Automated coronary artery atherosclerosis detection and weakly supervised localization on coronary CT angiography with a deep 3-dimensional convolutional neural network

On the evaluation of deep learning interpretability methods for medical images under the scope of faithfulness

BACKGROUND AND OBJECTIVE

Evaluating the interpretability of Deep Learning models is crucial for building trust and gaining insights into their decision-making processes. In this work, we employ class activation map based attribution methods in a setting where only High-Resolution Class Activation Mapping (HiResCAM) is known to produce faithful explanations. The objective is to evaluate the quality of the attribution maps using quantitative metrics and investigate whether faithfulness aligns with the metrics results.

METHODS

We fine-tune pre-trained deep learning architectures over four medical image datasets in order to calculate attribution maps. The maps are evaluated on a threefold metrics basis utilizing well-established evaluation scores.

RESULTS

Our experimental findings suggest that the Area Over Perturbation Curve (AOPC) and Max-Sensitivity scores favor the HiResCAM maps. On the other hand, the Heatmap Assisted Accuracy Score (HAAS) does not provide insights to our comparison as it evaluates almost all maps as inaccurate. To this purpose we further compare our calculated values against values obtained over a diverse group of models which are trained on non-medical benchmark datasets, to eventually achieve more responsive results.

CONCLUSION

This study develops a series of experiments to discuss the connection between faithfulness and quantitative metrics over medical attribution maps. HiResCAM preserves the gradient effect on a pixel level ultimately producing high-resolution, informative and resilient mappings. In turn, this is depicted in the results of AOPC and Max-Sensitivity metrics, successfully identifying the faithful algorithm. In regards to HAAS, our experiments yield that it is sensitive over complex medical patterns, commonly characterized by strong color dependency and multiple attention areas.

Using artificial intelligence to study atherosclerosis from computed tomography imaging: A state-of-the-art review of the current literature

With the enormous progress in the field of cardiovascular imaging in recent years, computed tomography (CT) has become readily available to phenotype atherosclerotic coronary artery disease. New analytical methods using artificial intelligence (AI) enable the analysis of complex phenotypic information of atherosclerotic plaques. In particular, deep learning-based approaches using convolutional neural networks (CNNs) facilitate tasks such as lesion detection, segmentation, and classification. New radiotranscriptomic techniques even capture underlying bio-histochemical processes through higher-order structural analysis of voxels on CT images. In the near future, the international large-scale Oxford Risk Factors And Non-invasive Imaging (ORFAN) study will provide a powerful platform for testing and validating prognostic AI-based models. The goal is the transition of these new approaches from research settings into a clinical workflow. In this review, we present an overview of existing AI-based techniques with focus on imaging biomarkers to determine the degree of coronary inflammation, coronary plaques, and the associated risk. Further, current limitations using AI-based approaches as well as the priorities to address these challenges will be discussed. This will pave the way for an AI-enabled risk assessment tool to detect vulnerable atherosclerotic plaques and to guide treatment strategies for patients.

CAD-RADS scoring of coronary CT angiography with Multi-Axis Vision Transformer: A clinically-inspired deep learning pipeline

BACKGROUND AND OBJECTIVE

The standard non-invasive imaging technique used to assess the severity and extent of Coronary Artery Disease (CAD) is Coronary Computed Tomography Angiography (CCTA). However, manual grading of each patient's CCTA according to the CAD-Reporting and Data System (CAD-RADS) scoring is time-consuming and operator-dependent, especially in borderline cases. This work proposes a fully automated, and visually explainable, deep learning pipeline to be used as a decision support system for the CAD screening procedure. The pipeline performs two classification tasks: firstly, identifying patients who require further clinical investigations and secondly, classifying patients into subgroups based on the degree of stenosis, according to commonly used CAD-RADS thresholds.

METHODS

The pipeline pre-processes multiplanar projections of the coronary arteries, extracted from the original CCTAs, and classifies them using a fine-tuned Multi-Axis Vision Transformer architecture. With the aim of emulating the current clinical practice, the model is trained to assign a per-patient score by stacking the bi-dimensional longitudinal cross-sections of the three main coronary arteries along channel dimension. Furthermore, it generates visually interpretable maps to assess the reliability of the predictions.

