Accurate and robust deep learning-based segmentation of the prostate clinical target volume in ultrasound images

Artificial intelligence uncertainty quantification in radiotherapy applications - A scoping review

BACKGROUND/PURPOSE

The use of artificial intelligence (AI) in radiotherapy (RT) is expanding rapidly. However, there exists a notable lack of clinician trust in AI models, underscoring the need for effective uncertainty quantification (UQ) methods. The purpose of this study was to scope existing literature related to UQ in RT, identify areas of improvement, and determine future directions.

METHODS

We followed the PRISMA-ScR scoping review reporting guidelines. We utilized the population (human cancer patients), concept (utilization of AI UQ), context (radiotherapy applications) framework to structure our search and screening process. We conducted a systematic search spanning seven databases, supplemented by manual curation, up to January 2024. Our search yielded a total of 8980 articles for initial review. Manuscript screening and data extraction was performed in Covidence. Data extraction categories included general study characteristics, RT characteristics, AI characteristics, and UQ characteristics.

RESULTS

We identified 56 articles published from 2015 to 2024. 10 domains of RT applications were represented; most studies evaluated auto-contouring (50 %), followed by image-synthesis (13 %), and multiple applications simultaneously (11 %). 12 disease sites were represented, with head and neck cancer being the most common disease site independent of application space (32 %). Imaging data was used in 91 % of studies, while only 13 % incorporated RT dose information. Most studies focused on failure detection as the main application of UQ (60 %), with Monte Carlo dropout being the most commonly implemented UQ method (32 %) followed by ensembling (16 %). 55 % of studies did not share code or datasets.

CONCLUSION

Our review revealed a lack of diversity in UQ for RT applications beyond auto-contouring. Moreover, we identified a clear need to study additional UQ methods, such as conformal prediction. Our results may incentivize the development of guidelines for reporting and implementation of UQ in RT.

Artificial Intelligence Uncertainty Quantification in Radiotherapy Applications - A Scoping Review

Background/purpose

The use of artificial intelligence (AI) in radiotherapy (RT) is expanding rapidly. However, there exists a notable lack of clinician trust in AI models, underscoring the need for effective uncertainty quantification (UQ) methods. The purpose of this study was to scope existing literature related to UQ in RT, identify areas of improvement, and determine future directions.

Methods

We followed the PRISMA-ScR scoping review reporting guidelines. We utilized the population (human cancer patients), concept (utilization of AI UQ), context (radiotherapy applications) framework to structure our search and screening process. We conducted a systematic search spanning seven databases, supplemented by manual curation, up to January 2024. Our search yielded a total of 8980 articles for initial review. Manuscript screening and data extraction was performed in Covidence. Data extraction categories included general study characteristics, RT characteristics, AI characteristics, and UQ characteristics.

Results

We identified 56 articles published from 2015-2024. 10 domains of RT applications were represented; most studies evaluated auto-contouring (50%), followed by image-synthesis (13%), and multiple applications simultaneously (11%). 12 disease sites were represented, with head and neck cancer being the most common disease site independent of application space (32%). Imaging data was used in 91% of studies, while only 13% incorporated RT dose information. Most studies focused on failure detection as the main application of UQ (60%), with Monte Carlo dropout being the most commonly implemented UQ method (32%) followed by ensembling (16%). 55% of studies did not share code or datasets.

Conclusion

Our review revealed a lack of diversity in UQ for RT applications beyond auto-contouring. Moreover, there was a clear need to study additional UQ methods, such as conformal prediction. Our results may incentivize the development of guidelines for reporting and implementation of UQ in RT.

The use of deep learning in interventional radiotherapy (brachytherapy): A review with a focus on open source and open data

Deep learning advanced to one of the most important technologies in almost all medical fields. Especially in areas, related to medical imaging it plays a big role. However, in interventional radiotherapy (brachytherapy) deep learning is still in an early phase. In this review, first, we investigated and scrutinised the role of deep learning in all processes of interventional radiotherapy and directly related fields. Additionally, we summarised the most recent developments. For better understanding, we provide explanations of key terms and approaches to solving common deep learning problems. To reproduce results of deep learning algorithms both source code and training data must be available. Therefore, a second focus of this work is on the analysis of the availability of open source, open data and open models. In our analysis, we were able to show that deep learning plays already a major role in some areas of interventional radiotherapy, but is still hardly present in others. Nevertheless, its impact is increasing with the years, partly self-propelled but also influenced by closely related fields. Open source, data and models are growing in number but are still scarce and unevenly distributed among different research groups. The reluctance in publishing code, data and models limits reproducibility and restricts evaluation to mono-institutional datasets. The conclusion of our analysis is that deep learning can positively change the workflow of interventional radiotherapy but there is still room for improvements when it comes to reproducible results and standardised evaluation methods.

Deep learning-based ultrasound auto-segmentation of the prostate with brachytherapy implanted needles

BACKGROUND

Accurate segmentation of the clinical target volume (CTV) corresponding to the prostate with or without proximal seminal vesicles is required on transrectal ultrasound (TRUS) images during prostate brachytherapy procedures. Implanted needles cause artifacts that may make this task difficult and time-consuming. Thus, previous studies have focused on the simpler problem of segmentation in the absence of needles at the cost of reduced clinical utility.

PURPOSE

To use a convolutional neural network (CNN) algorithm for segmentation of the prostatic CTV in TRUS images post-needle insertion obtained from prostate brachytherapy procedures to better meet the demands of the clinical procedure.

METHODS

A dataset consisting of 144 3-dimensional (3D) TRUS images with implanted metal brachytherapy needles and associated manual CTV segmentations was used for training a 2-dimensional (2D) U-Net CNN using a Dice Similarity Coefficient (DSC) loss function. These were split by patient, with 119 used for training and 25 reserved for testing. The 3D TRUS training images were resliced at radial (around the axis normal to the coronal plane) and oblique angles through the center of the 3D image, as well as axial, coronal, and sagittal planes to obtain 3689 2D TRUS images and masks for training. The network generated boundary predictions on 300 2D TRUS images obtained from reslicing each of the 25 3D TRUS images used for testing into 12 radial slices (15° apart), which were then reconstructed into 3D surfaces. Performance metrics included DSC, recall, precision, unsigned and signed volume percentage differences (VPD/sVPD), mean surface distance (MSD), and Hausdorff distance (HD). In addition, we studied whether providing algorithm-predicted boundaries to the physicians and allowing modifications increased the agreement between physicians. This was performed by providing a subset of 3D TRUS images of five patients to five physicians who segmented the CTV using clinical software and repeated this at least 1 week apart. The five physicians were given the algorithm boundary predictions and allowed to modify them, and the resulting inter- and intra-physician variability was evaluated.

RESULTS

Median DSC, recall, precision, VPD, sVPD, MSD, and HD of the 3D-reconstructed algorithm segmentations were 87.2 [84.1, 88.8]%, 89.0 [86.3, 92.4]%, 86.6 [78.5, 90.8]%, 10.3 [4.5, 18.4]%, 2.0 [-4.5, 18.4]%, 1.6 [1.2, 2.0] mm, and 6.0 [5.3, 8.0] mm, respectively. Segmentation time for a set of 12 2D radial images was 2.46 [2.44, 2.48] s. With and without U-Net starting points, the intra-physician median DSCs were 97.0 [96.3, 97.8]%, and 94.4 [92.5, 95.4]% (p < 0.0001), respectively, while the inter-physician median DSCs were 94.8 [93.3, 96.8]% and 90.2 [88.7, 92.1]%, respectively (p < 0.0001). The median segmentation time for physicians, with and without U-Net-generated CTV boundaries, were 257.5 [211.8, 300.0] s and 288.0 [232.0, 333.5] s, respectively (p = 0.1034).

CONCLUSIONS

Our algorithm performed at a level similar to physicians in a fraction of the time. The use of algorithm-generated boundaries as a starting point and allowing modifications reduced physician variability, although it did not significantly reduce the time compared to manual segmentations.

