Cumulative Dosage of Intrathecal Chemotherapy Agents Predicts White Matter Integrity in Long-Term Survivors of Acute Lymphoblastic Leukemia: A PETALE Study.

Acute lymphoblastic leukemia (ALL) stands as the most prevalent form of pediatric cancer in North America, with a current five-year survival rate of 85%. While more children achieved ALL remission and transition into adulthood, the prevalence of long-term treatment-related effects, especially neurocognitive sequelae, remains significant. This study pursues two objectives. Firstly, it investigates if Magnetization Transfer Ratio (MTR), a method assessing myelin integrity, is sensitive to white matter (WM) microstructural changes in long-term ALL survivors and whether these relate to cognitive impairments. Secondly, it examines the dose-related effects of chemotherapy agents on the MTR and its relationship to other risk factors such as female sex, early age diagnosis, and cranial radiotherapy. Magnetization transfer imaging was utilized to assess WM integrity in 35 survivors at a mean of 18.9 years after the onset of ALL (range since diagnosis: 6.9-26.8). Additionally, 21 controls matched for age, sex, and education level, with no history of cancer, were included. MTR was extracted from both the entire brain's WM and the corpus callosum through semi-automated procedures. The results indicated lower MTR means in survivors, which is linked to cognitive function. Negative associations between MTR means and intrathecal agents' (MTX, cytarabine, and hydrocortisone) cumulative doses received were highlighted. This study offers valuable insights into the connections between myelin deterioration, cognitive impairment, and the implications of IT chemotherapy, enhancing our understanding of ALL survivorship dynamics. It underscores MTR's relevance in monitoring neurotoxicity during oncological drug follow-up examinations.

Attention-gated U-Net networks for simultaneous axial/sagittal planes segmentation of injured spinal cords.

Magnetic resonance imaging is currently the gold standard for the evaluation of spinal cord injuries. Automatic analysis of these injuries is however challenging, as MRI resolutions vary for different planes of analysis and physiological features are often distorted around these injuries. This study proposes a new CNN-based segmentation method in which information is exchanged between two networks analyzing the scans from different planes. Our aim was to develop a robust method for automatic segmentation of the spinal cord in patients having suffered traumatic injuries. The database consisted of 106 sagittal MRI scans from 94 patients with traumatic spinal cord injuries. Our method used an innovative approach where the scans were analyzed in series under the axial and sagittal plane by two different convolutional networks. The results were compared with those of Deepseg 2D from the Spinal Cord Toolbox (SCT), which was taken as state-of-the-art. Comparisons were evaluated using K-Fold cross-validation combined with statistical t-test results on separate test data. Our method achieved significantly better results than Deepseg 2D, with an average Dice coefficient of 0.95 against 0.88 for Deepseg 2D (p <0.001). Other metrics were also used to compare the segmentations, all of which showed significantly better results for our approach. In this study, we introduce a robust method for spinal cord segmentation which is capable of adequately segmenting spinal cords affected by traumatic injuries, improving upon the methods contained in SCT.

Automated detection of regions of interest in cartridge case images using deep learning.

This paper explores a deep-learning approach to evaluate the position of circular delimiters in cartridge case images. These delimiters define two regions of interest (ROI), corresponding to the breech face and the firing pin impressions, and are placed manually or by an image-processing algorithm. This positioning bears a significant impact on the performance of the image-matching algorithms for firearm identification, and an automated evaluation method would be beneficial to any computerized system. Our contribution consists in optimizing and training U-Net segmentation models from digital images of cartridge cases, intending to locate ROIs automatically. For the experiments, we used high-resolution 2D images from 1195 samples of cartridge cases fired by different 9MM firearms. Our results show that the segmentation models, trained on augmented data sets, exhibit a performance of 95.6% IoU (Intersection over Union) and 99.3% DC (Dice Coefficient) with a loss of 0.014 for the breech face images; and a performance of 95.9% IoU and 99.5% DC with a loss of 0.011 for the firing pin images. We observed that the natural shapes of predicted circles reduce the performance of segmentation models compared with perfect circles on ground truth masks suggesting that our method provide a more accurate segmentation of the real ROI shape. In practice, we believe that these results could be useful for firearms identification. In future work, the predictions may be used to evaluate the quality of delimiters on specimens in a database, or they could determine the region of interest on a cartridge case image.