RESULTS

When run on a database of 1873 three-channel images of 253 patients collected at the Monzino Cardiology Center in Milan, the pipeline obtained an AUC of 0.87 and 0.93 for the two classification tasks, respectively.

CONCLUSION

According to our knowledge, this is the first model trained to assign CAD-RADS scores learning solely from patient scores and not requiring finer imaging annotation steps that are not part of the clinical routine.

Fully automated artificial intelligence-based coronary CT angiography image processing: efficiency, diagnostic capability, and risk stratification

OBJECTIVES

To prospectively investigate whether fully automated artificial intelligence (FAAI)-based coronary CT angiography (CCTA) image processing is non-inferior to semi-automated mode in efficiency, diagnostic ability, and risk stratification of coronary artery disease (CAD).

MATERIALS AND METHODS

Adults with indications for CCTA were prospectively and consecutively enrolled at two hospitals and randomly assigned to either FAAI-based or semi-automated image processing using equipment workstations. Outcome measures were workflow efficiency, diagnostic accuracy for obstructive CAD (≥ 50% stenosis), and cardiovascular events at 2-year follow-up. The endpoints included major adverse cardiovascular events, hospitalization for unstable angina, and recurrence of cardiac symptoms. The non-inferiority margin was 3 percentage difference in diagnostic accuracy and C-index.

RESULTS

In total, 1801 subjects (62.7 ± 11.1 years) were included, of whom 893 and 908 were assigned to the FAAI-based and semi-automated modes, respectively. Image processing times were 121.0 ± 18.6 and 433.5 ± 68.4 s, respectively (p <0.001). Scan-to-report release times were 6.4 ± 2.7 and 10.5 ± 3.8 h, respectively (p < 0.001). Of all subjects, 152 and 159 in the FAAI-based and semi-automated modes, respectively, subsequently underwent invasive coronary angiography. The diagnostic accuracies for obstructive CAD were 94.7% (89.9-97.7%) and 94.3% (89.5-97.4%), respectively (difference 0.4%). Of all subjects, 779 and 784 in the FAAI-based and semi-automated modes were followed for 589 ± 182 days, respectively, and the C-statistic for cardiovascular events were 0.75 (0.67 to 0.83) and 0.74 (0.66 to 0.82) (difference 1%).

CONCLUSIONS

FAAI-based CCTA image processing significantly improves workflow efficiency than semi-automated mode, and is non-inferior in diagnosing obstructive CAD and risk stratification for cardiovascular events.

CLINICAL RELEVANCE STATEMENT

Conventional coronary CT angiography image processing is semi-automated. This observation shows that fully automated artificial intelligence-based image processing greatly improves efficiency, and maintains high diagnostic accuracy and the effectiveness in stratifying patients for cardiovascular events.

KEY POINTS

• Coronary CT angiography (CCTA) relies heavily on high-quality and fast image processing. • Full-automation CCTA image processing is clinically non-inferior to the semi-automated mode. • Full automation can facilitate the application of CCTA in early detection of coronary artery disease.

Transfer Learning in Multiple Hypothesis Testing

In this investigation, a synthesis of Convolutional Neural Networks (CNNs) and Bayesian inference is presented, leading to a novel approach to the problem of Multiple Hypothesis Testing (MHT). Diverging from traditional paradigms, this study introduces a sequence-based uncalibrated Bayes factor approach to test many hypotheses using the same family of sampling parametric models. A two-step methodology is employed: initially, a learning phase is conducted utilizing simulated datasets encompassing a wide spectrum of null and alternative hypotheses, followed by a transfer phase applying this fitted model to real-world experimental sequences. The outcome is a CNN model capable of navigating the complex domain of MHT with improved precision over traditional methods, also demonstrating robustness under varying conditions, including the number of true nulls and dependencies between tests. Although indications of empirical evaluations are presented and show that the methodology will prove useful, more work is required to provide a full evaluation from a theoretical perspective. The potential of this innovative approach is further illustrated within the critical domain of genomics. Although formal proof of the consistency of the model remains elusive due to the inherent complexity of the algorithms, this paper also provides some theoretical insights and advocates for continued exploration of this methodology.

Artificial intelligence evaluation of coronary computed tomography angiography for coronary stenosis classification and diagnosis

BACKGROUND

Ruling out obstructive coronary artery disease (CAD) using coronary computed tomography angiography (CCTA) is time-consuming and challenging. This study developed a deep learning (DL) model to assist in detecting obstructive CAD on CCTA to streamline workflows.