MicroSegNet: A deep learning approach for prostate segmentation on micro-ultrasound images

Micro-ultrasound (micro-US) is a novel 29-MHz ultrasound technique that provides 3-4 times higher resolution than traditional ultrasound, potentially enabling low-cost, accurate diagnosis of prostate cancer. Accurate prostate segmentation is crucial for prostate volume measurement, cancer diagnosis, prostate biopsy, and treatment planning. However, prostate segmentation on micro-US is challenging due to artifacts and indistinct borders between the prostate, bladder, and urethra in the midline. This paper presents MicroSegNet, a multi-scale annotation-guided transformer UNet model designed specifically to tackle these challenges. During the training process, MicroSegNet focuses more on regions that are hard to segment (hard regions), characterized by discrepancies between expert and non-expert annotations. We achieve this by proposing an annotation-guided binary cross entropy (AG-BCE) loss that assigns a larger weight to prediction errors in hard regions and a lower weight to prediction errors in easy regions. The AG-BCE loss was seamlessly integrated into the training process through the utilization of multi-scale deep supervision, enabling MicroSegNet to capture global contextual dependencies and local information at various scales. We trained our model using micro-US images from 55 patients, followed by evaluation on 20 patients. Our MicroSegNet model achieved a Dice coefficient of 0.939 and a Hausdorff distance of 2.02 mm, outperforming several state-of-the-art segmentation methods, as well as three human annotators with different experience levels. Our code is publicly available at https://github.com/mirthAI/MicroSegNet and our dataset is publicly available at https://zenodo.org/records/10475293.

A review of traditional Chinese medicine diagnosis using machine learning: Inspection, auscultation-olfaction, inquiry, and palpation

Traditional Chinese medicine (TCM) is an essential part of the Chinese medical system and is recognized by the World Health Organization as an important alternative medicine. As an important part of TCM, TCM diagnosis is a method to understand a patient's illness, analyze its state, and identify syndromes. In the long-term clinical diagnosis practice of TCM, four fundamental and effective diagnostic methods of inspection, auscultation-olfaction, inquiry, and palpation (IAOIP) have been formed. However, the diagnostic information in TCM is diverse, and the diagnostic process depends on doctors' experience, which is subject to a high-level subjectivity. At present, the research on the automated diagnosis of TCM based on machine learning is booming. Machine learning, which includes deep learning, is an essential part of artificial intelligence (AI), which provides new ideas for the objective and AI-related research of TCM. This paper aims to review and summarize the current research status of machine learning in TCM diagnosis. First, we review some key factors for the application of machine learning in TCM diagnosis, including data, data preprocessing, machine learning models, and evaluation metrics. Second, we review and summarize the research and applications of machine learning methods in TCM IAOIP and the synthesis of the four diagnostic methods. Finally, we discuss the challenges and research directions of using machine learning methods for TCM diagnosis.

Deep learning for automated segmentation in radiotherapy: a narrative review

The segmentation of organs and structures is a critical component of radiation therapy planning, with manual segmentation being a laborious and time-consuming task. Interobserver variability can also impact the outcomes of radiation therapy. Deep neural networks have recently gained attention for their ability to automate segmentation tasks, with convolutional neural networks (CNNs) being a popular approach. This article provides a descriptive review of the literature on deep learning (DL) techniques for segmentation in radiation therapy planning. This review focuses on five clinical sub-sites and finds that U-net is the most commonly used CNN architecture. The studies using DL for image segmentation were included in brain, head and neck, lung, abdominal, and pelvic cancers. The majority of DL segmentation articles in radiation therapy planning have concentrated on normal tissue structures. N-fold cross-validation was commonly employed, without external validation. This research area is expanding quickly, and standardization of metrics and independent validation are critical to benchmarking and comparing proposed methods.

Densely connected convolutional networks for ultrasound image based lesion segmentation

Delineating lesion boundaries play a central role in diagnosing thyroid and breast cancers, making related therapy plans and evaluating therapeutic effects. However, it is often time-consuming and error-prone with limited reproducibility to manually annotate low-quality ultrasound (US) images, given high speckle noises, heterogeneous appearances, ambiguous boundaries etc., especially for nodular lesions with huge intra-class variance. It is hence appreciative but challenging for accurate lesion segmentations from US images in clinical practices. In this study, we propose a new densely connected convolutional network (called MDenseNet) architecture to automatically segment nodular lesions from 2D US images, which is first pre-trained over ImageNet database (called PMDenseNet) and then retrained upon the given US image datasets. Moreover, we also designed a deep MDenseNet with pre-training strategy (PDMDenseNet) for segmentation of thyroid and breast nodules by adding a dense block to increase the depth of our MDenseNet. Extensive experiments demonstrate that the proposed MDenseNet-based method can accurately extract multiple nodular lesions, with even complex shapes, from input thyroid and breast US images. Moreover, additional experiments show that the introduced MDenseNet-based method also outperforms three state-of-the-art convolutional neural networks in terms of accuracy and reproducibility. Meanwhile, promising results in nodular lesion segmentation from thyroid and breast US images illustrate its great potential in many other clinical segmentation tasks.

ADNet++: A few-shot learning framework for multi-class medical image volume segmentation with uncertainty-guided feature refinement

A major barrier to applying deep segmentation models in the medical domain is their typical data-hungry nature, requiring experts to collect and label large amounts of data for training. As a reaction, prototypical few-shot segmentation (FSS) models have recently gained traction as data-efficient alternatives. Nevertheless, despite the recent progress of these models, they still have some essential shortcomings that must be addressed. In this work, we focus on three of these shortcomings: (i) the lack of uncertainty estimation, (ii) the lack of a guiding mechanism to help locate edges and encourage spatial consistency in the segmentation maps, and (iii) the models' inability to do one-step multi-class segmentation. Without modifying or requiring a specific backbone architecture, we propose a modified prototype extraction module that facilitates the computation of uncertainty maps in prototypical FSS models, and show that the resulting maps are useful indicators of the model uncertainty. To improve the segmentation around boundaries and to encourage spatial consistency, we propose a novel feature refinement module that leverages structural information in the input space to help guide the segmentation in the feature space. Furthermore, we demonstrate how uncertainty maps can be used to automatically guide this feature refinement. Finally, to avoid ambiguous voxel predictions that occur when images are segmented class-by-class, we propose a procedure to perform one-step multi-class FSS. The efficiency of our proposed methodology is evaluated on two representative datasets for abdominal organ segmentation (CHAOS dataset and BTCV dataset) and one dataset for cardiac segmentation (MS-CMRSeg dataset). The results show that our proposed methodology significantly (one-sided Wilcoxon signed rank test, p<0.05) improves the baseline, increasing the overall dice score with +5.2, +5.1, and +2.8 percentage points for the CHAOS dataset, the BTCV dataset, and the MS-CMRSeg dataset, respectively.

Observer preference of artificial intelligence-generated versus clinical prostate contours for ultrasound-based high dose rate brachytherapy

BACKGROUND

For trans-rectal ultrasound (TRUS)-based high dose rate (HDR) prostate brachytherapy, prostate contouring can be challenging due to artifacts from implanted needles, bleeding, and calcifications.

PURPOSE

To evaluate the geometric accuracy and observer preference of an artificial intelligence (AI) algorithm for generating prostate contours on TRUS images with implanted needles.

METHODS

We conducted a retrospective study of 150 patients, who underwent HDR brachytherapy. These patients were randomly divided into training (104), validation (26) and testing (20) sets. An AI algorithm was trained/validated utilizing the TRUS image and reference (clinical) contours. The algorithm then provided contours for the test set. For evaluation, we calculated the Dice coefficient between AI and reference prostate contours. We then presented AI and reference contours to eight clinician observers, and asked observers to select their preference. Observers were blinded to the source of contours. We calculated the percentage of cases in which observers preferred AI contours. Lastly, we evaluate whether the presence of AI contours improved the geometric accuracy of prostate contours provided by five resident observers for a 10-patient subset.

RESULTS

The median Dice coefficient between AI and reference contours was 0.92 (IQR: 0.90-0.94). Observers preferred AI contours for a median of 57.5% (IQR: 47.5, 65.0) of the test cases. For resident observers, the presence of AI contours was associated with a 0.107 (95% CI: 0.086, 0.128; p < 0.001) improvement in Dice coefficient for the 10-patient subset.

