Segmentation Of Liver Lesions Research Articles

Fully Convolutional DenseNet with Attention Mechanism for Liver Lesion Segmentation in MRI Images

The goal of this research is to achieve accurate segmentation of liver tumors in noncontrast T2-weighted magnetic resonance imaging. As liver tumors and adjacent organs are represented by pixels of very similar gray intensity, segmentation is challenging, and the presence of different sizes of liver tumor makes segmentation more difficult. Differing from previous work to capture contextual information using multiscale feature fusion with concatenation, attention mechanism is added to our segmentation model to extract precise global contextual information for pixel labeling without requiring complex dilated convolution. This study describe a liver lesion segmentation model derived from FC-DenseNet with attention mechanism. Specifically, a global attention module (GAM) is added to up-sampling path, and high-level features are processed by the GAM to generating weighting information for guiding high resolution detail features recovery. High-level features are very effective for accurate category classification, but relatively weak at pixel classification and predicting restoration of the original resolution, so the fusion of high-level semantic features and low-level detail features can improve segmentation accuracy. A weighted focal loss function is used to solve the problem of lesion area occupying a relatively low proportion of the whole image, and to deal with the disequilibrium of foreground and background in the training liver lesion images. Experimental results show our segmentation model can automatically segment liver tumors from complete MRI images, and the addition of the GAM model can effectively improve liver tumor segmentation. Our algorithms have obvious advantages over other CNN algorithms and traditional manual methods of feature extraction.

Cascaded Deeply Supervised Convolutional Networks for Liver Lesion Segmentation

Liver lesion segmentation from abdomen computed tomography (CT) with deep neural networks remains challenging due to the small volume and the unclear boundary. To effectively tackle these problems, in this paper, we propose a cascaded deeply supervised convolutional networks (CDS-Net). The cascaded deep supervision (CDS) mechanism uses auxiliary losses to construct a cascaded segmentation method in a single network, focusing the network attention on pixels that are more difficult to classify, so that the network can segment the lesion more effectively. CDS mechanism can be easily integrated into standard CNN models and it helps to increase the model sensitivity and prediction accuracy. Based on CDS mechanism, we propose a cascaded deep supervised ResUNet, which is an end-to-end liver lesion segmentation network. We conduct experiments on LiTS and 3DIRCADb dataset. Our method has achieved competitive results compared with other state-of-the-art ones.

An automatic method for segmentation of liver lesions in computed tomography images using deep neural networks

Liver cancer is one of the major causes of death by cancer. The early detection of lesions in the liver provides a better chance of treatment and cure of the disease. Computed tomography (CT) is one of the most used imaging techniques for the detection and diagnosis of liver lesions. However, the manual segmentation of liver and tumors, aside from being time-consuming, can still cause errors and may vary among specialists. Because of this hard work, computer-aided detection (CAD) and computer-aided diagnosis (CADx) systems have been developed to assist specialists in the detection and characterization of lesions in the liver and reduce the required time for diagnosis. The automatic segmentation of these lesions is a complex task since they present variability in contrast, shape, size, and location. In this work, a method to automatically segment liver lesions in CT images is proposed. The proposed method, which presents two deep convolutional neural networks (CNN) models, consists of five main steps: (1) image acquisition, (2) image pre-processing, (3) initial segmentation using RetinaNet, (4) lesion segmentation using U-Net, and (5) segmentation refinement. The proposed method was evaluated using a set of 131 CT images from the LiTS dataset, and the best result obtained a matthews correlation coefficient (MCC) of 83.62%, a sensitivity of 83.86%, a specificity of 99.96%, a Dice coefficient of 82.99%, a volumetric overlap error (VOE) of 27.89%, and a relative volume difference (RVD) of 1.69%. We show in our method that the problem of segmentation of liver lesions in CT images can be efficiently solved through the use of deep CNNs to define the scope of the problem and to precisely segment lesions.