METHODS

In total, 2929 DICOM files and 7945 labels were extracted from curved planar reformatted CCTA images. A modified Inception V3 model was adopted. To validate the artificial intelligence (AI) model, two cardiologists labelled and adjudicated the classification of coronary stenosis on CCTA. The model was trained to differentiate the coronary artery into binary stenosis classifications <50% and ≥50% stenosis. Using the quantitative coronary angiography (QCA) consensus results as a reference standard, the performance of the AI model and CCTA radiology readers was compared by calculating Cohen's kappa coefficients at patient and vessel levels. The net reclassification index was used to evaluate the net benefit of the DL model.

RESULTS

The diagnostic accuracy of the AI model was 92.3% and 88.4% at the patient and vessel levels, respectively. Compared with CCTA radiology readers, the AI model had a better agreement for binary stenosis classification at both patient and vessel levels (Cohen kappa coefficient: .79 vs. .39 and .77 vs. .40, p < .0001). The AI model also exhibited significantly improved model discrimination and reclassification (Net reclassification index = .350; Z = 4.194; p < .001).

CONCLUSIONS

The developed AI model identified obstructive CAD, and the model results correlated well with QCA results. Incorporating the model into the reporting system of CCTA may improve workflows.

Detecting and Characterizing Inferior Vena Cava Filters on Abdominal Computed Tomography with Data-Driven Computational Frameworks

Two data-driven algorithms were developed for detecting and characterizing Inferior Vena Cava (IVC) filters on abdominal computed tomography to assist healthcare providers with the appropriate management of these devices to decrease complications: one based on 2-dimensional data and transfer learning (2D + TL) and an augmented version of the same algorithm which accounts for the 3-dimensional information leveraging recurrent convolutional neural networks (3D + RCNN). The study contains 2048 abdominal computed tomography studies obtained from 439 patients who underwent IVC filter placement during the 10-year period from January 1st, 2009, to January 1st, 2019. Among these, 399 patients had retrievable filters, and 40 had non-retrievable filter types. The reference annotations for the filter location were obtained through a custom-developed interface. The ground truth annotations for the filter types were determined based on the electronic medical record and physician review of imaging. The initial stage of the framework returns a list of locations containing metallic objects based on the density of the structure. The second stage processes the candidate locations and determines which one contains an IVC filter. The final stage of the pipeline classifies the filter types as retrievable vs. non-retrievable. The computational models are trained using Tensorflow Keras API on an Nvidia Quadro GV100 system. We utilized a fine-tuning supervised training strategy to conduct our experiments. We find that the system achieves high sensitivity on detecting the filter locations with a high confidence value. The 2D + TL model achieved a sensitivity of 0.911 and a precision of 0.804, and the 3D + RCNN model achieved a sensitivity of 0.923 and a precision of 0.853 for filter detection. The system confidence for the IVC location predictions is high: 0.993 for 2D + TL and 0.996 for 3D + RCNN. The filter type prediction component of the system achieved 0.945 sensitivity, 0.882 specificity, and 0.97 AUC score with 2D + TL and 0. 940 sensitivity, 0.927 specificity, and 0.975 AUC score with 3D + RCNN. With the intent to create tools to improve patient outcomes, this study describes the initial phase of a computational framework to support healthcare providers in detecting patients with retained IVC filters, so an individualized decision can be made to remove these devices when appropriate, to decrease complications. To our knowledge, this is the first study that curates abdominal computed tomography (CT) scans and presents an algorithm for automated detection and characterization of IVC filters.

Context-aware deep network for coronary artery stenosis classification in coronary CT angiography

Automatic coronary artery stenosis grading plays an important role in the diagnosis of coronary artery disease. Due to the difficulty of learning the informative features from varying grades of stenosis, it is still a challenging task to identify coronary artery stenosis from coronary CT angiography (CCTA). In this paper, we propose a context-aware deep network (CADN) for coronary artery stenosis classification. The proposed method integrates 3D CNN with Transformer to improve the feature representation of coronary artery stenosis in CCTA. We evaluate the proposed method on a multicenter dataset (APOLLO study with NCT05509010). Experimental results show that our proposed method can achieve the accuracy of 0.84, 0.83, and 0.86 for stenosis diagnosis on the lesion, artery, and patient levels, respectively.