Unsupervised synthesis of realistic coronary artery X-ray angiogram.

PURPOSE: Medical image analysis suffers from a sparsity of annotated data necessary in learning-based models. Cardiorespiratory simulators have been developed to counter the lack of data. However, the resulting data often lack realism. Hence, the proposed method aims to synthesize realistic and fully customizable angiograms of coronary arteries for the training of learning-based biomedical tasks, for cardiologists performing interventions, and for cardiologist trainees. METHODS: 3D models of coronary arteries are generated with a fully customizable realistic cardiorespiratory simulator. The transfer of X-ray angiography style to simulator-generated images is performed using a new vessel-specific adaptation of the CycleGAN model. The CycleGAN model is paired with a vesselness-based loss function that is designed as a vessel-specific structural integrity constraint. RESULTS: Validation is performed both on the style and on the preservation of the shape of the arteries of the images. The results show a PSNR of 14.125, an SSIM of 0.898, and an overlapping of 89.5% using the Dice coefficient. CONCLUSION: We proposed a novel fluoroscopy-based style transfer method for the enhancement of the realism of simulated coronary artery angiograms. The results show that the proposed model is capable of accurately transferring the style of X-ray angiograms to the simulations while keeping the integrity of the structures of interest (i.e., the topology of the coronary arteries).

Patient-specific Cardio-respiratory Motion Prediction in X-ray Angiography using LSTM Networks.

Objective.To develop a novel patient-specific cardio-respiratory motion prediction approach for X-ray angiography time series based on a simple long short-term memory (LSTM) model.Approach.The cardio-respiratory motion behavior in an X-ray image sequence was represented as a sequence of 2D affine transformation matrices, which provide the displacement information of contrasted moving objects (arteries and medical devices) in a sequence. The displacement information includes translation, rotation, shearing, and scaling in 2D. A many-to-many LSTM model was developed to predict 2D transformation parameters in matrix form for future frames based on previously generated images. The method was developed with 64 simulated phantom datasets (pediatric and adult patients) using a realistic cardio-respiratory motion simulator (XCAT) and was validated using 10 different patient X-ray angiography sequences.Main results.Using this method we achieved less than 1 mm prediction error for complex cardio-respiratory motion prediction. The following mean prediction error values were recorded over all the simulated sequences: 0.39 mm (for both motions), 0.33 mm (for only cardiac motion), and 0.47 mm (for only respiratory motion). The mean prediction error for the patient dataset was 0.58 mm.Significance.This study paves the road for a patient-specific cardio-respiratory motion prediction model, which might improve navigation guidance during cardiac interventions.

Navigation guidance for ventricular septal defect closure in heart phantoms.

PURPOSE: Transesophageal echocardiography (TEE) is the preferred imaging modality in a hybrid procedure used to close ventricular septal defects (VSDs). However, the limited field of view of TEE hinders the maneuvering of surgical instruments inside the beating heart. This study evaluates the accuracy of a method that aims to support navigation guidance in the hybrid procedure. METHODS: A cardiologist maneuvered a needle to puncture the patient's heart and to access a VSD, guided by information displayed in a virtual environment. The information displayed included a model of the patient's heart and a virtual needle that reproduced the position and orientation of the real needle in real time. The physical and the virtual worlds were calibrated with a landmark registration and an iterative closest point algorithms, using an electromagnetic measurement system (EMS). For experiments, we developed a setup that included heart phantoms representing the patient's heart. RESULTS: Experimental results from two pediatric cases studied suggested that the information provided for guidance was accurate enough when the landmark registration algorithm was fed with coordinates of seven points clearly identified on the surfaces of the physical and virtual hearts. Indeed, with a registration error of 2.28 mm RMS, it was possible to successfully access two VSDs (6.2 mm and 6.3 mm in diameter) in all the attempts with a needle (5 attempts) and a guidewire (7 attempts). CONCLUSION: We found that information provided in a virtual environment facilitates guidance in the hybrid procedure for VSD closure. A clear identification of anatomical details in the heart surfaces is key to the accuracy of the procedure.

Characterization of blood-mimicking fluids for echocardiography imaging of ventricular septal defects.