CONCLUSION

The AI algorithm provided high-quality prostate contours on TRUS with implanted needles. Further prospective study is needed to better understand how to incorporate AI prostate contours into the TRUS-based HDR brachytherapy workflow.

The Big Bang of Deep Learning in Ultrasound-Guided Surgery: A Review

Ultrasound (US) imaging is a paramount modality in many image-guided surgeries and percutaneous interventions, thanks to its high portability, temporal resolution, and cost-efficiency. However, due to its imaging principles, the US is often noisy and difficult to interpret. Appropriate image processing can greatly enhance the applicability of the imaging modality in clinical practice. Compared with the classic iterative optimization and machine learning (ML) approach, deep learning (DL) algorithms have shown great performance in terms of accuracy and efficiency for US processing. In this work, we conduct a comprehensive review on deep-learning algorithms in the applications of US-guided interventions, summarize the current trends, and suggest future directions on the topic.

Artificial intelligence applications in brachytherapy: A literature review

PURPOSE

Artificial intelligence (AI) has the potential to simplify and optimize various steps of the brachytherapy workflow, and this literature review aims to provide an overview of the work done in this field.

METHODS AND MATERIALS

We conducted a literature search in June 2022 on PubMed, Embase, and Cochrane for papers that proposed AI applications in brachytherapy.

RESULTS

A total of 80 papers satisfied inclusion/exclusion criteria. These papers were categorized as follows: segmentation (24), registration and image processing (6), preplanning (13), dose prediction and treatment planning (11), applicator/catheter/needle reconstruction (16), and quality assurance (10). AI techniques ranged from classical models such as support vector machines and decision tree-based learning to newer techniques such as U-Net and deep reinforcement learning, and were applied to facilitate small steps of a process (e.g., optimizing applicator selection) or even automate the entire step of the workflow (e.g., end-to-end preplanning). Many of these algorithms demonstrated human-level performance and offer significant improvements in speed.

CONCLUSIONS

AI has potential to augment, automate, and/or accelerate many steps of the brachytherapy workflow. We recommend that future studies adhere to standard reporting guidelines. We also stress the importance of using larger sample sizes and reporting results using clinically interpretable measures.

Research progress on deep learning in magnetic resonance imaging-based diagnosis and treatment of prostate cancer: a review on the current status and perspectives

Multiparametric magnetic resonance imaging (mpMRI) has emerged as a first-line screening and diagnostic tool for prostate cancer, aiding in treatment selection and noninvasive radiotherapy guidance. However, the manual interpretation of MRI data is challenging and time-consuming, which may impact sensitivity and specificity. With recent technological advances, artificial intelligence (AI) in the form of computer-aided diagnosis (CAD) based on MRI data has been applied to prostate cancer diagnosis and treatment. Among AI techniques, deep learning involving convolutional neural networks contributes to detection, segmentation, scoring, grading, and prognostic evaluation of prostate cancer. CAD systems have automatic operation, rapid processing, and accuracy, incorporating multiple sequences of multiparametric MRI data of the prostate gland into the deep learning model. Thus, they have become a research direction of great interest, especially in smart healthcare. This review highlights the current progress of deep learning technology in MRI-based diagnosis and treatment of prostate cancer. The key elements of deep learning-based MRI image processing in CAD systems and radiotherapy of prostate cancer are briefly described, making it understandable not only for radiologists but also for general physicians without specialized imaging interpretation training. Deep learning technology enables lesion identification, detection, and segmentation, grading and scoring of prostate cancer, and prediction of postoperative recurrence and prognostic outcomes. The diagnostic accuracy of deep learning can be improved by optimizing models and algorithms, expanding medical database resources, and combining multi-omics data and comprehensive analysis of various morphological data. Deep learning has the potential to become the key diagnostic method in prostate cancer diagnosis and treatment in the future.

Ultrasound Prostate Segmentation Using Adaptive Selection Principal Curve and Smooth Mathematical Model

Accurate prostate segmentation in ultrasound images is crucial for the clinical diagnosis of prostate cancer and for performing image-guided prostate surgery. However, it is challenging to accurately segment the prostate in ultrasound images due to their low signal-to-noise ratio, the low contrast between the prostate and neighboring tissues, and the diffuse or invisible boundaries of the prostate. In this paper, we develop a novel hybrid method for segmentation of the prostate in ultrasound images that generates accurate contours of the prostate from a range of datasets. Our method involves three key steps: (1) application of a principal curve-based method to obtain a data sequence comprising data coordinates and their corresponding projection index; (2) use of the projection index as training input for a fractional-order-based neural network that increases the accuracy of results; and (3) generation of a smooth mathematical map (expressed via the parameters of the neural network) that affords a smooth prostate boundary, which represents the output of the neural network (i.e., optimized vertices) and matches the ground truth contour. Experimental evaluation of our method and several other state-of-the-art segmentation methods on datasets of prostate ultrasound images generated at multiple institutions demonstrated that our method exhibited the best capability. Furthermore, our method is robust as it can be applied to segment prostate ultrasound images obtained at multiple institutions based on various evaluation metrics.

Uncertainty Estimation for Heatmap-Based Landmark Localization

Automatic anatomical landmark localization has made great strides by leveraging deep learning methods in recent years. The ability to quantify the uncertainty of these predictions is a vital component needed for these methods to be adopted in clinical settings, where it is imperative that erroneous predictions are caught and corrected. We propose Quantile Binning, a data-driven method to categorize predictions by uncertainty with estimated error bounds. Our framework can be applied to any continuous uncertainty measure, allowing straightforward identification of the best subset of predictions with accompanying estimated error bounds. We facilitate easy comparison between uncertainty measures by constructing two evaluation metrics derived from Quantile Binning. We compare and contrast three epistemic uncertainty measures (two baselines, and a proposed method combining aspects of the two), derived from two heatmap-based landmark localization model paradigms (U-Net and patch-based). We show results across three datasets, including a publicly available Cephalometric dataset. We illustrate how filtering out gross mispredictions caught in our Quantile Bins significantly improves the proportion of predictions under an acceptable error threshold. Finally, we demonstrate that Quantile Binning remains effective on landmarks with high aleatoric uncertainty caused by inherent landmark ambiguity, and offer recommendations on which uncertainty measure to use and how to use it. The code and data are available at https://github.com/schobs/qbin.

Learning to segment fetal brain tissue from noisy annotations

Automatic fetal brain tissue segmentation can enhance the quantitative assessment of brain development at this critical stage. Deep learning methods represent the state of the art in medical image segmentation and have also achieved impressive results in brain segmentation. However, effective training of a deep learning model to perform this task requires a large number of training images to represent the rapid development of the transient fetal brain structures. On the other hand, manual multi-label segmentation of a large number of 3D images is prohibitive. To address this challenge, we segmented 272 training images, covering 19-39 gestational weeks, using an automatic multi-atlas segmentation strategy based on deformable registration and probabilistic atlas fusion, and manually corrected large errors in those segmentations. Since this process generated a large training dataset with noisy segmentations, we developed a novel label smoothing procedure and a loss function to train a deep learning model with smoothed noisy segmentations. Our proposed methods properly account for the uncertainty in tissue boundaries. We evaluated our method on 23 manually-segmented test images of a separate set of fetuses. Results show that our method achieves an average Dice similarity coefficient of 0.893 and 0.916 for the transient structures of younger and older fetuses, respectively. Our method generated results that were significantly more accurate than several state-of-the-art methods including nnU-Net that achieved the closest results to our method. Our trained model can serve as a valuable tool to enhance the accuracy and reproducibility of fetal brain analysis in MRI.

Improving Calibration and Out-of-Distribution Detection in Deep Models for Medical Image Segmentation

Convolutional Neural Networks (CNNs) have proved to be powerful medical image segmentation models. In this study, we address some of the main unresolved issues regarding these models. Specifically, training of these models on small medical image datasets is still challenging, with many studies promoting techniques such as transfer learning. Moreover, these models are infamous for producing over-confident predictions and for failing silently when presented with out-of-distribution (OOD) test data. In this paper, for improving prediction calibration we advocate for multi-task learning, i.e., training a single model on several different datasets, spanning different organs of interest and different imaging modalities. We show that multi-task learning can significantly improve model confidence calibration. For OOD detection, we propose a novel method based on spectral analysis of CNN feature maps. We show that different datasets, representing different imaging modalities and/or different organs of interest, have distinct spectral signatures, which can be used to identify whether or not a test image is similar to the images used for training. We show that our proposed method is more accurate than several competing methods, including methods based on prediction uncertainty and image classification.