Survey of explainable artificial intelligence techniques for biomedical imaging with deep neural networks

Artificial Intelligence (AI) techniques of deep learning have revolutionized the disease diagnosis with their outstanding image classification performance. In spite of the outstanding results, the widespread adoption of these techniques in clinical practice is still taking place at a moderate pace. One of the major hindrance is that a trained Deep Neural Networks (DNN) model provides a prediction, but questions about why and how that prediction was made remain unanswered. This linkage is of utmost importance for the regulated healthcare domain to increase the trust in the automated diagnosis system by the practitioners, patients and other stakeholders. The application of deep learning for medical imaging has to be interpreted with caution due to the health and safety concerns similar to blame attribution in the case of an accident involving autonomous cars. The consequences of both a false positive and false negative cases are far reaching for patients' welfare and cannot be ignored. This is exacerbated by the fact that the state-of-the-art deep learning algorithms comprise of complex interconnected structures, millions of parameters, and a 'black box' nature, offering little understanding of their inner working unlike the traditional machine learning algorithms. Explainable AI (XAI) techniques help to understand model predictions which help develop trust in the system, accelerate the disease diagnosis, and meet adherence to regulatory requirements. This survey provides a comprehensive review of the promising field of XAI for biomedical imaging diagnostics. We also provide a categorization of the XAI techniques, discuss the open challenges, and provide future directions for XAI which would be of interest to clinicians, regulators and model developers.

Developing a Deep-Learning-Based Coronary Artery Disease Detection Technique Using Computer Tomography Images

Coronary artery disease (CAD) is one of the major causes of fatalities across the globe. The recent developments in convolutional neural networks (CNN) allow researchers to detect CAD from computed tomography (CT) images. The CAD detection model assists physicians in identifying cardiac disease at earlier stages. The recent CAD detection models demand a high computational cost and a more significant number of images. Therefore, this study intends to develop a CNN-based CAD detection model. The researchers apply an image enhancement technique to improve the CT image quality. The authors employed You look only once (YOLO) V7 for extracting the features. Aquila optimization is used for optimizing the hyperparameters of the UNet++ model to predict CAD. The proposed feature extraction technique and hyperparameter tuning approach reduces the computational costs and improves the performance of the UNet++ model. Two datasets are utilized for evaluating the performance of the proposed CAD detection model. The experimental outcomes suggest that the proposed method achieves an accuracy, recall, precision, F1-score, Matthews correlation coefficient, and Kappa of 99.4, 98.5, 98.65, 98.6, 95.35, and 95 and 99.5, 98.95, 98.95, 98.95, 96.35, and 96.25 for datasets 1 and 2, respectively. In addition, the proposed model outperforms the recent techniques by obtaining the area under the receiver operating characteristic and precision-recall curve of 0.97 and 0.95, and 0.96 and 0.94 for datasets 1 and 2, respectively. Moreover, the proposed model obtained a better confidence interval and standard deviation of [98.64–98.72] and 0.0014, and [97.41–97.49] and 0.0019 for datasets 1 and 2, respectively. The study’s findings suggest that the proposed model can support physicians in identifying CAD with limited resources.

Deep Learning for Head and Neck CT Angiography: Stenosis and Plaque Classification

Background Studies have rarely investigated stenosis detection from head and neck CT angiography scans because accurate interpretation is time consuming and labor intensive. Purpose To develop an automated convolutional neural network-based method for accurate stenosis detection and plaque classification in head and neck CT angiography images and compare its performance with that of radiologists. Materials and Methods A deep learning (DL) algorithm was constructed and trained with use of head and neck CT angiography images that were collected retrospectively from four tertiary hospitals between March 2020 and July 2021. CT scans were partitioned into training, validation, and independent test sets at a ratio of 7:2:1. An independent test set of CT angiography scans was collected prospectively between October 2021 and December 2021 in one of the four tertiary centers. Stenosis grade categories were as follows: mild stenosis (<50%), moderate stenosis (50%-69%), severe stenosis (70%-99%), and occlusion (100%). The stenosis diagnosis and plaque classification of the algorithm were compared with the ground truth of consensus by two radiologists (with more than 10 years of experience). The performance of the models was analyzed in terms of accuracy, sensitivity, specificity, and areas under the receiver operating characteristic curve. Results There were 3266 patients (mean age ± SD, 62 years ± 12; 2096 men) evaluated. The consistency between radiologists and the DL-assisted algorithm on plaque classification was 85.6% (320 of 374 cases [95% CI: 83.2, 88.6]) on a per-vessel basis. Moreover, the artificial intelligence model assisted in visual assessment, such as increasing confidence in the degree of stenosis. This reduced the time needed for diagnosis and report writing of radiologists from 28.8 minutes ± 5.6 to 12.4 minutes ± 2.0 (P < .001). Conclusion A deep learning algorithm for head and neck CT angiography interpretation accurately determined vessel stenosis and plaque classification and had equivalent diagnostic performance when compared with experienced radiologists. © RSNA, 2023 Supplemental material is available for this article.

