Artificial intelligence in computed tomography plaque characterization: A review

Effects of adaptive or fixed thresholds and different platforms on the assessment of plaque characteristics using coronary computed tomography angiography

BACKGROUND

Coronary computed tomography angiography (CCTA) is used to evaluate components of atherosclerosis. Either adaptive or diverse, fixed Hounsfield Units (HU) are used to define components such as low attenuation (LAP), mixed (MP) and calcified plaque (CP). Comparisons of different platforms and different thresholding approaches have not been extensively evaluated. We compare two fixed threshold options to an adaptive threshold option within a specific platform and to fixed threshold options measured with another platform.

METHODS

Coronary segments (n ​= ​24) of good image quality, with well-defined boundaries and representing a broad range of atheroma were analyzed for LAP, MP and CP. Thresholds for LAP vs MP and MP vs CP were either Fixed30/350, Fixed75/350 or based on an automatically determined Adaptive option. Pearson correlation and Bland-Altman analyses were undertaken.

RESULTS

Within a single platform, measures were highly correlated irrespective of use of Adaptive or Fixed30/350 and Fixed75/350 thresholds (R ​≥ ​0.819, p ​< ​0.000001). The correlation slope for measures of LAP progressively diminished comparing the Adaptive versus Fixed30/350 and the Fixed75/350 versus the Fixed30/350 approaches but bias was small. Between-platform comparisons yielded less optimal results, particularly with respect to measures of LAP and with one platform yielding both very small LAP volumes and very small ranges of volumes.

CONCLUSION

Measures of plaque components are highly correlated irrespective of use of Adaptive or Fixed threshold approaches within a given platform. But measures are more affected by the specific proprietary algorithms employed than by specific thresholding options, especially for LAP.

Carotid stenosis and cryptogenic stroke

OBJECTIVES

Cryptogenic stroke represents a type of ischemic stroke with an unknown origin, presenting a significant challenge in both stroke management and prevention. According to the Trial of Org 10,172 in Acute Stroke Treatment criteria, a stroke is categorized as being caused by large artery atherosclerosis only when there is >50% luminal narrowing of the ipsilateral internal carotid artery. However, nonstenosing carotid artery plaques can be an underlying cause of ischemic stroke. Indeed, emerging evidence documents that some features of plaque vulnerability may act as an independent risk factor, regardless of the degree of stenosis, in precipitating cerebrovascular events. This review, drawing from an array of imaging-based studies, explores the predictive values of carotid imaging modalities in the detection of nonstenosing carotid plaque (<50%), that could be the cause of a cerebrovascular event when some features of vulnerability are present.

METHODS

Google Scholar, Scopus, and PubMed were searched for articles on cryptogenic stroke and those reporting the association between cryptogenic stroke and imaging features of carotid plaque vulnerability.

RESULTS

Despite extensive diagnostic evaluations, the etiology of a considerable proportion of strokes remains undetermined, contributing to the recurrence rate and persistent morbidity in affected individuals. Advances in imaging modalities, such as magnetic resonance imaging, computed tomography scans, and ultrasound examination, facilitate more accurate detection of nonstenosing carotid artery plaque and allow better stratification of stroke risk, leading to a more tailored treatment strategy.

CONCLUSIONS

Early detection of nonstenosing carotid plaque with features of vulnerability through carotid imaging techniques impacts the clinical management of cryptogenic stroke, resulting in refined stroke subtype classification and improved patient management. Additional research is required to validate these findings and recommend the integration of these state-of-the-art imaging methodologies into standard diagnostic protocols to improve stroke management and prevention.

Advancements in non-invasive imaging of atherosclerosis: Future perspectives

Atherosclerosis is a chronic inflammatory disease characterized by the buildup of plaques in arterial walls, leading to cardiovascular diseases and high morbidity and mortality rates worldwide. Non-invasive imaging techniques play a crucial role in evaluating patients with suspected or established atherosclerosis. However, there is a growing body of evidence suggesting the need to visualize the underlying processes of plaque progression and rupture to enhance risk stratification. This review explores recent advancements in non-invasive assessment of atherosclerosis, focusing on computed tomography, magnetic resonance imaging, and nuclear imaging. These advancements provide valuable insights into the assessment and management of atherosclerosis, potentially leading to better risk stratification and improved patient outcomes.