3D pyramidal densely connected network with cross-frame uncertainty guidance for intravascular ultrasound sequence segmentation

Objective. Automatic extraction of external elastic membrane border (EEM) and lumen-intima border (LIB) in intravascular ultrasound (IVUS) sequences aids atherosclerosis diagnosis. Existing IVUS segmentation networks ignored longitudinal relations among sequential images and neglected that IVUS images of different vascular conditions vary largely in intricacy and informativeness. As a result, they suffered from performance degradation in complicated parts in IVUS sequences.Approach. In this paper, we develop a 3D Pyramidal Densely-connected Network (PDN) with Adaptive learning and post-Correction guided by a novel cross-frame uncertainty (CFU). The proposed method is named PDN-AC. Specifically, the PDN enables the longitudinal information exploitation and the effective perception of size-varied vessel regions in IVUS samples, by pyramidally connecting multi-scale 3D dilated convolutions. Additionally, the CFU enhances the robustness of the method to complicated pathology from the frame-level (f-CFU) and pixel-level (p-CFU) via exploiting cross-frame knowledge in IVUS sequences. The f-CFU weighs the complexity of IVUS frames and steers an adaptive sampling during the PDN training. The p-CFU visualizes uncertain pixels probably misclassified by the PDN and guides an active contour-based post-correction.Main results. Human and animal experiments were conducted on IVUS datasets acquired from atherosclerosis patients and pigs. Results showed that the f-CFU weighted adaptive sampling reduced the Hausdorff distance (HD) by 10.53%/7.69% in EEM/LIB detection. Improvements achieved by the p-CFU guided post-correction were 2.94%/5.56%.Significance. The PDN-AC attained mean Jaccard values of 0.90/0.87 and HD values of 0.33/0.34 mm in EEM/LIB detection, preferable to state-of-the-art IVUS segmentation methods.

Improving the Segmentation Accuracy of Ovarian-Tumor Ultrasound Images Using Image Inpainting

Diagnostic results can be radically influenced by the quality of 2D ovarian-tumor ultrasound images. However, clinically processed 2D ovarian-tumor ultrasound images contain many artificially recognized symbols, such as fingers, crosses, dashed lines, and letters which assist artificial intelligence (AI) in image recognition. These symbols are widely distributed within the lesion's boundary, which can also affect the useful feature-extraction-utilizing networks and thus decrease the accuracy of lesion classification and segmentation. Image inpainting techniques are used for noise and object elimination from images. To solve this problem, we observed the MMOTU dataset and built a 2D ovarian-tumor ultrasound image inpainting dataset by finely annotating the various symbols in the images. A novel framework called mask-guided generative adversarial network (MGGAN) is presented in this paper for 2D ovarian-tumor ultrasound images to remove various symbols from the images. The MGGAN performs to a high standard in corrupted regions by using an attention mechanism in the generator to pay more attention to valid information and ignore symbol information, making lesion boundaries more realistic. Moreover, fast Fourier convolutions (FFCs) and residual networks are used to increase the global field of perception; thus, our model can be applied to high-resolution ultrasound images. The greatest benefit of this algorithm is that it achieves pixel-level inpainting of distorted regions without clean images. Compared with other models, our model achieveed better results with only one stage in terms of objective and subjective evaluations. Our model obtained the best results for 256 × 256 and 512 × 512 resolutions. At a resolution of 256 × 256, our model achieved 0.9246 for SSIM, 22.66 for FID, and 0.07806 for LPIPS. At a resolution of 512 × 512, our model achieved 0.9208 for SSIM, 25.52 for FID, and 0.08300 for LPIPS. Our method can considerably improve the accuracy of computerized ovarian tumor diagnosis. The segmentation accuracy was improved from 71.51% to 76.06% for the Unet model and from 61.13% to 66.65% for the PSPnet model in clean images.

Predictive uncertainty estimation for out-of-distribution detection in digital pathology

Machine learning model deployment in clinical practice demands real-time risk assessment to identify situations in which the model is uncertain. Once deployed, models should be accurate for classes seen during training while providing informative estimates of uncertainty to flag abnormalities and unseen classes for further analysis. Although recent developments in uncertainty estimation have resulted in an increasing number of methods, a rigorous empirical evaluation of their performance on large-scale digital pathology datasets is lacking. This work provides a benchmark for evaluating prevalent methods on multiple datasets by comparing the uncertainty estimates on both in-distribution and realistic near and far out-of-distribution (OOD) data on a whole-slide level. To this end, we aggregate uncertainty values from patch-based classifiers to whole-slide level uncertainty scores. We show that results found in classical computer vision benchmarks do not always translate to the medical imaging setting. Specifically, we demonstrate that deep ensembles perform best at detecting far-OOD data but can be outperformed on a more challenging near-OOD detection task by multi-head ensembles trained for optimal ensemble diversity. Furthermore, we demonstrate the harmful impact OOD data can have on the performance of deployed machine learning models. Overall, we show that uncertainty estimates can be used to discriminate in-distribution from OOD data with high AUC scores. Still, model deployment might require careful tuning based on prior knowledge of prospective OOD data.

Domain generalization for prostate segmentation in transrectal ultrasound images: A multi-center study

Prostate biopsy and image-guided treatment procedures are often performed under the guidance of ultrasound fused with magnetic resonance images (MRI). Accurate image fusion relies on accurate segmentation of the prostate on ultrasound images. Yet, the reduced signal-to-noise ratio and artifacts (e.g., speckle and shadowing) in ultrasound images limit the performance of automated prostate segmentation techniques and generalizing these methods to new image domains is inherently difficult. In this study, we address these challenges by introducing a novel 2.5D deep neural network for prostate segmentation on ultrasound images. Our approach addresses the limitations of transfer learning and finetuning methods (i.e., drop in performance on the original training data when the model weights are updated) by combining a supervised domain adaptation technique and a knowledge distillation loss. The knowledge distillation loss allows the preservation of previously learned knowledge and reduces the performance drop after model finetuning on new datasets. Furthermore, our approach relies on an attention module that considers model feature positioning information to improve the segmentation accuracy. We trained our model on 764 subjects from one institution and finetuned our model using only ten subjects from subsequent institutions. We analyzed the performance of our method on three large datasets encompassing 2067 subjects from three different institutions. Our method achieved an average Dice Similarity Coefficient (Dice) of 94.0±0.03 and Hausdorff Distance (HD95) of 2.28 mm in an independent set of subjects from the first institution. Moreover, our model generalized well in the studies from the other two institutions (Dice: 91.0±0.03; HD95: 3.7 mm and Dice: 82.0±0.03; HD95: 7.1 mm). We introduced an approach that successfully segmented the prostate on ultrasound images in a multi-center study, suggesting its clinical potential to facilitate the accurate fusion of ultrasound and MRI images to drive biopsy and image-guided treatments.