PURPOSE: Ventricular septal defects (VSD) are congenital heart malformations and, in severe cases, they require complex interventions under echocardiography guidance. Heart phantoms can be helpful to train and to understand the complex hemodynamics of VSD. The goal of this study was to characterize the best blood mimicking fluids in such heart phantoms for modelling the hemodynamics of VSD patients using echocardiography. METHODS: Four fluid compositions were considered. Distilled water was used as a baseline, while the other three fluids were developed based on physical properties of human blood, such as the viscosity and the refractive index. Three bi-ventricular heart phantoms of three different pediatric patients with complex VSD were designed from preoperative CT imaging. Custom molds were printed in 3-D and the anatomical structure was casted in polyvinyl alcohol cryogel. The VSD in each heart phantom were observed using echocardiography and color Doppler imaging was used for the hemodynamic study. RESULTS: Heart phantoms with blood mimicking fluids of 30% glycerol and 27% glycerol, 10% sodium iodide were found to be anatomically realistic under echocardiography imaging. Hemodynamic parameters such as the pressure gradient and the volume of the shunt were characterized using color Doppler imaging. CONCLUSION: Proper composition of blood mimicking fluids are important for improving the realism in echocardiographic heart phantoms and they contribute to better understand the complex hemodynamic of VSD under echocardiography.

Generative learning approach for radiation dose reduction in X-ray guided cardiac interventions.

BACKGROUND: Navigation guidance in cardiac interventions is provided by X-ray angiography. Cumulative radiation exposure is a serious concern for pediatric cardiac interventions. PURPOSE: A generative learning-based approach is proposed to predict X-ray angiography frames to reduce the radiation exposure for pediatric cardiac interventions while preserving the image quality. METHODS: Frame predictions are based on a model-free motion estimation approach using a long short-term memory architecture and a content predictor using a convolutional neural network structure. The presented model thus estimates contrast-enhanced vascular structures such as the coronary arteries and their motion in X-ray sequences in an end-to-end system. This work was validated with 56 simulated and 52 patients' X-ray angiography sequences. RESULTS: Using the predicted images can reduce the number of pulses by up to three new frames without affecting the image quality. The average required acquisition can drop by 30% per second for a 15 fps acquisition. The average structural similarity index measurement was 97% for the simulated dataset and 82% for the patients' dataset. CONCLUSIONS: Frame prediction using a learning-based method is promising for minimizing radiation dose exposure. The required pulse rate is reduced while preserving the frame rate and the image quality. With proper integration in X-ray angiography systems, this method can pave the way for improved dose management.

Design of heart phantoms for ultrasound imaging of ventricular septal defects.

PURPOSE: Ventricular septal defects (VSDs) are common congenital heart malformations. Echocardiography used during VSD hybrid cardiac procedures requires extensive training for image acquisition and interpretation. Cardiac surgery simulators with heart phantoms have shown usefulness for such training, but they are limited in visualization and characterization of complex VSD. This study explores a new method to build patient-specific heart phantoms with VSD, with proper tissue echogenicity for ultrasound imaging. METHODS: Heart phantoms were designed from preoperative imaging of three patients with complex VSDs. Each whole heart phantom, including atrial and ventricular septums, was obtained by manual segmentation and by surface reconstruction, then by molding and by casting in different materials. Heart phantoms in silicone and polyvinyl alcohol cryogel (PVA-C) were considered, and they were reconstructed in 3-D using 2-D freehand ultrasound imaging. RESULTS: An electromagnetic measurement system was used to measure the mean VSD diameters from the heart phantoms. Errors were evaluated below 1.0 mm for mean VSD diameters between 6.2 and 7.5 mm. CONCLUSION: Patient-specific heart phantoms promise for representing complex heart malformations such as VSDs. PVA-C showed better tissue echogenicity than silicone for VSDs visualization and characterization.

Intravascular imaging of coronary artery: Bridging the gap between clinical needs and technical advances.

Coronary artery disease is the leading cause of mortality worldwide. Almost seven million deaths are reported each year due to coronary disease. Coronary artery events in the adult are primarily due to atherosclerosis with seventy-five percent of the related mortality caused by plaque rupture. Despite significant progress made to improve intravascular imaging of coronary arteries, there is still a large gap between clinical needs and technical developments. The goal of this review is to identify the gap elements between clinical knowledge and recent advances in the domain of medical image analysis. Efficient image analysis computational models should be designed with respect to the exact clinical needs, and detailed features of the tissues under review. In this review, we discuss the detailed clinical features of the intracoronary plaques for mathematical and biomedical researchers. We emphasize the importance of integrating this clinical knowledge validated by clinicians to investigate the potentially effective models for proper features efficiency in the scope of leveraging the state-of-the-art of coronary image analyses.