Artificial Intelligence in the Differential Diagnosis of Cardiomyopathy Phenotypes

Artificial intelligence (AI) is rapidly being applied to the medical field, especially in the cardiovascular domain. AI approaches have demonstrated their applicability in the detection, diagnosis, and management of several cardiovascular diseases, enhancing disease stratification and typing. Cardiomyopathies are a leading cause of heart failure and life-threatening ventricular arrhythmias. Identifying the etiologies is fundamental for the management and diagnostic pathway of these heart muscle diseases, requiring the integration of various data, including personal and family history, clinical examination, electrocardiography, and laboratory investigations, as well as multimodality imaging, making the clinical diagnosis challenging. In this scenario, AI has demonstrated its capability to capture subtle connections from a multitude of multiparametric datasets, enabling the discovery of hidden relationships in data and handling more complex tasks than traditional methods. This review aims to present a comprehensive overview of the main concepts related to AI and its subset. Additionally, we review the existing literature on AI-based models in the differential diagnosis of cardiomyopathy phenotypes, and we finally examine the advantages and limitations of these AI approaches.

Deep Learning Paradigm and Its Bias for Coronary Artery Wall Segmentation in Intravascular Ultrasound Scans: A Closer Look

BACKGROUND AND MOTIVATION

Coronary artery disease (CAD) has the highest mortality rate; therefore, its diagnosis is vital. Intravascular ultrasound (IVUS) is a high-resolution imaging solution that can image coronary arteries, but the diagnosis software via wall segmentation and quantification has been evolving. In this study, a deep learning (DL) paradigm was explored along with its bias.

METHODS

Using a PRISMA model, 145 best UNet-based and non-UNet-based methods for wall segmentation were selected and analyzed for their characteristics and scientific and clinical validation. This study computed the coronary wall thickness by estimating the inner and outer borders of the coronary artery IVUS cross-sectional scans. Further, the review explored the bias in the DL system for the first time when it comes to wall segmentation in IVUS scans. Three bias methods, namely (i) ranking, (ii) radial, and (iii) regional area, were applied and compared using a Venn diagram. Finally, the study presented explainable AI (XAI) paradigms in the DL framework.

FINDINGS AND CONCLUSIONS

UNet provides a powerful paradigm for the segmentation of coronary walls in IVUS scans due to its ability to extract automated features at different scales in encoders, reconstruct the segmented image using decoders, and embed the variants in skip connections. Most of the research was hampered by a lack of motivation for XAI and pruned AI (PAI) models. None of the UNet models met the criteria for bias-free design. For clinical assessment and settings, it is necessary to move from a paper-to-practice approach.

Advancing the Management of Long COVID by Integrating into Health Informatics Domain: Current and Future Perspectives

The ongoing COVID-19 pandemic has profoundly affected millions of lives globally, with some individuals experiencing persistent symptoms even after recovering. Understanding and managing the long-term sequelae of COVID-19 is crucial for research, prevention, and control. To effectively monitor the health of those affected, maintaining up-to-date health records is essential, and digital health informatics apps for surveillance play a pivotal role. In this review, we overview the existing literature on identifying and characterizing long COVID manifestations through hierarchical classification based on Human Phenotype Ontology (HPO). We outline the aspects of the National COVID Cohort Collaborative (N3C) and Researching COVID to Enhance Recovery (RECOVER) initiative in artificial intelligence (AI) to identify long COVID. Through knowledge exploration, we present a concept map of clinical pathways for long COVID, which offers insights into the data required and explores innovative frameworks for health informatics apps for tackling the long-term effects of COVID-19. This study achieves two main objectives by comprehensively reviewing long COVID identification and characterization techniques, making it the first paper to explore incorporating long COVID as a variable risk factor within a digital health informatics application. By achieving these objectives, it provides valuable insights on long COVID's challenges and impact on public health.

