thickness, and muscle layer infiltration from transvaginal ultrasound images after treatment. The diameter, echo, boundary, shape, composition, and enhancement degree of lesions between the two groups have a significant difference. Moreover, the levels of serum HE4 and CA125 in both groups decreased after treatment, and the decrease in the TCM + WM group was more obvious than that in the WM group. There were statistically significant differences between the two groups in the occurrence of myelosuppression, abnormal liver function, decreased platelet number, gastrointestinal reactions, leukopenia, and cardiotoxicity. After three years of follow-up, the cumulative survival rate of the TCM + WM group was 76.32%, and the cumulative survival rate of the WM group was 57.89%. Conclusion Radix Astragali injection combined with megestrol acetate tablets has obvious therapeutic effects against endometrial cancer. Through vaginal ultrasonography and MRI, it can significantly improve the size, shape, and blood flow signals of patients' lesions, reduce the level of serum inflammatory factors and tumor markers HE4 and CA125, reduce the incidence of toxic and side reactions, improve the patient's immunity, improve the patient's condition significantly, and prolong the survival time of patients.

Evaluation of Ischemic Penumbra in Stroke Patients Based on Deep Learning and Multimodal CT.

In order to investigate the value of multimodal CT for quantitative assessment of collateral circulation, ischemic semidark zone, core infarct volume in patients with acute ischemic stroke (AIS), and prognosis assessment in intravenous thrombolytic therapy, segmentation model which is based on the self-attention mechanism is prone to generate attention coefficient maps with incorrect regions of interest. Moreover, the stroke lesion is not clearly characterized, and lesion boundary is poorly differentiated from normal brain tissue, thus affecting the segmentation performance. To address this problem, a primary and secondary path attention compensation network structure is proposed, which is based on the improved global attention upsampling U-Net model. The main path network is responsible for performing accurate lesion segmentation and outputting segmentation results. Likewise, the auxiliary path network generates loose auxiliary attention compensation coefficients, which compensate for possible attention coefficient errors in the main path network. Two hybrid loss functions are proposed to realize the respective functions of main and auxiliary path networks. It is experimentally demonstrated that both the improved global attention upsampling U-Net and the proposed primary and secondary path attention compensation networks show significant improvement in segmentation performance. Moreover, patients with good collateral circulation have a small final infarct area volume and a good clinical prognosis after intravenous thrombolysis. Quantitative assessment of collateral circulation and ischemic semidark zone by multimodal CT can better predict the clinical prognosis of intravenous thrombolysis.

NIMG-26. IMPROVING THE GENERALIZABILITY OF DEEP LEARNING FOR T2-LESION SEGMENTATION OF GLIOMAS IN THE POST-TREATMENT SETTING

Abstract Although current advances for automated glioma lesion segmentation and volumetric measurements using deep learning have yielded high performance on newly-diagnosed patients, response assessment in neuro-oncology still relies on manually-drawn, cross-sectional areas of the tumor because these models do not generalize to patients in the post-treatment setting, where they are most needed in the clinic. Surgical resections, adjuvant treatment, or disease progression can alter the characteristics of these lesions on T2-weighted imaging, causing measures of segmentation accuracy, typically measured by Dice coefficients of overlap (DCs), to drop by ~15%. To improve the generalizability of T2-lesion segmentation to patients with glioma post-treatment, we evaluated the effects of: 1) training with different proportions of newly-diagnosed and treated gliomas, 2) applying transfer learning from pre- to post-treatment domains, and 3) incorporating a loss term that spatially weights the lesion boundaries with greater emphasis in training. Using 425 patients (208 newly-diagnosed, 217 post-Tx, with 25 treated patients withheld as a test set) and a top-performing model previously trained on newly-diagnosed gliomas, we found that DCs increased by 10% (to 0.84) then plateaued after including ~25% of post-treatment patients in training. Transfer learning (pre-training on newly-diagnosed and finetuning with post-treatment data) significantly improved Hausdorf distances (HDs), a measure more sensitive to changes at the lesion boundaries, by 17% after including 26% post-treatment images in training, while DCs remained similar. Although modifying our loss functions with boundary-weighted penalizations resulted in comparable DCs to using standard DC loss, HD measures were further reduced by 26%, suggesting that HDs may be a more sensitive metric to subtle changes in segmentation accuracy than DCs. Current work is evaluating their utility in providing accurate volumes for real-time response assessment in the clinic using workflows that have recently been deployed on our clinical PACs system.