Segmentation of focal liver lesions and virtual resection based on computed tomography data

The aim of the study was to test the work of plugins for segmentation and virtual resection of focal liver lesions based on CT data. Materials and methods. Analysis of CT data of the abdominal organs with bolus contrast enhancement in 80 patients with focal liver lesions was carried out. Segmentation and 3D-modeling of the CT data was carried out by radiologists and the surgeon in the ‘Autoplan’ system. Results . The liver nosological structure in patients was determined (the most common were hemangiomas in 21.25% of 80 patients, cysts in 20%, parasitic cysts in 20%, etc.), according to the computed tomography results. The segmentation of the liver, its focal lesions, arteries and veins was carried out using the ‘Autoplan’ system. The surgeon determined the volume of the parenchyma and focal liver formations using the standard function ‘volume of segmentation’, chose the optimal treatment tactics and performed a virtual liver resection. In some cases, the use of segmentation and preoperative planning made it possible to avoid an inefficient surgery. The effectiveness of modeling changed the treatment tactics of 42 patients. Conclusion. The obtained results indicate that the use of the ‘Autoplan’ system plugins for planning an abdominal surgery allows doctors: 1) to carry out the segmentation of liver, focal lesions and blood vessels; 2) to determine the location of a focal formation in a particular segment, their combination; 3) to perform a virtual resection, evaluate the structures passing through it; 4) to choose the optimal tactics of intervention or abandon it due to objective anatomical reasons.

Segmentation of Liver Lesions Without Contrast Agents With Radiomics-Guided Densely UNet-Nested GAN

Segmentation of liver lesions on non-contrast magnetic resonance imaging (MRI) is critical for patient management and treatment planning. In clinical treatment, the imaging process suffers from high-risk, expensive, and time-consuming due to using contrast agents (CA). Furthermore, manual segmentation has the disadvantages of tedious, low-reproducibility, and high misdiagnosis rate. Although some deep-learning based works have attempted for liver lesions segmentation, they are all limited to the use of contrast-enhanced MRI. To avoid the limitations comes from CA, we proposed a Radiomics-guided Densely-UNet-Nested Generative Adversarial Networks (Radiomics-guided DUN-GAN) for automatic segmentation of liver lesions on non-contrast MRI. Radiomics-guided DUN-GAN includes a DUN segmentor and a Radiomics-guided discriminator. It uses radiomics feature of the multi-phase contrast image as prior knowledge to guide the extraction of key implicit contrast radiomics (ICR) features in non-contrast images, thus achieving the direct lesions segmentation without CA for the first time. In the DUN segmentor, an innovative nested structure of Densely-UNet-connection reliably completes the segmentation. The nested structure extracts global features, semantic features, and ICR features by reasonably sharing features and maximizing information flow. Those features are fused with a new direction strategy of multi-integration features to improve the segmentation ability. In the innovative Radiomics-guided discriminator, the radiomics feature combined with the semantic feature enhances the discrimination of Radiomics-guided discriminator. Moreover, it guides the segmentor for multiple feature extraction via using the adversarial mechanism. Radiomics-guided DUN-GAN learns the mapping relationship between images, extracting the key ICR in the non-contrast image, and finally completing the accurate segmentation. Radiomics-guided DUN-GAN obtained the Dice Similarity Coefficient results of 93.47± 0.83% for the segmentation of lesions in non-contrast images from 250 clinical subjects. The results verify the Radiomics-guided DUN-GAN is accurate and robust, and it has the possibility of becoming a safe, inexpensive, and time-saving medical assistant tool in clinical diagnosis.

Multi-stage fuzzy swarm intelligence for automatic hepatic lesion segmentation from CT scans

Segmentation of liver and hepatic lesions using computed tomography (CT) is a critical and challenging task for doctors to accurately identify liver abnormalities and to reduce the risk of liver surgery. This study proposed a novel dynamic approach to improve the fuzzy c-means (FCM) clustering algorithm for automatic localization and segmentation of liver and hepatic lesions from CT scans. More specifically, we developed a powerful optimization approach in terms of accuracy, speed, and optimal convergence based on fast-FCM, chaos theory, and bio-inspired ant lion optimizer (ALO), named (CALOFCM), for automatic liver and hepatic lesion segmentation. We employed ALO to guide the FCM to determine the optimal cluster centroids for segmentation processes. We used chaos theory to improve the performance of ALO in terms of convergence speed and local minima avoidance. In addition, chaos theory-based ALO prevented the FCM from getting stuck in local minima and increased computational performance, thus increasing stability, reducing sensitivity in the iterative process, and allowing the best centroids to be used by FCM. We validated the proposed approach on a group of patients with abdominal liver CT images, and the results showed good detection and segmentation performance compared with other popular techniques. This new hybrid approach allowed for the clinical diagnosis of hepatic lesions earlier and more systematically, thereby helping medical experts in their decision-making.