Clinical Characteristics and Outcomes of Aortic Arch Emergencies: Takayasu Disease, Fibromuscular Dysplasia, and Aortic Arch Pathologies: A Retrospective Study and Review of the Literature

Non-atherosclerotic aortic arch pathologies (NA-AAPs) and anatomical variants are characterized as rare cardiovascular diseases with a low incidence rate, below 1 case per 2000 population, but enormous heterogeneity in terms of anatomical variants, i.e., Takayasu disease (TAK) and fibromuscular dysplasia (FMD). In specific clinical scenarios, NA-AAPs constitute life-threatening disorders.

METHODS

In this study, 82 (1.07%) consecutive patients with NA-AAPs (including 38 TAKs, 26 FMDs, and 18 other AAPs) out of 7645 patients who underwent endovascular treatment (EVT) for the aortic arch and its side-branch diseases at a single institution between 2002 and 2022 were retrospectively reviewed. The recorded demographic, biochemical, diagnostic, operative, and postoperative factors were reviewed, and the functional outcomes were determined during follow-up. A systematic review of the literature was also performed.

RESULTS

The study group comprised 65 (79.3%) female and 17 (21.7%) male subjects with a mean age of 46.1 ± 14.9 years. Overall, 62 (75.6%) patients were diagnosed with either cerebral ischemia symptoms or aortic arch dissection on admission. The EVT was feasible in 59 (72%) patients, whereas 23 (28%) patients were referred for medical treatment. In EVT patients, severe periprocedural complications occurred in two (3.39%) patients, including one periprocedural death and one cerebral hyperperfusion syndrome. During a median follow-up period of 64 months, cardiovascular events occurred in 24 (29.6%) patients (5 deaths, 13 ISs, and 6 myocardial infarctions). Repeated EVT for the index lesion was performed in 21/59 (35.6%) patients, including 19/33 (57.6%) in TAK and 2/13 (15.4%) in FMD. In the AAP group, one patient required additional stent-graft implantation for progressing dissection to the iliac arteries at 12 months. A baseline white blood count (odds ratio [HR]: 1.25, 95% confidence interval [CI]: 1.11-1.39; p < 0.001) was the only independent prognostic factor for recurrent stenosis, while a baseline hemoglobin level (HR: 0.73, 95%CI: 0.59-0.89; p = 0.002) and coronary involvement (HR: 4.11, 95%CI: 1.74-9.71; p = 0.001) were independently associated with a risk of major cardiac and cerebral events according to the multivariate Cox proportional hazards regression analysis.

CONCLUSIONS

This study showed that AAPs should not be neglected in clinical settings, as it can be a life-threatening condition requiring a multidisciplinary approach. The knowledge of prognostic risk factors for adverse outcomes may improve surveillance in this group of patients.

Computed Tomography Assessment of Coronary Atherosclerosis: From Threshold-Based Evaluation to Histologically Validated Plaque Quantification

Arterial plaque rupture and thrombosis is the primary cause of major cardiovascular and neurovascular events. The identification of atherosclerosis, especially high-risk plaques, is therefore crucial to identify high-risk patients and to implement preventive therapies. Computed tomography angiography has the ability to visualize and characterize vascular plaques. The standard methods for plaque evaluation rely on the assessment of plaque burden, stenosis severity, the presence of positive remodeling, napkin ring sign, and spotty calcification, as well as Hounsfield Unit (HU)-based thresholding for plaque quantification; the latter with multiple shortcomings. Semiautomated threshold-based segmentation techniques with predefined HU ranges identify and quantify limited plaque characteristics, such as low attenuation, non-calcified, and calcified plaque components. Contrary to HU-based thresholds, histologically validated plaque characterization, and quantification, an emerging Artificial intelligence-based approach has the ability to differentiate specific tissue types based on a biological correlate, such as lipid-rich necrotic core and intraplaque hemorrhage that determine plaque vulnerability. In this article, we review the relevance of plaque characterization and quantification and discuss the benefits and limitations of the currently available plaque assessment and classification techniques.