Skin Lesion Extraction Using Multiscale Morphological Local Variance Reconstruction Based Watershed Transform and Fast Fuzzy C-Means Clustering

Early identification of melanocytic skin lesions increases the survival rate for skin cancer patients. Automated melanocytic skin lesion extraction from dermoscopic images using the computer vision approach is a challenging task as the lesions present in the image can be of different colors, there may be a variation of contrast near the lesion boundaries, lesions may have different sizes and shapes, etc. Therefore, lesion extraction from dermoscopic images is a fundamental step for automated melanoma identification. In this article, a watershed transform based on the fast fuzzy c-means (FCM) clustering algorithm is proposed for the extraction of melanocytic skin lesion from dermoscopic images. Initially, the proposed method removes the artifacts from the dermoscopic images and enhances the texture regions. Further, it is filtered using a Gaussian filter and a local variance filter to enhance the lesion boundary regions. Later, the watershed transform based on MMLVR (multiscale morphological local variance reconstruction) is introduced to acquire the superpixels of the image with accurate boundary regions. Finally, the fast FCM clustering technique is implemented in the superpixels of the image to attain the final lesion extraction result. The proposed method is tested in the three publicly available skin lesion image datasets, i.e., ISIC 2016, ISIC 2017 and ISIC 2018. Experimental evaluation shows that the proposed method achieves a good result.

How i do it: biportal endoscopic thoracic decompression for ossification of the ligamentum flavum.

Currently, decompressive laminectomy with or without concomitant fusion is a standard treatment for ossification of the ligamentum flavum (OLF). However, conventional thoracic decompressive laminectomy is not free of the inevitable destruction of the posterior ligamentous complex, and facet injury may lead to various sequelae. We used the biportal endoscopic technique for posterior thoracic decompression (BE-PTD) and describe the steps with discussion regarding the indications, advantages, possible complications, and ways to overcome complications. BE-PTD can obtain endoscopic visualizations of all the boundaries of the OLF lesion and achieve direct neural decompression of thoracic OLF.

DE-Net: A deep edge network with boundary information for automatic skin lesion segmentation

Automatic skin lesion segmentation is one of the most important tasks for computer-aided diagnosis of skin cancer. Although many deep learning-based methods have been proposed, most of them do not pay attention to the boundary information that is critical for achieving more accurate segmentation. In this work, an end-to-end deep edge convolutional neural network (DE-Net) based on the encoder-decoder structure is proposed to pay more attention to the skin lesion boundaries. In the decoder process of the proposed DE-Net, an edge information guided module (EIGM) is designed to introduce the boundary information of the original image and fuse it with different levels of contextual information. Furthermore, to generate more accurate skin lesion boundaries without losing the entire performance and guide the module to highlight more boundary information, we propose an entirety-center-edge (ECE) loss function that can further optimize the boundary details on the basis of the necessary segmentation results. The proposed loss function is backbone-independent and has better performance than other commonly used loss functions in segmentation tasks. In the experiment, the ISIC-2017 dataset is employed to evaluate the effectiveness of the proposed method. It achieves the performance of 0.8792 and 0.8053 on the metrics Dice coefficients and Jaccard index, respectively. Furthermore, we also evaluate the proposed method on ISIC-2016 and PH2 datasets to demonstrate its generalizability. Experimental results demonstrate that the proposed method can outperform the state-of-the-art methods on all these three datasets and has strong generalizability.

Pancreatic cancer segmentation in unregistered multi-parametric MRI with adversarial learning and multi-scale supervision