Liver Segmentation on a Variety of Computed Tomography (CT) Images Based on Convolutional Neural Networks Combined with Connected Components

Liver Segmentation on a Variety of Computed Tomography (CT) Images Based on Convolutional Neural Networks Combined with Connected Components

Semiautomatic Segmentation and Radiomics for Dual-Energy CT: A Pilot Study to Differentiate Benign and Malignant Hepatic Lesions.

OBJECTIVE. This study assessed a machine learning-based dual-energy CT (DECT) tumor analysis prototype for semiautomatic segmentation and radiomic analysis of benign and malignant liver lesions seen on contrast-enhanced DECT. MATERIALS AND METHODS. This institutional review board-approved study included 103 adult patients (mean age, 65 ± 15 [SD] years; 53 men, 50 women) with benign (60/103) or malignant (43/103) hepatic lesions on contrast-enhanced dual-source DECT. Most malignant lesions were histologically proven; benign lesions were either stable on follow-up CT or had characteristic benign features on MRI. Low- and high-kilovoltage datasets were deidentified, exported offline, and processed with the DECT tumor analysis for semiautomatic segmentation of the volume and rim of each liver lesion. For each segmentation, contrast enhancement and iodine concentrations as well as radiomic features were derived for different DECT image series. Statistical analyses were performed to determine if DECT tumor analysis and radiomics can differentiate benign from malignant liver lesions. RESULTS. Normalized iodine concentration and mean iodine concentration in the benign and malignant lesions were significantly different (p < 0.0001-0.0084; AUC, 0.695-0.856). Iodine quantification and radiomic features from lesion rims (AUC, ≤ 0.877) had higher accuracy for differentiating liver lesions compared with the values from lesion volumes (AUC, ≤ 0.856). There was no difference in the accuracies of DECT iodine quantification (AUC, 0.91) and radiomics (AUC, 0.90) for characterizing liver lesions. CONCLUSION. DECT radiomics were more accurate than iodine quantification for differentiating solid benign and malignant hepatic lesions.

Correction

Correction

A Novel Bio-Inspired Deep Learning Approach for Liver Cancer Diagnosis

Current research on computer-aided diagnosis (CAD) of liver cancer is based on traditional feature engineering methods, which have several drawbacks including redundant features and high computational cost. Recent deep learning models overcome these problems by implicitly capturing intricate structures from large-scale medical image data. However, they are still affected by network hyperparameters and topology. Hence, the state of the art in this area can be further optimized by integrating bio-inspired concepts into deep learning models. This work proposes a novel bio-inspired deep learning approach for optimizing predictive results of liver cancer. This approach contributes to the literature in two ways. Firstly, a novel hybrid segmentation algorithm is proposed to extract liver lesions from computed tomography (CT) images using SegNet network, UNet network, and artificial bee colony optimization (ABC), namely, SegNet-UNet-ABC. This algorithm uses the SegNet for separating liver from the abdominal CT scan, then the UNet is used to extract lesions from the liver. In parallel, the ABC algorithm is hybridized with each network to tune its hyperparameters, as they highly affect the segmentation performance. Secondly, a hybrid algorithm of the LeNet-5 model and ABC algorithm, namely, LeNet-5/ABC, is proposed as feature extractor and classifier of liver lesions. The LeNet-5/ABC algorithm uses the ABC to select the optimal topology for constructing the LeNet-5 network, as network structure affects learning time and classification accuracy. For assessing performance of the two proposed algorithms, comparisons have been made to the state-of-the-art algorithms on liver lesion segmentation and classification. The results reveal that the SegNet-UNet-ABC is superior to other compared algorithms regarding Jaccard index, Dice index, correlation coefficient, and convergence time. Moreover, the LeNet-5/ABC algorithm outperforms other algorithms regarding specificity, F1-score, accuracy, and computational time.

The impact of irreversible image data compression on post-processing algorithms in computed tomography.