How scan parameter choice affects deep learning-based coronary artery disease assessment from computed tomography

Recently, algorithms capable of assessing the severity of Coronary Artery Disease (CAD) in form of the Coronary Artery Disease-Reporting and Data System (CAD-RADS) grade from Coronary Computed Tomography Angiography (CCTA) scans using Deep Learning (DL) were proposed. Before considering to apply these algorithms in clinical practice, their robustness regarding different commonly used Computed Tomography (CT)-specific image formation parameters-including denoising strength, slab combination, and reconstruction kernel-needs to be evaluated. For this study, we reconstructed a data set of 500 patient CCTA scans under seven image formation parameter configurations. We select one default configuration and evaluate how varying individual parameters impacts the performance and stability of a typical algorithm for automated CAD assessment from CCTA. This algorithm consists of multiple preprocessing and a DL prediction step. We evaluate the influence of the parameter changes on the entire pipeline and additionally on only the DL step by propagating the centerline extraction results of the default configuration to all others. We consider the standard deviation of the CAD severity prediction grade difference between the default and variation configurations to assess the stability w.r.t. parameter changes. For the full pipeline we observe slight instability (± 0.226 CAD-RADS) for all variations. Predictions are more stable with centerlines propagated from the default to the variation configurations (± 0.122 CAD-RADS), especially for differing denoising strengths (± 0.046 CAD-RADS). However, stacking slabs with sharp boundaries instead of mixing slabs in overlapping regions (called true stack ± 0.313 CAD-RADS) and increasing the sharpness of the reconstruction kernel (± 0.150 CAD-RADS) leads to unstable predictions. Regarding the clinically relevant tasks of excluding CAD (called rule-out; AUC default 0.957, min 0.937) and excluding obstructive CAD (called hold-out; AUC default 0.971, min 0.964) the performance remains on a high level for all variations. Concluding, an influence of reconstruction parameters on the predictions is observed. Especially, scans reconstructed with the true stack parameter need to be treated with caution when using a DL-based method. Also, reconstruction kernels which are underrepresented in the training data increase the prediction uncertainty.

Artificial intelligence in atherosclerotic disease: Applications and trends

Atherosclerotic cardiovascular disease (ASCVD) is the most common cause of death globally. Increasing amounts of highly diverse ASCVD data are becoming available and artificial intelligence (AI) techniques now bear the promise of utilizing them to improve diagnosis, advance understanding of disease pathogenesis, enable outcome prediction, assist with clinical decision making and promote precision medicine approaches. Machine learning (ML) algorithms in particular, are already employed in cardiovascular imaging applications to facilitate automated disease detection and experts believe that ML will transform the field in the coming years. Current review first describes the key concepts of AI applications from a clinical standpoint. We then provide a focused overview of current AI applications in four main ASCVD domains: coronary artery disease (CAD), peripheral arterial disease (PAD), abdominal aortic aneurysm (AAA), and carotid artery disease. For each domain, applications are presented with refer to the primary imaging modality used [e.g., computed tomography (CT) or invasive angiography] and the key aim of the applied AI approaches, which include disease detection, phenotyping, outcome prediction, and assistance with clinical decision making. We conclude with the strengths and limitations of AI applications and provide future perspectives.