Cardiac MRI segmentation with sparse annotations: Ensembling deep learning uncertainty and shape priors

The performance of deep learning for cardiac magnetic resonance imaging (MRI) segmentation is oftentimes degraded when using small datasets and sparse annotations for training or adapting a pre-trained model to previously unseen datasets. Here, we developed and evaluated an approach to addressing some of these issues to facilitate broader use of deep learning for short-axis cardiac MRI segmentation. We developed a globally optimal label fusion (GOLF) algorithm that enforced spatial smoothness to generate consensus segmentation from segmentation predictions provided by a deep learning ensemble algorithm. The GOLF consensus was entered into an uncertainty-guided coupled continuous kernel cut (ugCCKC) algorithm that employed normalized cut, image-grid continuous regularization, and "nesting" and circular shape priors of the left ventricular myocardium (LVM) and cavity (LVC). In addition, the uncertainty measurements derived from the segmentation predictions were used to constrain the similarity of GOLF and final segmentation. We optimized ugCCKC through upper bound relaxation, for which we developed an efficient coupled continuous max-flow algorithm implemented in an iterative manner. We showed that GOLF yielded average symmetric surface distance (ASSD) 0.2-0.8 mm lower than an averaging method with higher or similar Dice similarity coefficient (DSC). We also demonstrated that ugCCKC incorporating the shape priors improved DSC by 0.01-0.05 and reduced ASSD by 0.1-0.9 mm. In addition, we integrated GOLF and ugCCKC into a deep learning ensemble algorithm by refining the segmentation of an unannotated dataset and using the refined segmentation to update the trained models. With the proposed framework, we demonstrated the capability of using relatively small datasets (5-10 subjects) with sparse (5-25% slices labeled) annotations to train a deep learning algorithm, while achieving DSC of 0.871-0.893 for LVM and 0.933-0.959 for LVC on the LVQuan dataset, and these were 0.844-0.871 for LVM and 0.923-0.931 for LVC on the ACDC dataset. Furthermore, we showed that the proposed approach can be adapted to substantially alleviate the domain shift issue. Moreover, we calculated a number of commonly used LV function measurements using the derived segmentation and observed strong correlations (Pearson r=0.77-1.00, p<0.001) between algorithm and manual LV function analyses. These results suggest that the developed approaches can be used to facilitate broader application of deep learning in research and clinical cardiac MR imaging workflow.

AI-Based Automatic Detection and Classification of Diabetic Retinopathy Using U-Net and Deep Learning

Artificial intelligence is widely applied to automate Diabetic retinopathy diagnosis. Diabetes-related retinal vascular disease is one of the world’s most common leading causes of blindness and vision impairment. Therefore, automated DR detection systems would greatly benefit the early screening and treatment of DR and prevent vision loss caused by it. Researchers have proposed several systems to detect abnormalities in retinal images in the past few years. However, Diabetic Retinopathy automatic detection methods have traditionally been based on hand-crafted feature extraction from the retinal images and using a classifier to obtain the final classification. DNN (Deep neural networks) have made several changes in the previous few years to assist overcome the problem mentioned above. We suggested a two-stage novel approach for automated DR classification in this research. Due to the low fraction of positive instances in the asymmetric Optic Disk (OD) and blood vessels (BV) detection system, preprocessing and data augmentation techniques are used to enhance the image quality and quantity. The first step uses two independent U-Net models for OD (optic disc) and BV (blood vessel) segmentation. In the second stage, the symmetric hybrid CNN-SVD model was created after preprocessing to extract and choose the most discriminant features following OD and BV extraction using Inception-V3 based on transfer learning, and detects DR by recognizing retinal biomarkers such as MA (microaneurysms), HM (hemorrhages), and exudates (EX). On EyePACS-1, Messidor-2, and DIARETDB0, the proposed methodology demonstrated state-of-the-art performance, with an average accuracy of 97.92%, 94.59%, and 93.52%, respectively. Extensive testing and comparisons with baseline approaches indicate the efficacy of the suggested methodology.

FECC-Net: A Novel Feature Enhancement and Context Capture Network Based on Brain MRI Images for Lesion Segmentation

In recent years, the increasing incidence of morbidity of brain stroke has made fast and accurate segmentation of lesion areas from brain MRI images important. With the development of deep learning, segmentation methods based on the computer have become a solution to assist clinicians in early diagnosis and treatment planning. Nevertheless, the variety of lesion sizes in brain MRI images and the roughness of the boundary of the lesion pose challenges to the accuracy of the segmentation algorithm. Current mainstream medical segmentation models are not able to solve these challenges due to their insufficient use of image features and context information. This paper proposes a novel feature enhancement and context capture network (FECC-Net), which is mainly composed of an atrous spatial pyramid pooling (ASPP) module and an enhanced encoder. In particular, the ASPP model uses parallel convolution operations with different sampling rates to enrich multi-scale features and fully capture image context information in order to process lesions of different sizes. The enhanced encoder obtains deep semantic features and shallow boundary features in the feature extraction process to achieve image feature enhancement, which is helpful for restoration of the lesion boundaries. We divide the pathological image into three levels according to the number of pixels in the real mask area and evaluate FECC-Net on an open dataset called Anatomical Tracings of Lesions After Stroke (ATLAS). The experimental results show that our FECC-Net outperforms mainstream methods, such as DoubleU-Net and TransUNet. Especially in small target tasks, FECC-Net is 4.09% ahead of DoubleU-Net on the main indicator DSC. Therefore, FECC-Net is encouraging and can be relied upon for brain MRI image applications.

Semi-Automatic Prostate Segmentation From Ultrasound Images Using Machine Learning and Principal Curve Based on Interpretable Mathematical Model Expression

Accurate prostate segmentation in transrectal ultrasound (TRUS) is a challenging problem due to the low contrast of TRUS images and the presence of imaging artifacts such as speckle and shadow regions. To address this issue, we propose a semi-automatic model termed Hybrid Segmentation Model (H-SegMod) for prostate Region of Interest (ROI) segmentation in TRUS images. H-SegMod contains two cascaded stages. The first stage is to obtain the vertices sequences based on an improved principal curve-based model, where a few radiologist-selected seed points are used as prior. The second stage is to find a map function for describing the smooth prostate contour based on an improved machine learning model. Experimental results show that our proposed model achieved superior segmentation results compared with several other state-of-the-art models, achieving an average Dice Similarity Coefficient (DSC), Jaccard Similarity Coefficient (Ω), and Accuracy (ACC) of 96.5%, 95.2%, and 96.3%, respectively.

Recent Advances in Machine Learning Applied to Ultrasound Imaging

Machine learning (ML) methods are pervading an increasing number of fields of application because of their capacity to effectively solve a wide variety of challenging problems. The employment of ML techniques in ultrasound imaging applications started several years ago but the scientific interest in this issue has increased exponentially in the last few years. The present work reviews the most recent (2019 onwards) implementations of machine learning techniques for two of the most popular ultrasound imaging fields, medical diagnostics and non-destructive evaluation. The former, which covers the major part of the review, was analyzed by classifying studies according to the human organ investigated and the methodology (e.g., detection, segmentation, and/or classification) adopted, while for the latter, some solutions to the detection/classification of material defects or particular patterns are reported. Finally, the main merits of machine learning that emerged from the study analysis are summarized and discussed.

Deep-learning segmentation of ultrasound images for automated calculation of the hydronephrosis area to renal parenchyma ratio

Purpose

We investigated the feasibility of measuring the hydronephrosis area to renal parenchyma (HARP) ratio from ultrasound images using a deep-learning network.

Materials and Methods

The coronal renal ultrasound images of 195 pediatric and adolescent patients who underwent pyeloplasty to repair ureteropelvic junction obstruction were retrospectively reviewed. After excluding cases without a representative longitudinal renal image, we used a dataset of 168 images for deep-learning segmentation. Ten novel networks, such as combinations of DeepLabV3+ and UNet++, were assessed for their ability to calculate hydronephrosis and kidney areas, and the ensemble method was applied for further improvement. By dividing the image set into four, cross-validation was conducted, and the segmentation performance of the deep-learning network was evaluated using sensitivity, specificity, and dice similarity coefficients by comparison with the manually traced area.

Results

All 10 networks and ensemble methods showed good visual correlation with the manually traced kidney and hydronephrosis areas. The dice similarity coefficient of the 10-model ensemble was 0.9108 on average, and the best 5-model ensemble had a dice similarity coefficient of 0.9113 on average. We included patients with severe hydronephrosis who underwent renal ultrasonography at a single institution; thus, external validation of our algorithm in a heterogeneous ultrasonography examination setup with a diverse set of instruments is recommended.

Conclusions

Deep-learning-based calculation of the HARP ratio is feasible and showed high accuracy for imaging of the severity of hydronephrosis using ultrasonography. This algorithm can help physicians make more accurate and reproducible diagnoses of hydronephrosis using ultrasonography.