PURPOSE We aimed to evaluate the influence of irreversible image compression at varying levels on image post-processing algorithms (3D volume rendering of angiographs, computer-assisted detection of lung nodules, segmentation and volumetry of liver lesions, and automated evaluation of functional cardiac imaging) in computed tomography (CT). METHODS Uncompressed CT image data (30 angiographs of the lower limbs, 38 lung exams, 20 liver exams and 30 cardiac exams) were anonymized and subsequently compressed using the JPEG2000 algorithm with compression ratios of 8:1, 10:1, and 15:1. Volume renderings of CT angiographies obtained from compressed and uncompressed data were compared using objective and subjective measures. Computer-assisted detection of lung nodules was performed on compressed and uncompressed image data and compared with respect to diagnostic performance. Segmentation and volumetry of liver lesions as well as measurement of ejection fraction on cardiac studies was performed on compressed and uncompressed datasets; differences in measurements were analyzed. RESULTS No differences could be detected for the 3D volume renderings and no statistically significant differences in performance were found for the computer-assisted detection algorithm. Measurements in volumetry of liver lesions and functional cardiac imaging showed good to excellent reliability. CONCLUSION Irreversible image compression within the limits proposed by the European Society of Radiology has no significant influence on commonly used image post-processing algorithms in CT.

Liver Cancer Detection Using Hybridized Fully Convolutional Neural Network Based on Deep Learning Framework

Liver cancer is one of the world’s largest causes of death to humans. It is a difficult task and time consuming to identify the cancer tissue manually in the present scenario. The segmentation of liver lesions in CT images can be used to assess the tumor load, plan treatments predict, and monitor the clinical response. In this paper, the Hybridized Fully Convolutional Neural Network (HFCNN) has been proposed for liver tumor segmentation, which has been modeled mathematically to resolve the current issue of liver cancer. For semantic segmentation, HFCNN has been used as a powerful tool for liver cancer analysis. Whereas the CT-based lesion-type definition defines the diagnosis and therapeutic strategy, the distinction between cancer and non-cancer lesions is crucial. It demands highly qualified experience, expertise, and resources. However, a deep end-to-end learning approach to help discrimination in abdominal CT images of the liver between liver metastases of colorectal cancer and benign cysts has been analyzed. Our method includes the successful extraction of features from Inception combined with residual and pre-trained weights. Feature maps have been consistent with the original image voxel features, and The importance of features seemed to represent the most relevant imaging criteria for every class. This deep learning system shows the concept of illumination portions of the decision-making process of a pre-trained deep neural network, through an analysis of inner layers and the description of features that lead to predictions.

Liver Tumor Segmentation and Subsequent Risk Prediction Based on Deeplabv3+

As the largest glandular organ in the human body, liver has a large number of blood vessels and is connected with many important organs, such as spleen, pancreas and gallbladder, etc. The segmentation of liver and its lesions on medical images can help doctors accurately diagnose liver tumor and assess the probability of subsequent deterioration of the patient. Generally speaking, it is not only subjective but also wastes time if doctors rely on experience to manually analyze liver CT images. Therefore, it has been extensively studied in recent years. The segmentation of liver lesions is a kind of challenging task due to the low contrast ratio between the liver, lesions and nearby organs. To this end, we proposed to use the DeepLabV3+ semantic segmentation model based on the tensorflow architecture to segment the CT image of liver and locate the lesion positions. It combined deep convolutional neural networks (DCNNs) and probabilistic graphical model (DenseCRFs) and has been proven to have very good performance in a variety of computer vision tasks.

Hierarchical Fine-Tuning for joint Liver Lesion Segmentation and Lesion Classification in CT.

We present an automatic method for joint liver lesion segmentation and classification using a hierarchical fine-tuning framework. Our dataset is small, containing 332 2-D CT examinations with lesion annotated into 3 lesion types: cysts, hemangiomas, and metastases. Using a cascaded U-net that performs segmentation and classification simultaneously, we trained a strong lesion segmentation model on the dataset of MICCAI 2017 Liver Tumor Segmentation (LiTS) Challenge. We used the trained weights to fine-tune a slightly modified model to obtain improved lesion segmentation and classification, on the smaller dataset. Since pre-training was done with similar data on a related task, we were able to learn more representative features (especially higher-level features in the U-Net's encoder), and improve pixel-wise classification results. We show an improvement of over 10% in Dice score and classification accuracy, compared to a baseline model. We further improve the classification performance by hierarchically freezing the encoder part of the network and achieve an improvement of over 15% in Dice score and classification accuracy. We compare our results with an existing method and show an improvement of 14% in the success rate and 12% in the classification accuracy.