Artificial Intelligence Applications in Cardiovascular Magnetic Resonance Imaging: Are We on the Path to Avoiding the Administration of Contrast Media?

In recent years, cardiovascular imaging examinations have experienced exponential growth due to technological innovation, and this trend is consistent with the most recent chest pain guidelines. Contrast media have a crucial role in cardiovascular magnetic resonance (CMR) imaging, allowing for more precise characterization of different cardiovascular diseases. However, contrast media have contraindications and side effects that limit their clinical application in determinant patients. The application of artificial intelligence (AI)-based techniques to CMR imaging has led to the development of non-contrast models. These AI models utilize non-contrast imaging data, either independently or in combination with clinical and demographic data, as input to generate diagnostic or prognostic algorithms. In this review, we provide an overview of the main concepts pertaining to AI, review the existing literature on non-contrast AI models in CMR, and finally, discuss the strengths and limitations of these AI models and their possible future development.

Dual-Source Photon-Counting Computed Tomography-Part I: Clinical Overview of Cardiac CT and Coronary CT Angiography Applications

The photon-counting detector (PCD) is a new computed tomography detector technology (photon-counting computed tomography, PCCT) that provides substantial benefits for cardiac and coronary artery imaging. Compared with conventional CT, PCCT has multi-energy capability, increased spatial resolution and soft tissue contrast with near-null electronic noise, reduced radiation exposure, and optimization of the use of contrast agents. This new technology promises to overcome several limitations of traditional cardiac and coronary CT angiography (CCT/CCTA) including reduction in blooming artifacts in heavy calcified coronary plaques or beam-hardening artifacts in patients with coronary stents, and a more precise assessment of the degree of stenosis and plaque characteristic thanks to its better spatial resolution. Another potential application of PCCT is the use of a double-contrast agent to characterize myocardial tissue. In this current overview of the existing PCCT literature, we describe the strengths, limitations, recent applications, and promising developments of employing PCCT technology in CCT.

Pearls and Pitfalls of Carotid Artery Imaging: Ultrasound, Computed Tomography Angiography, and MR Imaging

Stroke represents a major cause of morbidity and mortality worldwide with carotid atherosclerosis responsible for a large proportion of ischemic strokes. Given the high burden of the disease , early diagnosis and optimal secondary prevention are essential elements in clinical practice. For a long time, the degree of stenosis had been considered the parameter to judge the severity of carotid atherosclerosis. Over the last 30 years, literature has shifted attention from stenosis to structural characteristics of atherosclerotic lesion, eventually leading to the "vulnerable plaque" model. These "vulnerable plaques" frequently demonstrate high-risk imaging features that can be assessed by various non-invasive imaging modalities.

Exploring the EVolution in PrognOstic CapabiLity of MUltisequence Cardiac MagneTIc ResOnance in PatieNts Affected by Takotsubo Cardiomyopathy Based on Machine Learning Analysis: Design and Rationale of the EVOLUTION Study

PURPOSE

Takotsubo cardiomyopathy (TTC) is a transient but severe acute myocardial dysfunction with a wide range of outcomes from favorable to life-threatening. The current risk stratification scores of TTC patients do not include cardiac magnetic resonance (CMR) parameters. To date, it is still unknown whether and how clinical, trans-thoracic echocardiography (TTE), and CMR data can be integrated to improve risk stratification.