Clinician's guide to trustworthy and responsible artificial intelligence in cardiovascular imaging

A growing number of artificial intelligence (AI)-based systems are being proposed and developed in cardiology, driven by the increasing need to deal with the vast amount of clinical and imaging data with the ultimate aim of advancing patient care, diagnosis and prognostication. However, there is a critical gap between the development and clinical deployment of AI tools. A key consideration for implementing AI tools into real-life clinical practice is their "trustworthiness" by end-users. Namely, we must ensure that AI systems can be trusted and adopted by all parties involved, including clinicians and patients. Here we provide a summary of the concepts involved in developing a "trustworthy AI system." We describe the main risks of AI applications and potential mitigation techniques for the wider application of these promising techniques in the context of cardiovascular imaging. Finally, we show why trustworthy AI concepts are important governing forces of AI development.

Using artificial intelligence in the development of diagnostic models of coronary artery disease with imaging markers: A scoping review

Background

Coronary artery disease (CAD) is a progressive disease of the blood vessels supplying the heart, which leads to coronary artery stenosis or obstruction and is life-threatening. Early diagnosis of CAD is essential for timely intervention. Imaging tests are widely used in diagnosing CAD, and artificial intelligence (AI) technology is used to shed light on the development of new imaging diagnostic markers.

Objective

We aim to investigate and summarize how AI algorithms are used in the development of diagnostic models of CAD with imaging markers.

Methods

This scoping review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) guideline. Eligible articles were searched in PubMed and Embase. Based on the predefined included criteria, articles on coronary heart disease were selected for this scoping review. Data extraction was independently conducted by two reviewers, and a narrative synthesis approach was used in the analysis.

Results

A total of 46 articles were included in the scoping review. The most common types of imaging methods complemented by AI included single-photon emission computed tomography (15/46, 32.6%) and coronary computed tomography angiography (15/46, 32.6%). Deep learning (DL) (41/46, 89.2%) algorithms were used more often than machine learning algorithms (5/46, 10.8%). The models yielded good model performance in terms of accuracy, sensitivity, specificity, and AUC. However, most of the primary studies used a relatively small sample (n < 500) in model development, and only few studies (4/46, 8.7%) carried out external validation of the AI model.

Conclusion

As non-invasive diagnostic methods, imaging markers integrated with AI have exhibited considerable potential in the diagnosis of CAD. External validation of model performance and evaluation of clinical use aid in the confirmation of the added value of markers in practice.

Systematic review registration

[https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022306638], identifier [CRD42022306638].

Artificial intelligence can detect left ventricular dilatation on contrast-enhanced thoracic computer tomography relative to cardiac magnetic resonance imaging

OBJECTIVES

To assess the diagnostic accuracy of an automated algorithm to detect left ventricular (LV) dilatation on non-ECG gated CT, using cardiac magnetic resonance (CMR) as reference standard.

METHODS

Consecutive patients with contrast-enhanced CT thorax and CMR within 31 days (2016-2020) were analysed (n = 84). LV dilatation was defined against age-, sex- and body surface area-specific values for CMR. CTs underwent automated artificial intelligence(AI)-derived analysis that segmented ventricular chambers, presenting maximal LV diameter and volume. Area under the receiver operator curve (AUC-ROC) analysis identified CT thresholds with ≥90% sensitivity and highest specificity and ≥90% specificity with highest sensitivity. Youden's Index was used to identify thresholds with optimised sensitivity and specificity.

RESULTS

Automated diameter analysis was feasible in 92% of cases (77/84; 45 men, age 61 ± 14 years, mean CT to CMR interval 10 ± 8 days). Relative to CMR as a reference standard, 45% had LV dilatation. In males, an automated LV diameter measurement of ≥55.5 mm was ≥90% specific for CMR-defined LV dilatation (positive predictive value (PPV) 85.7%, negative predictive value (NPV) 61.2%, accuracy 68.9%). In females, an LV diameter of ≥49.7 mm was ≥90% specific for CMR-defined LV dilatation (PPV 66.7%, NPV 73.1%, accuracy 71.9%). AI CT volumetry data did not significantly improve AUC performance.

CONCLUSION

Fully automated AI-derived analysis LV dilatation on routine unselected non-gated contrast-enhanced CT thorax studies is feasible. We have defined thresholds for the detection of LV dilatation on CT relative to CMR, which could be used to routinely screen for dilated cardiomyopathy at the time of CT.