Deep Learning for Automated Segmentation of Liver Lesions at CT in Patients with Colorectal Cancer Liver Metastases.

To evaluate the performance, agreement, and efficiency of a fully convolutional network (FCN) for liver lesion detection and segmentation at CT examinations in patients with colorectal liver metastases (CLMs). This retrospective study evaluated an automated method using an FCN that was trained, validated, and tested with 115, 15, and 26 contrast material-enhanced CT examinations containing 261, 22, and 105 lesions, respectively. Manual detection and segmentation by a radiologist was the reference standard. Performance of fully automated and user-corrected segmentations was compared with that of manual segmentations. The interuser agreement and interaction time of manual and user-corrected segmentations were assessed. Analyses included sensitivity and positive predictive value of detection, segmentation accuracy, Cohen κ, Bland-Altman analyses, and analysis of variance. In the test cohort, for lesion size smaller than 10 mm (n = 30), 10-20 mm (n = 35), and larger than 20 mm (n = 40), the detection sensitivity of the automated method was 10%, 71%, and 85%; positive predictive value was 25%, 83%, and 94%; Dice similarity coefficient was 0.14, 0.53, and 0.68; maximum symmetric surface distance was 5.2, 6.0, and 10.4 mm; and average symmetric surface distance was 2.7, 1.7, and 2.8 mm, respectively. For manual and user-corrected segmentation, κ values were 0.42 (95% confidence interval: 0.24, 0.63) and 0.52 (95% confidence interval: 0.36, 0.72); normalized interreader agreement for lesion volume was -0.10 ± 0.07 (95% confidence interval) and -0.10 ± 0.08; and mean interaction time was 7.7 minutes ± 2.4 (standard deviation) and 4.8 minutes ± 2.1 (P < .001), respectively. Automated detection and segmentation of CLM by using deep learning with convolutional neural networks, when manually corrected, improved efficiency but did not substantially change agreement on volumetric measurements.© RSNA, 2019Supplemental material is available for this article.

Computer-Aided Segmentation of Liver Lesions in CT Scans Using Cascaded Convolutional Neural Networks and Genetically Optimised Classifier

Abdominal CT scans have been widely studied and researched by medical professionals in recent years. CT scans have proved effective for the task of detection of liver abnormalities in patients. Computer-aided automatic segmentation of the liver can serve as an elementary step for radiologists to trace anomalies in the liver. In this paper, we have explored deep learning techniques first and foremost for the extraction of liver from the abdominal CT scan and then, consequently, to segment the lesions from a tumour-ridden liver. A cascaded model of convolutional neural networks is used to segment lesions once tumour has been detected in the liver by GA-ANN which has been fed textural liver features using LTEM for its classification procedure. A high DICE index has been obtained of 0.9557 for liver segmentation and 0.6976 for lesion segmentation.

3D segmentation of liver and its lesions using optimized geometric contours

An optimised accurate segmentation of liver and hepatic lesions from Computed Tomography (CT) scans can support the study of deriving quantitative biomarkers for accurate clinical diagnosis and successful conclusion of computer-assisted diagnosis and therapy. In spite of several existence of research, fully automatic segmentation of liver and its lesions still remains a challenging task. This paper presents a method for delineation of liver and its lesions in CT volume images using energy optimised geometric active contours. The abdominal region was initially determined based on maximum intensity projection (MIP) and thresholding methods to extract the liver volume of interest (VOI). Two geometric contours using level set method were initialized and cascaded for a combined segmentation of the liver and its lesions. In the first step, a contour to segment the liver was initialized that served as VOI input for a second geometric contour. The second contour exclusively segmented lesions within the predicted liver VOIs of first step. Experiments on CT image data obtained dice coefficient of 0.900, jaccard coefficient of 0.818 and other surface distance measures demonstrate the advantages of the proposed methodology in both exactness and competence.