Robust Scatterer Number Density Segmentation of Ultrasound Images

Quantitative ultrasound (QUS) aims to reveal information about the tissue microstructure using backscattered echo signals from clinical scanners. Among different QUS parameters, scatterer number density is an important property that can affect the estimation of other QUS parameters. Scatterer number density can be classified into high or low scatterer densities. If there are more than ten scatterers inside the resolution cell, the envelope data are considered as fully developed speckle (FDS) and, otherwise, as underdeveloped speckle (UDS). In conventional methods, the envelope data are divided into small overlapping windows (a strategy here we refer to as patching), and statistical parameters, such as SNR and skewness, are employed to classify each patch of envelope data. However, these parameters are system-dependent, meaning that their distribution can change by the imaging settings and patch size. Therefore, reference phantoms that have known scatterer number density are imaged with the same imaging settings to mitigate system dependency. In this article, we aim to segment regions of ultrasound data without any patching. A large dataset is generated, which has different shapes of scatterer number density and mean scatterer amplitude using a fast simulation method. We employ a convolutional neural network (CNN) for the segmentation task and investigate the effect of domain shift when the network is tested on different datasets with different imaging settings. Nakagami parametric image is employed for multitask learning to improve performance. Furthermore, inspired by the reference phantom methods in QUS, a domain adaptation stage is proposed, which requires only two frames of data from FDS and UDS classes. We evaluate our method for different experimental phantoms and in vivo data.

Effect of dataset size, image quality, and image type on deep learning-based automatic prostate segmentation in 3D ultrasound

Three-dimensional (3D) transrectal ultrasound (TRUS) is utilized in prostate cancer diagnosis and treatment, necessitating time-consuming manual prostate segmentation. We have previously developed an automatic 3D prostate segmentation algorithm involving deep learning prediction on radially sampled 2D images followed by 3D reconstruction, trained on a large, clinically diverse dataset with variable image quality. As large clinical datasets are rare, widespread adoption of automatic segmentation could be facilitated with efficient 2D-based approaches and the development of an image quality grading method. The complete training dataset of 6761 2D images, resliced from 206 3D TRUS volumes acquired using end-fire and side-fire acquisition methods, was split to train two separate networks using either end-fire or side-fire images. Split datasets were reduced to 1000, 500, 250, and 100 2D images. For deep learning prediction, modified U-Net and U-Net++ architectures were implemented and compared using an unseen test dataset of 40 3D TRUS volumes. A 3D TRUS image quality grading scale with three factors (acquisition quality, artifact severity, and boundary visibility) was developed to assess the impact on segmentation performance. For the complete training dataset, U-Net and U-Net++ networks demonstrated equivalent performance, but when trained using split end-fire/side-fire datasets, U-Net++ significantly outperformed the U-Net. Compared to the complete training datasets, U-Net++ trained using reduced-size end-fire and side-fire datasets demonstrated equivalent performance down to 500 training images. For this dataset, image quality had no impact on segmentation performance for end-fire images but did have a significant effect for side-fire images, with boundary visibility having the largest impact. Our algorithm provided fast (<1.5 s) and accurate 3D segmentations across clinically diverse images, demonstrating generalizability and efficiency when employed on smaller datasets, supporting the potential for widespread use, even when data is scarce. The development of an image quality grading scale provides a quantitative tool for assessing segmentation performance.

Medical Image Segmentation Using Transformer Networks

Deep learning models represent the state of the art in medical image segmentation. Most of these models are fully-convolutional networks (FCNs), namely each layer processes the output of the preceding layer with convolution operations. The convolution operation enjoys several important properties such as sparse interactions, parameter sharing, and translation equivariance. Because of these properties, FCNs possess a strong and useful inductive bias for image modeling and analysis. However, they also have certain important shortcomings, such as performing a fixed and pre-determined operation on a test image regardless of its content and difficulty in modeling long-range interactions. In this work we show that a different deep neural network architecture, based entirely on self-attention between neighboring image patches and without any convolution operations, can achieve more accurate segmentations than FCNs. Our proposed model is based directly on the transformer network architecture. Given a 3D image block, our network divides it into non-overlapping 3D patches and computes a 1D embedding for each patch. The network predicts the segmentation map for the block based on the self-attention between these patch embeddings. Furthermore, in order to address the common problem of scarcity of labeled medical images, we propose methods for pre-training this model on large corpora of unlabeled images. Our experiments show that the proposed model can achieve segmentation accuracies that are better than several state of the art FCN architectures on two datasets. Our proposed network can be trained using only tens of labeled images. Moreover, with the proposed pre-training strategies, our network outperforms FCNs when labeled training data is small.

The Challenges Facing Deep Learning based Catheter Localization for Ultrasound Guided High-Dose-Rate Prostate Brachytherapy

BACKGROUND

Automated catheter localization for ultrasound guided high-dose-rate prostate brachytherapy faces challenges relating to imaging noise and artifacts. To date, catheter reconstruction during the clinical procedure is performed manually. Deep learning has been successfully applied to a wide variety of complex tasks and has the potential to tackle the unique challenges associated with multiple catheter localization on ultrasound. Such a task is well suited for automation, with the potential to improve productivity and reliability.

PURPOSE

We developed a deep learning model for automated catheter reconstruction and investigated potential factors influencing model performance. The model was designed to integrate into a clinical workflow, with a proposed reconstruction confidence metric to aid in planner verification.

METHODS

Datasets from 242 patients treated from 2016 to 2020 were collected retrospectively. The anonymized dataset comprises of 31,000 transverse images reconstructed from 3D sagittal ultrasound acquisitions and 3,500 implanted catheters manually localized by the planner. Each catheter was retrospectively ranked based on the severity of imaging artifacts affecting reconstruction difficulty. The U-NET deep learning architecture was trained to localize implanted catheters on transverse images. A five-fold cross-validation method was used, allowing for evaluation over the entire dataset. The post-processing software combined the predictions with patient-specific implant information to reconstructed catheters in 3D space, uniquely matched to the implanted grid positions. A reconstruction confidence metric was calculated based on the number and probability of localized predictions per catheter. For each patient, deep learning prediction and post-processing reconstruction was completed in under two minutes on a non-performance PC.

RESULTS

Overall, 80% of catheter reconstructions were accurate, within 2 mm along 90% of the length. The catheter tip was often not detected and required extrapolation during reconstruction. The reconstruction accuracy was 89% for the easiest catheter ranking and decreased to 13% for the highest difficulty ranking, when the aid of live ultrasound would have been recommended. Even when limited to the easiest ranked catheters, the reconstruction accuracy decreased at distal grid positions, down to 50%. Individual implantation style was found to influence the frequency of severe artifacts, slightly impacting the model accuracy. A reconstruction confidence metric identified the difficult catheters, removed the observed individual variation, and increased the overall accuracy to 91% while excluding 27% of the reconstructions.

CONCLUSIONS

The deep learning model localized implanted catheters over a large clinical dataset, with overall promising results. The model faced challenges due to ultrasound artifacts and image degradation distal to the probe, underlining the continued importance of maintaining image quality and minimizing artifacts. A potential workflow for integration into the clinical procedure was demonstrated, including the use of a confidence metric to predict low accuracy reconstructions. Comparison between models evaluated on different datasets should also consider underlying differences, such as the frequency and severity of imaging artifacts. This article is protected by copyright. All rights reserved.

Estimation of the Prostate Volume from Abdominal Ultrasound Images by Image-Patch Voting

Estimation of the prostate volume with ultrasound offers many advantages such as portability, low cost, harmlessness, and suitability for real-time operation. Abdominal Ultrasound (AUS) is a practical procedure that deserves more attention in automated prostate-volume-estimation studies. As the experts usually consider automatic end-to-end volume-estimation procedures as non-transparent and uninterpretable systems, we proposed an expert-in-the-loop automatic system that follows the classical prostate-volume-estimation procedures. Our system directly estimates the diameter parameters of the standard ellipsoid formula to produce the prostate volume. To obtain the diameters, our system detects four diameter endpoints from the transverse and two diameter endpoints from the sagittal AUS images as defined by the classical procedure. These endpoints are estimated using a new image-patch voting method to address characteristic problems of AUS images. We formed a novel prostate AUS data set from 305 patients with both transverse and sagittal planes. The data set includes MRI images for 75 of these patients. At least one expert manually marked all the data. Extensive experiments performed on this data set showed that the proposed system results ranged among experts’ volume estimations, and our system can be used in clinical practice.