Automated pancreatic cancer segmentation is crucial for successful clinical aid diagnosis and surgical planning. However, the tiny size and inconspicuous boundaries of pancreatic cancer lesions lead to poor segmentation performance with single-modality imaging. The commonly used registration-based multi-modal fusion strategies may not only introduce uncertainties arising from registration, but also underutilize the complementary information between different modalities. Thus, to achieve different modality-based tumor segmentation, we propose for the first time a registration-free multi-modal and multi-scale adversarial segmentation network (MMSA-Net), which consist of a shared encoder and a dual decoder. Specifically, MMSA-Net combine two complementary modules, inter-modality adversarial learning and intra-modality multi-scale adversarial supervision, to obtain mode-specific segmentation results while facilitating multi-modal fusion. The inter-modality adversarial learning module facilitates the fusion of modality-shared features among different modalities by strengthening the similarity of features extracted by the shared encoder. The intra-modality multi-scale adversarial supervision emphasizes modality-specific features in different decoding paths, inherits and fuses modality-shared features while preserving the feature specificity of each modality, thus outputting competitive modality-specific segmentation results for each modality. Quantitative and qualitative experimental results on multi-parametric MRI pancreatic cancer data show that our method can effectively improve the performance of multi-modal segmentation. The method proposed in this work is expected to be another potential paradigm for addressing multi-modal segmentation tasks in addition to registration. Our source codes will be released at https://github.com/SJTUBME-QianLab/PancreaticCancer_Seg_AdvLearning, once the manuscript is accepted for publication.

Characterizing omental PET/CT findings for differentiating tuberculous peritonitis from peritoneal carcinomatosis.

To characterize and investigate PET/CT findings in the omentum in the differentiation of tuberculous peritonitis (TBP) and peritoneal carcinoma (PC). Thirty-nine patients with TBP and 113 patients with PC who underwent PET/CT were retrospectively enrolled. The omental uptake intensity, distribution characteristics, contracture, size and boundary of soft-tissue lesions, and CT patterns were reviewed. Absent and focal FDG uptake in the lesser omentum was more common in the PC patients (P = 0.034 and P = 0.017, respectively), and diffuse FDG uptake in the lesser omentum was more common in the TBP patients (P < 0.001). An apron-like pattern in the greater omentum commonly occurred in the TBP patients (P = 0.004). Micronodules (< 5mm) were more common in the TBP patients (P < 0.001), and masses (> 3cm) were more common in the PC patients (P = 0.001). Smudged and nodular patterns occurred more frequently in the TBP patients than in the PC patients (P < 0.001 and P = 0.003, respectively), and the caked pattern occurred more frequently in the PC patients (P < 0.001). There was no significant difference in the FDG uptake intensity and the boundary of soft-tissue lesions between the TBP and PC patients (P = 0.191 and P = 0.061, respectively). Diffuse FDG uptake, an apron-like pattern, micronodules, and a smudged and nodular pattern might be significant differential features of TBP. Absent and/or focal FDG uptake, mass, and a caked pattern might be significant differential features of PC.

Region-to-boundary deep learning model with multi-scale feature fusion for medical image segmentation

Accurately locating and segmenting lesions, organs, and tissues from medical images are necessary prerequisites for disease diagnosis, monitoring, and treatment planning. Semantic segmentation refers to the classification of each pixel/voxel in two-dimensional or three-dimensional space, which is beneficial to clinical parameter measurement and disease diagnosis. Due to the diversity of features such as size, shape, location, and intensity, segmenting lesions or organs from medical images has always been a challenging worldwide topic. Especially for low-contrast medical images, boundary recognition is particularly difficult. In this paper, we propose a novel region-to-boundary deep learning model to provide a feasible solution to alleviate this problem. First, we use a U-shaped network with two branches behind the last layer, one of which generates the target probability map, and the other obtains the corresponding signed distance map. Secondly, with the help of the signed distance map and obtained multi-scale features, we focus on the boundary of the target lesions or organs to be segmented. Finally, we fuse the region and boundary features and acquire the final results. We conduct extensive experiments on two public data sets and compare with seven the representative methods. The results show that the proposed model is superior to the comparative methods in most evaluation metrics, especially boundary tracking.

Adaptive Attention Convolutional Neural Network for Liver Tumor Segmentation.