Automatic Identification of the Optimal Reference Frame for Segmentation and Quantification of Focal Liver Lesions in Contrast-Enhanced Ultrasound

Post-examination interpretation of contrast-enhanced ultrasound (CEUS) cineloops of focal liver lesions (FLLs) requires offline manual assessment by experienced radiologists, which is time-consuming and generates subjective results. Such assessment usually starts by manually identifying a reference frame, where FLL and healthy parenchyma are well-distinguished. This study proposes an automatic computational method to objectively identify the optimal reference frame for distinguishing and hence delineating an FLL, by statistically analyzing the temporal intensity variation across the spatially discretized ultrasonographic image. Level of confidence and clinical value of the proposed method were quantitatively evaluated on retrospective multi-institutional data (n = 64) and compared with expert interpretations. Results support the proposed method for facilitating easier, quicker and reproducible assessment of FLLs, further increasing the radiologists' confidence in diagnostic decisions. Finally, our method yields a useful training tool for radiologists, widening CEUS use in non-specialist centers, potentially leading to reduced turnaround times and lower patient anxiety and healthcare costs.

MR-based synthetic CT generation using a deep convolutional neural network method

Interests have been rapidly growing in the field of radiotherapy to replace CT with magnetic resonance imaging (MRI), due to superior soft tissue contrast offered by MRI and the desire to reduce unnecessary radiation dose. MR-only radiotherapy also simplifies clinical workflow and avoids uncertainties in aligning MR with CT. Methods, however, are needed to derive CT-equivalent representations, often known as synthetic CT (sCT), from patient MR images for dose calculation and DRR-based patient positioning. Synthetic CT estimation is also important for PET attenuation correction in hybrid PET-MR systems. We propose in this work a novel deep convolutional neural network (DCNN) method for sCT generation and evaluate its performance on a set of brain tumor patient images. The proposed method builds upon recent developments of deep learning and convolutional neural networks in the computer vision literature. The proposed DCNN model has 27 convolutional layers interleaved with pooling and unpooling layers and 35 million free parameters, which can be trained to learn a direct end-to-end mapping from MR images to their corresponding CTs. Training such a large model on our limited data is made possible through the principle of transfer learning and by initializing model weights from a pretrained model. Eighteen brain tumor patients with both CT and T1-weighted MR images are used as experimental data and a sixfold cross-validation study is performed. Each sCT generated is compared against the real CT image of the same patient on a voxel-by-voxel basis. Comparison is also made with respect to an atlas-based approach that involves deformable atlas registration and patch-based atlas fusion. The proposed DCNN method produced a mean absolute error (MAE) below 85 HU for 13 of the 18 test subjects. The overall average MAE was 84.8 ± 17.3 HU for all subjects, which was found to be significantly better than the average MAE of 94.5 ± 17.8 HU for the atlas-based method. The DCNN method also provided significantly better accuracy when being evaluated using two other metrics: the mean squared error (188.6 ± 33.7 versus 198.3 ± 33.0) and the Pearson correlation coefficient(0.906 ± 0.03 versus 0.896 ± 0.03). Although training a DCNN model can be slow, training only need be done once. Applying a trained model to generate a complete sCT volume for each new patient MR image only took 9 s, which was much faster than the atlas-based approach. A DCNN model method was developed, and shown to be able to produce highly accurate sCT estimations from conventional, single-sequence MR images in near real time. Quantitative results also showed that the proposed method competed favorably with an atlas-based method, in terms of both accuracy and computation speed at test time. Further validation on dose computation accuracy and on a larger patient cohort is warranted. Extensions of the method are also possible to further improve accuracy or to handle multi-sequence MR images.

Adaptive local window for level set segmentation of CT and MRI liver lesions.

We propose a novel method, the adaptive local window, for improving level set segmentation technique. The window is estimated separately for each contour point, over iterations of the segmentation process, and for each individual object. Our method considers the object scale, the spatial texture, and the changes of the energy functional over iterations. Global and local statistics are considered by calculating several gray level co-occurrence matrices. We demonstrate the capabilities of the method in the domain of medical imaging for segmenting 233 images with liver lesions. To illustrate the strength of our method, those lesions were screened by either Computed Tomography or Magnetic Resonance Imaging. Moreover, we analyzed images using three different energy models. We compared our method to a global level set segmentation, to a local framework that uses predefined fixed-size square windows and to a local region-scalable fitting model. The results indicate that our proposed method outperforms the other methods in terms of agreement with the manual marking and dependence on contour initialization or the energy model used. In case of complex lesions, such as low contrast lesions, heterogeneous lesions, or lesions with a noisy background, our method shows significantly better segmentation with an improvement of 0.25 ± 0.13 in Dice similarity coefficient, compared with state of the art fixed-size local windows (Wilcoxon, p < 0.001).