METHODS

EVOLUTION (Exploring the eVolution in prognOstic capabiLity of mUlti-sequence cardiac magneTIc resOnance in patieNts affected by Takotsubo cardiomyopathy) is a multicenter, international registry of TTC patients who will undergo a clinical, TTE, and CMR evaluation. Clinical data including demographics, risk factors, comorbidities, laboratory values, ECG, and results from TTE and CMR analysis will be collected, and each patient will be followed-up for in-hospital and long-term outcomes. Clinical outcome measures during hospitalization will include cardiovascular death, pulmonary edema, arrhythmias, stroke, or transient ischemic attack.Clinical long-term outcome measures will include cardiovascular death, pulmonary edema, heart failure, arrhythmias, sudden cardiac death, and major adverse cardiac and cerebrovascular events defined as a composite endpoint of death from any cause, myocardial infarction, recurrence of TTC, transient ischemic attack, and stroke. We will develop a comprehensive clinical and imaging score that predicts TTC outcomes and test the value of machine learning models, incorporating clinical and imaging parameters to predict prognosis.

CONCLUSIONS

The main goal of the study is to develop a comprehensive clinical and imaging score, that includes TTE and CMR data, in a large cohort of TTC patients for risk stratification and outcome prediction as a basis for possible changes in patient management.

Does Coronary Plaque Morphology Matter Beyond Plaque Burden?

PURPOSE OF REVIEW

Imaging of adverse coronary plaque features by coronary computed tomography angiography (CCTA) has advanced greatly and at a fast pace. We aim to describe the evolution, present and future in plaque analysis, and its value in comparison to plaque burden.

RECENT FINDINGS

Recently, it has been demonstrated that in addition to plaque burden, quantitative and qualitative assessment of coronary plaque by CCTA can improve the prediction of future major adverse cardiovascular events in diverse coronary artery disease scenarios. The detection of high-risk non-obstructive coronary plaque can lead to higher use of preventive medical therapies such as statins and aspirin, help identify culprit plaque, and differentiate between myocardial infarction types. Even more, over traditional plaque burden, plaque analysis including pericoronary inflammation can potentially be useful tools for tracking disease progression and response to medical therapy. The identification of the higher risk phenotypes with plaque burden, plaque characteristics, or ideally both can allow the allocation of targeted therapies and potentially monitor response. Further observational data are now required to investigate these key issues in diverse populations, followed by rigorous randomized controlled trials.

Machine learning approach in diagnosing Takotsubo cardiomyopathy: The role of the combined evaluation of atrial and ventricular strain, and parametric mapping

BACKGROUND

Cardiac magnetic resonance (CMR) with late gadolinium enhancement (LGE) is a key diagnostic tool in the differential diagnosis between non-ischemic cause of cardiac chest pain. Some patients are not eligible for a gadolinium contrast-enhanced CMR; in this scenario, the diagnosis remains challenging without invasive examination. Our purpose was to derive a machine learning model integrating some non-contrast CMR parameters and demographic factors to identify Takotsubo cardiomyopathy (TTC) in subjects with cardiac chest pain.

MATERIAL AND METHODS

Three groups of patients were retrospectively studied: TTC, acute myocarditis, and healthy controls. Global and regional left ventricular longitudinal, circumferential, and radial strain (RS) analysis included were assessed. Reservoir, conduit, and booster bi-atrial functions were evaluated by tissue-tracking. Parametric mapping values were also assessed in all the patients. Five different tree-based ensemble learning algorithms were tested concerning their ability in recognizing TTC in a fully cross-validated framework.

RESULTS

The CMR-based machine learning (ML) ensemble model, by using the Extremely Randomized Trees algorithm with Elastic Net feature selection, showed a sensitivity of 92% (95% CI 78-100), specificity of 86% (95% CI 80-92) and area under the ROC of 0.94 (95% CI 0.90-0.99) in diagnosing TTC. Among non-contrast CMR parameters, the Shapley additive explanations analysis revealed that left atrial (LA) strain and strain rate were the top imaging markers in identifying TTC patients.

CONCLUSIONS

Our study demonstrated that using a tree-based ensemble learning algorithm on non-contrast CMR parameters and demographic factors enables the identification of subjects with TTC with good diagnostic accuracy.