ADVANCES IN KNOWLEDGE

We show, for the first time, that a fully-automated AI-derived analysis of maximal LV chamber axial diameter on non-ECG-gated thoracic CT is feasible in unselected real-world cases and that the derived measures can predict LV dilatation relative to cardiac magnetic resonance imaging, the non-invasive reference standard for determining cardiac chamber size. We have derived sex-specific cut-off values to screen for LV dilatation on routine contrast-enhanced thoracic CT. Future work should validate these thresholds and determine if technology can alter clinical outcomes in a cost-effective manner.

Reproducibility and Repeatability of Coronary Computed Tomography Angiography (CCTA) Image Segmentation in Detecting Atherosclerosis: A Radiomics Study

Atherosclerosis is known as the leading factor in heart disease with the highest mortality rate among the Malaysian population. Usually, the gold standard for diagnosing atherosclerosis is by using the coronary computed tomography angiography (CCTA) technique to look for plaque within the coronary artery. However, qualitative diagnosis for noncalcified atherosclerosis is vulnerable to false-positive diagnoses, as well as inconsistent reporting between observers. In this study, we assess the reproducibility and repeatability of segmenting atherosclerotic lesions manually and semiautomatically in CCTA images to identify the most appropriate CCTA image segmentation method for radiomics analysis to quantitatively extract the atherosclerotic lesion. Thirty (30) CCTA images were taken retrospectively from the radiology image database of Hospital Canselor Tuanku Muhriz (HCTM), Kuala Lumpur, Malaysia. We extract 11,700 radiomics features which include the first-order, second-order and shape features from 180 times of image segmentation. The interest vessels were segmentized manually and semiautomatically using LIFEx (Version 7.0.15, Institut Curie, Orsay, France) software by two independent radiology experts, focusing on three main coronary blood vessels. As a result, manual segmentation with a soft-tissuewindowing setting yielded higher repeatability as compared to semiautomatic segmentation with a significant intraclass correlation coefficient (intra-CC) 0.961 for thefirst-order and shape features; intra-CC of 0.924 for thesecond-order features with p < 0.001. Meanwhile, the semiautomatic segmentation has higher reproducibility as compared to manual segmentation with significant interclass correlation coefficient (inter-CC) of 0.920 (first-order features) and a good interclass correlation coefficient of 0.839 for the second-order features with p < 0.001. The first-order, shape order and second-order features for both manual and semiautomatic segmentation have an excellent percentage of reproducibility and repeatability (intra-CC > 0.9). In conclusion, semi-automated segmentation is recommended for inter-observer study while manual segmentation with soft tissue-windowing can be used for single observer study.

Diagnostic Accuracy and Generalizability of a Deep Learning-Based Fully Automated Algorithm for Coronary Artery Stenosis Detection on CCTA: A Multi-Centre Registry Study

Aims: In this retrospective, multi-center study, we aimed to estimate the diagnostic accuracy and generalizability of an established deep learning (DL)-based fully automated algorithm in detecting coronary stenosis on coronary computed tomography angiography (CCTA).

Methods and results: A total of 527 patients (33.0% female, mean age: 62.2 ± 10.2 years) with suspected coronary artery disease (CAD) who underwent CCTA and invasive coronary angiography (ICA) were enrolled from 27 hospitals from January 2016 to August 2019. Using ICA as a standard reference, the diagnostic accuracy of the DL algorithm in the detection of ≥50% stenosis was compared to that of expert readers. In the vessel-based evaluation, the DL algorithm had a higher sensitivity (65.7%) and negative predictive value (NPV) (78.8%) and a significantly higher area under the curve (AUC) (0.83, p < 0.001). In the patient-based evaluation, the DL algorithm achieved a higher sensitivity (90.0%), NPV (52.2%) and AUC (0.81). Generalizability analysis of the DL algorithm was conducted by comparing its diagnostic performance in subgroups stratified by sex, age, geographic area and CT scanner type. The AUCs of the DL algorithm in the aforementioned subgroups ranged from 0.79 to 0.86 and from 0.75 to 0.93 in the vessel-based and patient-based evaluations, both without significant group differences (p > 0.05). The DL algorithm significantly reduced post-processing time (160 [IQR:139–192] seconds), in comparison to manual work (p < 0.001).

Conclusions: The DL algorithm performed no inferior to expert readers in CAD diagnosis on CCTA and had good generalizability and time efficiency.