PurposeAccurate segmentation of liver and liver tumors is critical for radiotherapy. Liver tumor segmentation, however, remains a difficult and relevant problem in the field of medical image processing because of the various factors like complex and variable location, size, and shape of liver tumors, low contrast between tumors and normal tissues, and blurred or difficult-to-define lesion boundaries. In this paper, we proposed a neural network (S-Net) that can incorporate attention mechanisms to end-to-end segmentation of liver tumors from CT images.MethodsFirst, this study adopted a classical coding-decoding structure to realize end-to-end segmentation. Next, we introduced an attention mechanism between the contraction path and the expansion path so that the network could encode a longer range of semantic information in the local features and find the corresponding relationship between different channels. Then, we introduced long-hop connections between the layers of the contraction path and the expansion path, so that the semantic information extracted in both paths could be fused. Finally, the application of closed operation was used to dissipate the narrow interruptions and long, thin divide. This eliminated small cavities and produced a noise reduction effect.ResultsIn this paper, we used the MICCAI 2017 liver tumor segmentation (LiTS) challenge dataset, 3DIRCADb dataset and doctors’ manual contours of Hubei Cancer Hospital dataset to test the network architecture. We calculated the Dice Global (DG) score, Dice per Case (DC) score, volumetric overlap error (VOE), average symmetric surface distance (ASSD), and root mean square error (RMSE) to evaluate the accuracy of the architecture for liver tumor segmentation. The segmentation DG for tumor was found to be 0.7555, DC was 0.613, VOE was 0.413, ASSD was 1.186 and RMSE was 1.804. For a small tumor, DG was 0.3246 and DC was 0.3082. For a large tumor, DG was 0.7819 and DC was 0.7632.ConclusionS-Net obtained more semantic information with the introduction of an attention mechanism and long jump connection. Experimental results showed that this method effectively improved the effect of tumor recognition in CT images and could be applied to assist doctors in clinical treatment.

Classification of focal liver lesions in CT images using convolutional neural networks with lesion information augmented patches and synthetic data augmentation.

We propose a deep learning method that classifies focal liver lesions (FLLs) into cysts, hemangiomas, and metastases from portal phase abdominal CT images. We propose a synthetic data augmentation process to alleviate the class imbalance and the Lesion INformation Augmented (LINA) patch to improve the learning efficiency. A dataset of 502 portal phase CT scans of 1,290 FLLs was used. First, to alleviate the class imbalance and to diversify the training data patterns, we suggest synthetic training data augmentation using DCGAN-based lesion mask synthesis and pix2pix-based mask-to-image translation. Second, to improve the learning efficiency of convolutional neural networks (CNNs) for the small lesions, we propose a novel type of input patch termed the LINA patch to emphasize the lesion texture information while also maintaining the lesion boundary information in the patches. Third, we construct a multi-scale CNN through a model ensemble of ResNet-18 CNNs trained on LINA patches of various mini-patch sizes. The experiments demonstrate that (a) synthetic data augmentation method shows characteristics different but complementary to those in conventional real data augmentation in augmenting data distributions, (b) the proposed LINA patches improve classification performance compared to those by existing types of CNN input patches due to the enhanced texture and boundary information in the small lesions, and (c) through an ensemble of LINA patch-trained CNNs with different mini-patch sizes, the multi-scale CNN further improves overall classification performance. As a result, the proposed method achieved an accuracy of 87.30%, showing improvements of 10.81%p and 15.0%p compared to the conventional image patch-trained CNN and texture feature-trained SVM, respectively. The proposed synthetic data augmentation method shows promising results in improving the data diversity and class imbalance, and the proposed LINA patches enhance the learning efficiency compared to the existing input image patches.

Holographic mixed-reality neuronavigation with a head-mounted device: technical feasibility and clinical application.

The authors aimed to evaluate the technical feasibility of a mixed-reality neuronavigation (MRN) system with a wearable head-mounted device (HMD) and to determine its clinical application and accuracy. A semiautomatic registration MRN system on HoloLens smart glasses was developed and tested for accuracy and feasibility. Thirty-seven patients with intracranial lesions were prospectively identified. For each patient, multimodal imaging-based holograms of lesions, markers, and surrounding eloquent structures were created and then imported to the MRN HMD. After a point-based registration, the holograms were projected onto the patient's head and observed through the HMD. The contour of the holograms was compared with standard neuronavigation (SN). The projection of the lesion boundaries perceived by the neurosurgeon on the patient's scalp was then marked with MRN and SN. The distance between the two contours generated by MRN and SN was measured so that the accuracy of MRN could be assessed. MRN localization was achieved in all patients. The mean additional time required for MRN was 36.3 ± 6.3 minutes, in which the mean registration time was 2.6 ± 0.9 minutes. A trend toward a shorter time required for preparation was observed with the increase of neurosurgeon experience with the MRN system. The overall median deviation was 4.1 mm (IQR 3.0 mm-4.7 mm), and 81.1% of the lesions localized by MRN were found to be highly consistent with SN (deviation < 5.0 mm). There was a significant difference between the supine position and the prone position (3.7 ± 1.1 mm vs 5.4 ± 0.9 mm, p = 0.001). The magnitudes of deviation vectors did not correlate with lesion volume (p = 0.126) or depth (p = 0.128). There was no significant difference in additional operating time between different operators (37.4 ± 4.8 minutes vs 34.6 ± 4.8 minutes, p = 0.237) or in localization deviation (3.7 ± 1.0 mm vs 4.6 ± 1.5 mm, p = 0.070). This study provided a complete set of a clinically applicable workflow on an easy-to-use MRN system using a wearable HMD, and has shown its technical feasibility and accuracy. Further development is required to improve the accuracy and clinical efficacy of this system.