TRANSLATIONAL OUTLOOK

Our results suggest that non-contrast CMR features can be implemented in a ML model to accurately identify TTC subjects. This model could be a valuable tool for aiding in the diagnosis of subjects with a contraindication to the contrast media. Furthermore, the left atrial conduit strain and strain rate were imaging markers that had a strong impact on TTC identification. Further prospective and longitudinal studies are needed to validate these findings and assess predictive performance in different cohorts, such as those with different ethnicities, and social backgrounds and undergoing different treatments.

Application of AI in cardiovascular multimodality imaging

Technical advances in artificial intelligence (AI) in cardiac imaging are rapidly improving the reproducibility of this approach and the possibility to reduce time necessary to generate a report. In cardiac computed tomography angiography (CCTA) the main application of AI in clinical practice is focused on detection of stenosis, characterization of coronary plaques, and detection of myocardial ischemia. In cardiac magnetic resonance (CMR) the application of AI is focused on post-processing and particularly on the segmentation of cardiac chambers during late gadolinium enhancement. In echocardiography, the application of AI is focused on segmentation of cardiac chambers and is helpful for valvular function and wall motion abnormalities. The common thread represented by all of these techniques aims to shorten the time of interpretation without loss of information compared to the standard approach. In this review we provide an overview of AI applications in multimodality cardiac imaging.

Role of cardiac MRI in the diagnosis of immune checkpoint inhibitor-associated myocarditis

Immune Checkpoint Inhibitor (ICI)-induced cardiotoxicity is a rare immune-related adverse event (irAE) characterized by a high mortality rate. From a pathological point of view, this condition can result from a series of causes, including binding of ICIs to target molecules on non-lymphocytic cells, cross-reaction of T lymphocytes against tumor antigens with off-target tissues, generation of autoantibodies, and production of pro-inflammatory cytokines. The diagnosis of ICI-induced cardiotoxicity can be challenging, and cardiac magnetic resonance (CMR) represents the diagnostic tool of choice in clinically stable patients with suspected myocarditis. CMR is gaining a central role in diagnosis and monitoring of cardiovascular damage in cancer patients, and it is entering international cardiology and oncology guidelines. In this narrative review, we summarized the clinical aspects of ICI-associated myocarditis, highlighting its radiological aspects and proposing a novel algorithm for the use of CMR.

Artificial Intelligence in Coronary CT Angiography: Current Status and Future Prospects

Coronary heart disease (CHD) is the leading cause of mortality in the world. Early detection and treatment of CHD are crucial. Currently, coronary CT angiography (CCTA) has been the prior choice for CHD screening and diagnosis, but it cannot meet the clinical needs in terms of examination quality, the accuracy of reporting, and the accuracy of prognosis analysis. In recent years, artificial intelligence (AI) has developed rapidly in the field of medicine; it played a key role in auxiliary diagnosis, disease mechanism analysis, and prognosis assessment, including a series of studies related to CHD. In this article, the application and research status of AI in CCTA were summarized and the prospects of this field were also described.

Multimodality Imaging in Ischemic Chronic Cardiomyopathy

Ischemic chronic cardiomyopathy (ICC) is still one of the most common cardiac diseases leading to the development of myocardial ischemia, infarction, or heart failure. The application of several imaging modalities can provide information regarding coronary anatomy, coronary artery disease, myocardial ischemia and tissue characterization. In particular, coronary computed tomography angiography (CCTA) can provide information regarding coronary plaque stenosis, its composition, and the possible evaluation of myocardial ischemia using fractional flow reserve CT or CT perfusion. Cardiac magnetic resonance (CMR) can be used to evaluate cardiac function as well as the presence of ischemia. In addition, CMR can be used to characterize the myocardial tissue of hibernated or infarcted myocardium. Echocardiography is the most widely used technique to achieve information regarding function and myocardial wall motion abnormalities during myocardial ischemia. Nuclear medicine can be used to evaluate perfusion in both qualitative and quantitative assessment. In this review we aim to provide an overview regarding the different noninvasive imaging techniques for the evaluation of ICC, providing information ranging from the anatomical assessment of coronary artery arteries to the assessment of ischemic myocardium and myocardial infarction. In particular this review is going to show the different noninvasive approaches based on the specific clinical history of patients with ICC.