Precise anatomical localization and classification of rib fractures on CT using a convolutional neural network

OBJECTIVE

To develop a convolutional neural network (CNN) model for the detection, precise anatomical localization (right 1-12th and left 1-12th) and classification (fresh, healing and old fractures) of rib fractures automatically, and to compare the performance with the experienced radiologists.

MATERIALS AND METHODS

A total of 640 rib fracture patients with 340,501 annotations were retrospectively collected from three hospitals. They consisted of a classification training dataset (n = 482), a localization training dataset (n = 30), an internal testing dataset (n = 90) and an external testing dataset (n = 38). RetinaNet with rib localization postprocessing and the result merging technique were employed to structure the CNN model. ROC curve, free-response ROC curve, AUC, precision, recall, and F1-score were calculated to choose the better option between model I (training classification and localization data together) and model II (adding an additional classification model to model I).

RESULTS

The detection and classification performance of rib fractures was better in model II than in model I. The sensitivity of localization reached 97.11% and 94.87% on the right and left ribs, respectively. In the external dataset with different CT scanner and slice thickness, model II showed better diagnostic performance. Moreover, the CNN model was superior in diagnosing fresh and healing fractures to 5 radiologists and consumed shorter diagnosis time.

CONCLUSIONS

Our CNN model was capable of detection, precise anatomical localization, and classification of rib fractures automatically.

Recent Trends in Artificial Intelligence-Assisted Coronary Atherosclerotic Plaque Characterization

Coronary artery disease is a major cause of morbidity and mortality worldwide. Its underlying histopathology is the atherosclerotic plaque, which comprises lipid, fibrous and—when chronic—calcium components. Intravascular ultrasound (IVUS) and intravascular optical coherence tomography (IVOCT) performed during invasive coronary angiography are reference standards for characterizing the atherosclerotic plaque. Fine image spatial resolution attainable with contemporary coronary computed tomographic angiography (CCTA) has enabled noninvasive plaque assessment, including identifying features associated with vulnerable plaques known to presage acute coronary events. Manual interpretation of IVUS, IVOCT and CCTA images demands scarce physician expertise and high time cost. This has motivated recent research into and development of artificial intelligence (AI)-assisted methods for image processing, feature extraction, plaque identification and characterization. We performed parallel searches of the medical and technical literature from 1995 to 2021 focusing respectively on human plaque characterization using various imaging modalities and the use of AI-assisted computer aided diagnosis (CAD) to detect and classify atherosclerotic plaques, including their composition and the presence of high-risk features denoting vulnerable plaques. A total of 122 publications were selected for evaluation and the analysis was summarized in terms of data sources, methods—machine versus deep learning—and performance metrics. Trends in AI-assisted plaque characterization are detailed and prospective research challenges discussed. Future directions for the development of accurate and efficient CAD systems to characterize plaque noninvasively using CCTA are proposed.

Automated Artery Localization and Vessel Wall Segmentation using Tracklet Refinement and Polar Conversion

Quantitative analysis of blood vessel wall structures is important to study atherosclerotic diseases and assess cardiovascular event risks. To achieve this, accurate identification of vessel luminal and outer wall contours is needed. Computer-assisted tools exist, but manual preprocessing steps, such as region of interest identification and/or boundary initialization, are still needed. In addition, prior knowledge of the ring shape of vessel walls has not been fully explored in designing segmentation methods. In this work, a fully automated artery localization and vessel wall segmentation system is proposed. A tracklet refinement algorithm was adapted to robustly identify the artery of interest from a neural network-based artery centerline identification architecture. Image patches were extracted from the centerlines and converted in a polar coordinate system for vessel wall segmentation. The segmentation method used 3D polar information and overcame problems such as contour discontinuity, complex vessel geometry, and interference from neighboring vessels. Verified by a large (>32000 images) carotid artery dataset collected from multiple sites, the proposed system was shown to better automatically segment the vessel wall than traditional vessel wall segmentation methods or standard convolutional neural network approaches. In addition, a segmentation uncertainty score was estimated to effectively identify slices likely to have errors and prompt manual confirmation of the segmentation. This robust vessel wall segmentation system has applications in different vascular beds and will facilitate vessel wall feature extraction and cardiovascular risk assessment.