B-PO02-133 HISTOLOGICAL CHARACTERIZATION OF REVERSIBLE AND IRREVERSIBLE VENTRICULAR LESION BOUNDARIES PRODUCED BY PULSED FIELD ABLATION

During pulsed field ablation (PFA), electric pulses induce reversible (R) or irreversible (IR) injury on cardiac myocytes. To perform histological examination of acute ventricular lesions produced by PFA, determining R and IR injury boundaries in a swine heart model. A 7F catheter with a 3.5mm ablation electrode (TactiCath, Abbott) was connected to a pulsed electric field (PEF) generator (CENTAURI, Galaxy Medical) and positioned in RV and LV under EnSite guidance in 2 closed chest pigs. Biphasic PEF current was delivered between the ablation electrode and a skin patch at 13 RV sites (28Amp, total pulse width of 1.4ms, 4 pulses) and 19 LV sites (35Amp, 1.6ms, 7 pulses). Two hours after ablation, triphenyl tetrazolium chloride (TTC) was administered. Pigs were sacrificed and hearts were excised and fixed in formalin. Hearts were sectioned and stained with hematoxylin and eosin (H&E) and Masson trichrome. Cytochrome c oxidase (COX) staining was also performed to examine mitochondrial activity to delineate R and IR lesion boundaries. Fig. Ablation lesions were well demarcated with TTC staining, showing a dark central zone surrounded by pale boundaries (PB). Histology showed destruction of myocyte architecture within the PB. A hyperstained (dark red) rim beyond the PB indicates R zone. COX staining showed no or low mitochondrial activity within the PB, consistent with IR ablated region. Enhanced activity of COX staining extended to unaffected normal myocardium, consistent with R zone. Acute ventricular lesions produced by PEF ablation show clear demarcation by TTC staining. COX staining suggests that IR lesions are surrounded by R zone.

A Microfluidic Model Artery for Studying the Mechanobiology of Endothelial Cells

Recent vascular mechanobiology studies find that endothelial cells (ECs) convert multiple mechanical forces into functional responses in a nonadditive way, suggesting that signaling pathways such as those regulating cytoskeleton may be shared among the processes of converting individual forces. However, previous in vitro EC-culture platforms are inherent with extraneous mechanical components, which may saturate or insufficiently activate the shared signaling pathways and accordingly, may misguide EC mechanobiological responses being investigated. Here, a more physiologically relevant model artery is reported that accurately reproduces most of the mechanical forces found in vivo, which can be individually varied in any combination to pathological levels to achieve diseased states. Arterial geometries of normal and diseased states are also realized. By mimicking mechanical microenvironments of early-stage atherosclerosis, it is demonstrated that the elevated levels of the different types of stress experienced by ECs strongly correlate with the disruption of barrier integrity, suggesting that boundaries of an initial lesion could be sites for efficient disease progression.

SLSNet: Skin lesion segmentation using a lightweight generative adversarial network

The determination of precise skin lesion boundaries in dermoscopic images using automated methods faces many challenges, most importantly, the presence of hair, inconspicuous lesion edges and low contrast in dermoscopic images, and variability in the color, texture and shapes of skin lesions. Existing deep learning-based skin lesion segmentation algorithms are expensive in terms of computational time and memory. Consequently, running such segmentation algorithms requires a powerful GPU and high bandwidth memory, which are not available in dermoscopy devices. Thus, this article aims to achieve precise skin lesion segmentation with minimum resources: a lightweight, efficient generative adversarial network (GAN) model called SLSNet, which combines 1-D kernel factorized networks, position and channel attention, and multiscale aggregation mechanisms with a GAN model. The 1-D kernel factorized network reduces the computational cost of 2D filtering. The position and channel attention modules enhance the discriminative ability between the lesion and non-lesion feature representations in spatial and channel dimensions, respectively. A multiscale block is also used to aggregate the coarse-to-fine features of input skin images and reduce the effect of the artifacts. SLSNet is evaluated on two publicly available datasets: ISBI 2017 and the ISIC 2018. Although SLSNet has only 2.35 million parameters, the experimental results demonstrate that it achieves segmentation results on a par with the state-of-the-art skin lesion segmentation methods with an accuracy of 97.61%, and Dice and Jaccard similarity coefficients of 90.63% and 81.98%, respectively. SLSNet can run at more than 110 frames per second (FPS) in a single GTX1080Ti GPU, which is faster than well-known deep learning-based image segmentation models, such as FCN. Therefore, SLSNet can be used for practical dermoscopic applications.