A review of artificial intelligence in prostate cancer detection on imaging

A multitude of studies have explored the role of artificial intelligence (AI) in providing diagnostic support to radiologists, pathologists, and urologists in prostate cancer detection, risk-stratification, and management. This review provides a comprehensive overview of relevant literature regarding the use of AI models in (1) detecting prostate cancer on radiology images (magnetic resonance and ultrasound imaging), (2) detecting prostate cancer on histopathology images of prostate biopsy tissue, and (3) assisting in supporting tasks for prostate cancer detection (prostate gland segmentation, MRI-histopathology registration, MRI-ultrasound registration). We discuss both the potential of these AI models to assist in the clinical workflow of prostate cancer diagnosis, as well as the current limitations including variability in training data sets, algorithms, and evaluation criteria. We also discuss ongoing challenges and what is needed to bridge the gap between academic research on AI for prostate cancer and commercial solutions that improve routine clinical care.

Application of deep learning to auto-delineation of target volumes and organs at risk in radiotherapy

The technological advancement heralded the arrival of precision radiotherapy (RT), thereby increasing the therapeutic ratio and decreasing the side effects from treatment. Contour of target volumes (TV) and organs at risk (OARs) in RT is a complicated process. In recent years, automatic contouring of TV and OARs has rapidly developed due to the advances in deep learning (DL). This technology has the potential to save time and to reduce intra- or inter-observer variability. In this paper, the authors provide an overview of RT, introduce the concept of DL, summarize the data characteristics of the included literature, summarize the possible challenges for DL in the future, and discuss the possible research directions.

Automated Recognition of Ultrasound Cardiac Views Based on Deep Learning with Graph Constraint

In transthoracic echocardiographic (TTE) examination, it is essential to identify the cardiac views accurately. Computer-aided recognition is expected to improve the accuracy of cardiac views of the TTE examination, particularly when obtained by non-trained providers. A new method for automatic recognition of cardiac views is proposed consisting of three processes. First, a spatial transform network is performed to learn cardiac shape changes during a cardiac cycle, which reduces intra-class variability. Second, a channel attention mechanism is introduced to adaptively recalibrate channel-wise feature responses. Finally, the structured signals by the similarities among cardiac views are transformed into the graph-based image embedding, which acts as unsupervised regularization constraints to improve the generalization accuracy. The proposed method is trained and tested in 171792 cardiac images from 584 subjects. The overall accuracy of the proposed method on cardiac image classification is 99.10%, and the mean AUC is 99.36%, better than known methods. Moreover, the overall accuracy is 97.73%, and the mean AUC is 98.59% on an independent test set with 37,883 images from 100 subjects. The proposed automated recognition model achieved comparable accuracy with true cardiac views, and thus can be applied clinically to help find standard cardiac views.

Transfer learning in medical image segmentation: New insights from analysis of the dynamics of model parameters and learned representations

We present a critical assessment of the role of transfer learning in training fully convolutional networks (FCNs) for medical image segmentation. We first show that although transfer learning reduces the training time on the target task, improvements in segmentation accuracy are highly task/data-dependent. Large improvements are observed only when the segmentation task is more challenging and the target training data is smaller. We shed light on these observations by investigating the impact of transfer learning on the evolution of model parameters and learned representations. We observe that convolutional filters change little during training and still look random at convergence. We further show that quite accurate FCNs can be built by freezing the encoder section of the network at random values and only training the decoder section. At least for medical image segmentation, this finding challenges the common belief that the encoder section needs to learn data/task-specific representations. We examine the evolution of FCN representations to gain a deeper insight into the effects of transfer learning on the training dynamics. Our analysis shows that although FCNs trained via transfer learning learn different representations than FCNs trained with random initialization, the variability among FCNs trained via transfer learning can be as high as that among FCNs trained with random initialization. Moreover, feature reuse is not restricted to the early encoder layers; rather, it can be more significant in deeper layers. These findings offer new insights and suggest alternative ways of training FCNs for medical image segmentation.

Artificial intelligence in radiation oncology

Artificial intelligence (AI) is a computer science that tries to mimic human-like intelligence in machines that use computer software and algorithms to perform specific tasks without direct human input. Machine learning (ML) is a subunit of AI that uses data-driven algorithms that learn to imitate human behavior based on a previous example or experience. Deep learning is an ML technique that uses deep neural networks to create a model. The growth and sharing of data, increasing computing power, and developments in AI have initiated a transformation in healthcare. Advances in radiation oncology have produced a significant amount of data that must be integrated with computed tomography imaging, dosimetry, and imaging performed before each fraction. Of the many algorithms used in radiation oncology, has advantages and limitations with different computational power requirements. The aim of this review is to summarize the radiotherapy (RT) process in workflow order by identifying specific areas in which quality and efficiency can be improved by ML. The RT stage is divided into seven stages: patient evaluation, simulation, contouring, planning, quality control, treatment application, and patient follow-up. A systematic evaluation of the applicability, limitations, and advantages of AI algorithms has been done for each stage.

Artificial intelligence in ultrasound

Ultrasound (US), a flexible green imaging modality, is expanding globally as a first-line imaging technique in various clinical fields following with the continual emergence of advanced ultrasonic technologies and the well-established US-based digital health system. Actually, in US practice, qualified physicians should manually collect and visually evaluate images for the detection, identification and monitoring of diseases. The diagnostic performance is inevitably reduced due to the intrinsic property of high operator-dependence from US. In contrast, artificial intelligence (AI) excels at automatically recognizing complex patterns and providing quantitative assessment for imaging data, showing high potential to assist physicians in acquiring more accurate and reproducible results. In this article, we will provide a general understanding of AI, machine learning (ML) and deep learning (DL) technologies; We then review the rapidly growing applications of AI-especially DL technology in the field of US-based on the following anatomical regions: thyroid, breast, abdomen and pelvis, obstetrics heart and blood vessels, musculoskeletal system and other organs by covering image quality control, anatomy localization, object detection, lesion segmentation, and computer-aided diagnosis and prognosis evaluation; Finally, we offer our perspective on the challenges and opportunities for the clinical practice of biomedical AI systems in US.

Using Spatial Probability Maps to Highlight Potential Inaccuracies in Deep Learning-Based Contours: Facilitating Online Adaptive Radiation Therapy

PURPOSE

Contouring organs at risk remains a largely manual task, which is time consuming and prone to variation. Deep learning-based delineation (DLD) shows promise both in terms of quality and speed, but it does not yet perform perfectly. Because of that, manual checking of DLD is still recommended. There are currently no commercial tools to focus attention on the areas of greatest uncertainty within a DLD contour. Therefore, we explore the use of spatial probability maps (SPMs) to help efficiency and reproducibility of DLD checking and correction, using the salivary glands as the paradigm.

METHODS AND MATERIALS

A 3-dimensional fully convolutional network was trained with 315/264 parotid/submandibular glands. Subsequently, SPMs were created using Monte Carlo dropout (MCD). The method was boosted by placing a Gaussian distribution (GD) over the model's parameters during sampling (MCD + GD). MCD and MCD + GD were quantitatively compared and the SPMs were visually inspected.

RESULTS

The addition of the GD appears to increase the method's ability to detect uncertainty. In general, this technique demonstrated uncertainty in areas that (1) have lower contrast, (2) are less consistently contoured by clinicians, and (3) deviate from the anatomic norm.

CONCLUSIONS

We believe the integration of uncertainty information into contours made using DLD is an important step in highlighting where a contour may be less reliable. We have shown how SPMs are one way to achieve this and how they may be integrated into the online adaptive radiation therapy workflow.