Automatic bone maturity grading from EOS radiographs in Adolescent Idiopathic Scoliosis.

Adolescent Idiopathic Scoliosis (AIS) is a deformation of the spine and it is routinely diagnosed using posteroanterior and lateral radiographs. The Risser sign used in skeletal maturity assessment is commonly accepted in AIS patient's management. However, the Risser sign is subject to inter-observer variability and it relies mainly on the observation of ossification on the iliac crests. This study proposes a new machine-learning-based approach for Risser sign skeletal maturity assessment using EOS radiographs. Regions of interest including right and left humeral heads; left and right femoral heads; and pelvis are extracted from the radiographs. First, a total of 24 image features is extracted from EOS radiographs using a ResNet101-type convolutional neural network (CNN), pre-trained from the ImageNet database. Then, a support vector machine (SVM) algorithm is used for the final Risser sign classification. The experimental results demonstrate an overall accuracy of 84%, 78%, and 80% respectively for iliac crests, humeral heads, and femoral heads. Class activation maps using Grad-CAM were also investigated to understand the features of our model. In conclusion, our machine learning approach is promising to incorporate a large number of image features for different regions of interest to improve Risser grading for skeletal maturity. Automatic classification could contribute to the management of AIS patients.

A deep learning-based model for characterization of atherosclerotic plaque in coronary arteries using optical coherence tomography images.

PURPOSE: Coronary artery events are mainly associated with atherosclerosis in adult population, which is recognized as accumulation of plaques in arterial wall tissues. Optical Coherence Tomography (OCT) is a light-based imaging system used in cardiology to analyze intracoronary tissue layers and pathological formations including plaque accumulation. This state-of-the-art catheter-based imaging system provides intracoronary cross-sectional images with high resolution of 10-15 µm. But interpretation of the acquired images is operator dependent, which is not only very time-consuming but also highly error prone from one observer to another. An automatic and accurate coronary plaque tagging using OCT image post-processing can contribute to wide adoption of the OCT system and reducing the diagnostic error rate. METHOD: In this study, we propose a combination of spatial pyramid pooling module with dilated convolutions for semantic segmentation to extract atherosclerotic tissues regardless of their types and training a sparse auto-encoder to reconstruct the input features and enlarge the training data as well as plaque type characterization in OCT images. RESULTS: The results demonstrate high precision of the proposed model with reduced computational complexity, which can be appropriate for real-time analysis of OCT images. At each step of the work, measured accuracy, sensitivity, specificity of more than 93% demonstrate high performance of the model. CONCLUSION: The main focus of this study is atherosclerotic tissue characterization using OCT imaging. This contributes to wide adoption of the OCT imaging system by providing clinicians with a fully automatic interpretation of various atherosclerotic tissues. Future studies will be focused on analyzing atherosclerotic vulnerable plaques, those coronary plaques which are prone to rupture.

Real-time biomechanics using the finite element method and machine learning: Review and perspective.

PURPOSE: The finite element method (FEM) is the preferred method to simulate phenomena in anatomical structures. However, purely FEM-based mechanical simulations require considerable time, limiting their use in clinical applications that require real-time responses, such as haptics simulators. Machine learning (ML) approaches have been proposed to help with the reduction of the required time. The present paper reviews cases where ML could help to generate faster simulations, without considerably affecting the performance results. METHODS: This review details the ML approaches used, considering the anatomical structures involved, the data collection strategies, the selected ML algorithms, with corresponding features, the metrics used for validation, and the resulting time gains. RESULTS: A total of 41 references were found. ML algorithms are mainly trained with FEM-based simulations in 32 publications. The preferred ML approach is neural networks, including deep learning in 35 publications. Tissue deformation is simulated in 18 applications, but other features are also considered. The average distance error and mean squared error are the most frequently used performance metrics, in 14 and 17 publications, respectively. The time gains were considerable, going from hours or minutes for purely FEM-based simulations to milliseconds, when using ML. CONCLUSIONS: ML algorithms can be used to accelerate FEM-based biomechanical simulations of anatomical structures, possibly reaching real-time responses. Fast and real-time simulations of anatomical structures, generated with ML algorithms, can help to reduce the time required by FEM-based simulations and accelerate their adoption in the clinical practice.