Automated Quantification of Choriocapillaris Lesion Area in Patients With Posterior Uveitis

To validate a custom algorithm for automated identification and quantification of clinically relevant inflammatory choriocapillaris (CC) lesions from en face swept-source optical coherence tomography (SS-OCTA) images. Observational case series. Twenty eyes of 14 patients with posterior uveitis were imaged. The machine-generated en face OCTA CC slabs were exported to a computing platform, where a custom algorithm performed unsupervised lesion boundary delineation and area quantification. Lesions identified by the algorithm (AG) were compared to those identified by 2 masked human graders (HG1 and HG2), using the Sørensen-Dice coefficient (DSC) and intraclass correlation coefficient (ICC). Intragrader and intravisit reliability were determined by coefficient of variation (CV) and DSC. The AG demonstrated excellent agreement with both HGs in determination of lesion area (HG1 vs AG ICC 0.92, 95% CI 0.81-0.97, HG2 vs AG ICC 0.91, 95% CI 0.78-0.97). The AG demonstrated good spatial overlap (DSC ≥0.70) with both HGs in 14 of 20 (70%) eyes and at least 1 HG in 16 of 20 (80%) eyes. Poor spatial overlap (DSC between 0.31 and 0.69) was associated with the presence of a choroidal neovascular membrane and low-contrast lesion boundaries. Intravisit repeatability for the AG was superior to both HGs (CV 2.6% vs >5%). This custom algorithm demonstrated a high degree of agreement with HGs in identification of inflammatory CC lesions and outperformed HGs in reproducibility. Automated CC lesion delineation will support the development of objective and quantitative biomarker of disease activity in patients with posterior uveitis.

Application value of conventional ultrasound combined with shear wave elastography in diagnosing triple negative breast cancer

At present, most ultrasound (US) studies on triple negative breast cancer (TNBC) are limited to conventional US features, so it is necessary to develop new joint diagnostic methods. The study aims to explore the values of conventional US and shear wave elastography (SWE) in differential diagnoses of TNBC and non-TNBC. A total of 120 breast cancer (BC) patients involving 120 lesions that were pathologically verified were retrospectively analyzed in this study. All participants had received both conventional US and SWE before surgery. Meanwhile, the participants were divided into a TNBC group or a non-TNBC group according to their immunohistochemical (IHC) results. The differences between the conventional US features (including lesion shape, growth location, margin, boundary, internal echo, micro-calcification, posterior echo, and internal blood supply) and the SWE image features [including mean lesion hardness (Emean), maximum (Emax), minimum (Emin), standard deviation (SD), and ratio to normal gland (Eratio)] of 2 groups were compared. The receiver operating characteristic (ROC) curve of the diagnosed lesion was calculated by the area under the curve (AUC). According to the findings of conventional US, the TNBC group mostly manifested as a micro-lobulated margin, with a clear boundary and no internal micro-calcification; the non-TNBC group mainly manifested as marginal angulation or burr, and hyper-echo halo in the boundary, accompanied with internal micro-calcification, and the difference was statistically significant (P<0.05); the internal thrombolysis in myocardial infarction (TIMI) and resistance index between the TNBC group and non-TNBC group were similar, and the differences were not statistically significant (P>0.05). The findings of SWE were as follows: differences in Emax, Emean, and Eratio values between 2 groups were statistically significant (P<0.05); and the areas under the ROC curve (AUC) of these three in diagnosing the lesions were 0.811, 0.781 and 0.770, respectively. Conventional US combined with SWE can comprehensively analyze the morphological, blood supply, and hardness features of breast lesions, and provide more reliable information for the differential diagnosis between TNBC and non-TNBC.