A fast technique for hyper-echoic region separation from brain ultrasound images using patch based thresholding and cubic B-spline based contour smoothing

Ultrasound image guided brain surgery (UGBS) requires an automatic and fast image segmentation method. The level-set and active contour based algorithms have been found to be useful for obtaining topology-independent boundaries between different image regions. But slow convergence limits their use in online US image segmentation. The performance of these algorithms deteriorates on US images because of the intensity inhomogeneity. This paper proposes an effective region-driven method for the segmentation of hyper-echoic (HE) regions suppressing the hypo-echoic and anechoic regions in brain US images. An automatic threshold estimation scheme is developed with a modified Niblack's approach. The separation of the hyper-echoic and non-hyper-echoic (NHE) regions is performed by successively applying patch based intensity thresholding and boundary smoothing. First, a patch based segmentation is performed, which separates roughly the two regions. The patch based approach in this process reduces the effect of intensity heterogeneity within an HE region. An iterative boundary correction step with reducing patch size improves further the regional topology and refines the boundary regions. For avoiding the slope and curvature discontinuities and obtaining distinct boundaries between HE and NHE regions, a cubic B-spline model of curve smoothing is applied. The proposed method is 50-100 times faster than the other level-set based image segmentation algorithms. The segmentation performance and the convergence speed of the proposed method are compared with four other competing level-set based algorithms. The computational results show that the proposed segmentation approach outperforms other level-set based techniques both subjectively and objectively.

Artificial Intelligence and Machine Learning in Prostate Cancer Patient Management-Current Trends and Future Perspectives

Artificial intelligence (AI) is the field of computer science that aims to build smart devices performing tasks that currently require human intelligence. Through machine learning (ML), the deep learning (DL) model is teaching computers to learn by example, something that human beings are doing naturally. AI is revolutionizing healthcare. Digital pathology is becoming highly assisted by AI to help researchers in analyzing larger data sets and providing faster and more accurate diagnoses of prostate cancer lesions. When applied to diagnostic imaging, AI has shown excellent accuracy in the detection of prostate lesions as well as in the prediction of patient outcomes in terms of survival and treatment response. The enormous quantity of data coming from the prostate tumor genome requires fast, reliable and accurate computing power provided by machine learning algorithms. Radiotherapy is an essential part of the treatment of prostate cancer and it is often difficult to predict its toxicity for the patients. Artificial intelligence could have a future potential role in predicting how a patient will react to the therapy side effects. These technologies could provide doctors with better insights on how to plan radiotherapy treatment. The extension of the capabilities of surgical robots for more autonomous tasks will allow them to use information from the surgical field, recognize issues and implement the proper actions without the need for human intervention.

Deep learning-based digitization of prostate brachytherapy needles in ultrasound images

PURPOSE

To develop, and evaluate the performance of, a deep learning-based three-dimensional (3D) convolutional neural network (CNN) artificial intelligence (AI) algorithm aimed at finding needles in ultrasound images used in prostate brachytherapy.

METHODS

Transrectal ultrasound (TRUS) image volumes from 1102 treatments were used to create a clinical ground truth (CGT) including 24422 individual needles that had been manually digitized by medical physicists during brachytherapy procedures. A 3D CNN U-net with 128 × 128 × 128 TRUS image volumes as input was trained using 17215 needle examples. Predictions of voxels constituting a needle were combined to yield a 3D linear function describing the localization of each needle in a TRUS volume. Manual and AI digitizations were compared in terms of the root-mean-square distance (RMSD) along each needle, expressed as median and interquartile range (IQR). The method was evaluated on a data set including 7207 needle examples. A subgroup of the evaluation data set (n = 188) was created, where the needles were digitized once more by a medical physicist (G1) trained in brachytherapy. The digitization procedure was timed.

RESULTS

The RMSD between the AI and CGT was 0.55 (IQR: 0.35-0.86) mm. In the smaller subset, the RMSD between AI and CGT was similar (0.52 [IQR: 0.33-0.79] mm) but significantly smaller (P < 0.001) than the difference of 0.75 (IQR: 0.49-1.20) mm between AI and G1. The difference between CGT and G1 was 0.80 (IQR: 0.48-1.18) mm, implying that the AI performed as well as the CGT in relation to G1. The mean time needed for human digitization was 10 min 11 sec, while the time needed for the AI was negligible.

CONCLUSIONS

A 3D CNN can be trained to identify needles in TRUS images. The performance of the network was similar to that of a medical physicist trained in brachytherapy. Incorporating a CNN for needle identification can shorten brachytherapy treatment procedures substantially.

Confidence Calibration and Predictive Uncertainty Estimation for Deep Medical Image Segmentation

Fully convolutional neural networks (FCNs), and in particular U-Nets, have achieved state-of-the-art results in semantic segmentation for numerous medical imaging applications. Moreover, batch normalization and Dice loss have been used successfully to stabilize and accelerate training. However, these networks are poorly calibrated i.e. they tend to produce overconfident predictions for both correct and erroneous classifications, making them unreliable and hard to interpret. In this paper, we study predictive uncertainty estimation in FCNs for medical image segmentation. We make the following contributions: 1) We systematically compare cross-entropy loss with Dice loss in terms of segmentation quality and uncertainty estimation of FCNs; 2) We propose model ensembling for confidence calibration of the FCNs trained with batch normalization and Dice loss; 3) We assess the ability of calibrated FCNs to predict segmentation quality of structures and detect out-of-distribution test examples. We conduct extensive experiments across three medical image segmentation applications of the brain, the heart, and the prostate to evaluate our contributions. The results of this study offer considerable insight into the predictive uncertainty estimation and out-of-distribution detection in medical image segmentation and provide practical recipes for confidence calibration. Moreover, we consistently demonstrate that model ensembling improves confidence calibration.

Segmenting the Semi-Conductive Shielding Layer of Cable Slice Images Using the Convolutional Neural Network

Being an important part of aerial insulated cable, the semiconductive shielding layer is made of a typical polymer material and can improve the cable transmission effects; the structural parameters will affect the cable quality directly. Then, the image processing of the semiconductive layer plays an essential role in the structural parameter measurements. However, the semiconductive layer images are often disturbed by the cutting marks, which affect the measurements seriously. In this paper, a novel method based on the convolutional neural network is proposed for image segmentation. In our proposed strategy, a deep fully convolutional network with a skip connection algorithm is defined as the main framework. The inception structure and residual connection are employed to fuse features extracted from the receptive fields with different sizes. Finally, an improved weighted loss function and refined algorithm are utilized for pixel classification. Experimental results show that our proposed algorithm achieves better performance than the current algorithms.

Current status of deep learning applications in abdominal ultrasonography

Deep learning is one of the most popular artificial intelligence techniques used in the medical field. Although it is at an early stage compared to deep learning analyses of computed tomography or magnetic resonance imaging, studies applying deep learning to ultrasound imaging have been actively conducted. This review analyzes recent studies that applied deep learning to ultrasound imaging of various abdominal organs and explains the challenges encountered in these applications.

A deep learning method for real-time intraoperative US image segmentation in prostate brachytherapy

PURPOSE

This paper addresses the detection of the clinical target volume (CTV) in transrectal ultrasound (TRUS) image-guided intraoperative for permanent prostate brachytherapy. Developing a robust and automatic method to detect the CTV on intraoperative TRUS images is clinically important to have faster and reproducible interventions that can benefit both the clinical workflow and patient health.

METHODS

We present a multi-task deep learning method for an automatic prostate CTV boundary detection in intraoperative TRUS images by leveraging both the low-level and high-level (prior shape) information. Our method includes a channel-wise feature calibration strategy for low-level feature extraction and learning-based prior knowledge modeling for prostate CTV shape reconstruction. It employs CTV shape reconstruction from automatically sampled boundary surface coordinates (pseudo-landmarks) to detect the low-contrast and noisy regions across the prostate boundary, while being less biased from shadowing, inherent speckles, and artifact signals from the needle and implanted radioactive seeds.

RESULTS

The proposed method was evaluated on a clinical database of 145 patients who underwent permanent prostate brachytherapy under TRUS guidance. Our method achieved a mean accuracy of [Formula: see text] and a mean surface distance error of [Formula: see text]. Extensive ablation and comparison studies show that our method outperformed previous deep learning-based methods by more than 7% for the Dice similarity coefficient and 6.9 mm reduced 3D Hausdorff distance error.

CONCLUSION

Our study demonstrates the potential of shape model-based deep learning methods for an efficient and accurate CTV segmentation in an ultrasound-guided intervention. Moreover, learning both low-level features and prior shape knowledge with channel-wise feature calibration can significantly improve the performance of deep learning methods in medical image segmentation.