A freehand ultrasound framework for spine assessment in 3D: a preliminary study.

X-ray imaging is currently the gold standard for the assessment of spinal deformities. The purpose of this study is to evaluate a freehand 3D ultrasound system for volumetric reconstruction of the spine. A setup consisting of an ultrasound scanner with a linear transducer, an electromagnetic measuring system and a workstation was used. We conducted 64 acquisitions of US images of 8 adults in a natural standing position, and we tested three setups: 1) Subjects are constrained to be close to a wall, 2) Subjects are unconstrained, and 3) Subjects are constrained to performing fast and slow acquisitions. The spinous processes were manually selected from the volume reconstruction from tracked ultrasound images to generate a 3D point-based model depicting the centerline of the spine. The results suggested that a freehand 3D ultrasound system can be suitable for representing the spine. Volumetric reconstructions can be computed and landmarking can be performed to model the surface of the spine in the 3D space. These reconstructions promise to generate computer-based descriptors to analyze the shape of the spine in the 3D space.Clinical Relevance- We provide clinicians with a protocol that could be integrated in clinical setups for the assessment and monitoring of AIS, based on US image acquisitions, which constitutes a radiation-free technology.

Semi-supervised generative adversarial networks for the segmentation of the left ventricle in pediatric MRI.

Segmentation of the left ventricle in magnetic resonance imaging (MRI) is important for assessing cardiac function. We present DT-GAN, a generative adversarial network (GAN) segmentation approach for the identification of the left ventricle in pediatric MRI. Segmentation of the left ventricle requires a large amount of annotated data; generating such data can be time-consuming and subject to observer variability. Additionally, it can be difficult to accomplish in a clinical setting. During the training of our GAN, we therefore introduce a semi-supervised semantic segmentation to reduce the number of images required for training, while maintaining a good segmentation accuracy. The GAN generator produces a segmentation label map and its discriminator outputs a confidence map, which gives the probability of a pixel coming from the label or from the generator. Moreover, we propose a new formulation of the GAN loss function based on distance transform and pixel-wise cross-entropy. This new loss function provides a better segmentation of boundary pixels, by favoring the correct classification of those pixels rather than focusing on pixels that are farther away from the boundary between anatomical structures. Our proposed method achieves a mean Hausdorff distance of 2.16 mm ± 0.42 mm (2.28 mm ± 0.21 mm for U-Net) and a Dice score of 0.88 ± 0.08 (0.91 ± 0.12 for U-Net) for the endocardium segmentation, using 50% of the annotated data. For the epicardium segmentation, we achieve a mean Hausdorff distance of 2.23 mm ± 0.35 mm (2.34 mm ± 0.39 mm for U-Net) and a Dice score of 0.93 mm ± 0.04 mm (0.89 ± 0.09 for U-Net). For the myocardium segmentation, we achieve a mean Hausdorff distance of 2.98 mm ± 0.43 mm (3.04 mm ± 0.27 mm for U-Net) and a Dice score of 0.79 mm ± 0.10 mm (0.74 ± 0.04 for U-Net). This new model could be very useful for the automatic analysis of cardiac MRI and for conducting large-scale studies based on MRI readings, with a limited amount of training data.

Convolutional Neural Networks for Automatic Risser Stage Assessment.

PURPOSE: To develop an automatic method for the assessment of the Risser stage using deep learning that could be used in the management panel of adolescent idiopathic scoliosis (AIS). MATERIALS AND METHODS: In this institutional review board approved-study, a total of 1830 posteroanterior radiographs of patients with AIS (age range, 10-18 years, 70% female) were collected retrospectively and graded manually by six trained readers using the United States Risser staging system. Each radiograph was preprocessed and cropped to include the entire pelvic region. A convolutional neural network was trained to automatically grade conventional radiographs according to the Risser classification. The network was then validated by comparing its accuracy against the interobserver variability of six trained graders from the authors' institution using the Fleiss κ statistical measure. RESULTS: Overall agreement between the six observers was fair, with a κ coefficient of 0.65 for the experienced graders and agreement of 74.5%. The automatic grading method obtained a κ coefficient of 0.72, which is a substantial agreement with the ground truth, and an overall accuracy of 78.0%. CONCLUSION: The high accuracy of the model presented here compared with human readers suggests that this work may provide a new method for standardization of Risser grading. The model could assist physicians with the task, as well as provide additional insights in the assessment of bone maturity based on radiographs.© RSNA, 2020.

An automatic diagnostic system of coronary artery lesions in Kawasaki disease using intravascular optical coherence tomography imaging.

Intravascular optical coherence tomography (IV-OCT) is a light-based imaging modality with high resolution, which employs near-infrared light to provide tomographic intracoronary images. Morbidity caused by coronary heart disease is a substantial cause of acute coronary syndrome and sudden cardiac death. The most common intracoronay complications caused by coronary artery disease are intimal hyperplasia, calcification, fibrosis, neovascularization and macrophage accumulation, which require efficient prevention strategies. OCT can provide discriminative information of the intracoronary tissues, which can be used to train a robust fully automatic tissue characterization model based on deep learning. In this study, we aimed to design a diagnostic model of coronary artery lesions. Particularly, we trained a random forest using convolutional neural network features to distinguish between normal and diseased arterial wall structure. Then, based on the arterial wall structure, fully convolutional network is designed to extract the tissue layers in normal cases, and pathological tissues regardless of lesion type in pathological cases. Then, the type of the lesions can be characterized with high precision using our previous model. The results demonstrate the robustness of the model with the approximate overall accuracy up to 90%.

CycleGAN for style transfer in X-ray angiography.

PURPOSE: We aim to perform generation of angiograms for various vascular structures as a mean of data augmentation in learning tasks. The task is to enhance the realism of vessels images generated from an anatomically realistic cardiorespiratory simulator to make them look like real angiographies. METHODS: The enhancement is performed by applying the CycleGAN deep network for transferring the style of real angiograms acquired during percutaneous interventions into a data set composed of realistically simulated arteries. RESULTS: The cycle consistency was evaluated by comparing an input simulated image with the one obtained after two cycles of image translation. An average structural similarity (SSIM) of 0.948 on our data sets has been obtained. The vessel preservation was measured by comparing segmentations of an input image and its corresponding enhanced image using Dice coefficient. CONCLUSIONS: We proposed an application of the CycleGAN deep network for enhancing the artificial data as an alternative to classical data augmentation techniques for medical applications, particularly focused on angiogram generation. We discussed success and failure cases, explaining conditions for the realistic data augmentation which respects both the complex physiology of arteries and the various patterns and textures generated by X-ray angiography.

Intra-Slice Motion Correction of Intravascular OCT Images Using Deep Features.

Intra-slice motion correction is an important step for analyzing volume variations and pathological formations from intravascular imaging. Optical coherence tomography (OCT) has been recently introduced for intravascular imaging and assessment of coronary artery disease. Two-dimensional (2-D) cross-sectional OCT images of coronary arteries play a crucial role to characterize the internal structure of the tissues. Adjacent images could be compounded; however, they might not fully match due to motion, which is a major hurdle for analyzing longitudinally each tissue in 3-D. The aim of this study is to develop a robust tissue-matching-based motion correction approach from a sequence of 2-D intracoronary OCT images. Our motion correction technique is based on the correlation between deep features obtained from a convolutional neural network (CNN) for each frame of a sequence. The optimal transformation of each frame is obtained by maximizing the similarity between the tissues of reference and moving frames. The results show a good alignment of the tissues after applying CNN features and determining the transformation parameters.

Epistemic Uncertainty Modeling for Vessel Segmentation.

X-ray angiograms are currently the gold-standard in percutaneous guidance during cardiovascular interventions. However, due to lack of contrast, to overlapping artifacts and to the rapid dilution of the contrast agent, they remain difficult to analyze either by cardiologists, or automatically by computers. Providing, a general yet accurate multi-arteries segmentation method along with the uncertainty linked to those segmentations would not only ease the analysis of medical imaging by cardiologists, but also provide a required pre-processing of the data for tasks ranging from 3D reconstruction to motion tracking of arteries. The proposed method has been validated on clinical data providing an average accuracy of 94.9%. Additionally, results show good transposition of learning from one type of artery to another. Epistemic uncertainty maps provide areas where the segmentation should be validated by an expert before being used, and could provide identification of regions of interest for data augmentation purposes.