Long-COVID diagnosis: from diagnostic to advanced AI-driven models

SARS-COV 2 is recognized to be responsible for a multi-organ syndrome. In most patients, symptoms are mild. However, in certain subjects, COVID-19 tends to progress more severely.

Most of the patients infected with SARS-COV2 fully recovered within some weeks. In a considerable number of patients, like many other viral infections, various long-lasting symptoms have been described, now defined as “long COVID-19 syndrome”.

Given the high number of contagious over the world, it is necessary to understand and comprehend this emerging pathology to enable early diagnosis and improve patents outcomes.

In this scenario, AI-based models can be applied in long-COVID-19 patients to assist clinicians and at the same time, to reduce the considerable impact on the care and rehabilitation unit.

The purpose of this manuscript is to review different aspects of long-COVID-19 syndrome from clinical presentation to diagnosis, highlighting the considerable impact that AI can have.

Automated deep learning-based paradigm for high-risk plaque detection in B-mode common carotid ultrasound scans: an asymptomatic Japanese cohort study

BACKGROUND

The death due to stroke is caused by embolism of the arteries which is due to the rupture of the atherosclerotic lesions in carotid arteries. The lesion formation is over time, and thus, early screening is recommended for asymptomatic and moderate-risk patients. The previous techniques adopted conventional methods or semi-automated and, more recently, machine learning solutions. A handful of studies have emerged based on solo deep learning (SDL) models such as UNet architecture.

METHODS

The proposed research is the first to adopt hybrid deep learning (HDL) artificial intelligence models such as SegNet-UNet. This model is benchmarked against UNet and advanced conventional models using scale-space such as AtheroEdge 2.0 (AtheroPoint, CA, USA). All our resultant statistics of the three systems were in the order of UNet, SegNet-UNet, and AtheroEdge 2.0.

RESULTS

Using the database of 379 ultrasound scans from a Japanese cohort of 190 patients having moderate risk and implementing the cross-validation deep learning framework, our system performance using area-under-the-curve (AUC) for UNet, SegNet-UNet, and AtheroEdge 2.0 were 0.93, 0.94, and 0.95 (p<0.001), respectively. The coefficient of correlation between the three systems and ground truth (GT) were: 0.82, 0.89, and 0.85 (p<0.001 for all three), respectively. The mean absolute area error for the three systems against manual GT was 4.07±4.70 mm2, 3.11±3.92 mm2, 3.72±4.76 mm2, respectively, proving the superior performance SegNet-UNet against UNet and AtheroEdge 2.0, respectively. Statistical tests were also conducted for their reliability and stability.

CONCLUSIONS

The proposed study demonstrates a fast, accurate, and reliable solution for early detection and quantification of plaque lesions in common carotid artery ultrasound scans. The system runs on a test US image in < 1 second, proving overall performance to be clinically reliable.

Hybrid deep learning segmentation models for atherosclerotic plaque in internal carotid artery B-mode ultrasound

The automated and accurate carotid plaque segmentation in B-mode ultrasound (US) is an essential part of stroke risk stratification. Previous segmented methods used AtheroEdge™ 2.0 (AtheroPoint™, Roseville, CA) for the common carotid artery (CCA). This study focuses on automated plaque segmentation in the internal carotid artery (ICA) using solo deep learning (SDL) and hybrid deep learning (HDL) models. The methodology consists of a novel design of 10 types of SDL/HDL models (AtheroEdge™ 3.0 systems (AtheroPoint™, Roseville, CA) with a depth of four layers each. Five of the models use cross-entropy (CE)-loss, and the other five models use Dice similarity coefficient (DSC)-loss functions derived from UNet, UNet+, SegNet, SegNet-UNet, and SegNet-UNet+. The K10 protocol (Train:Test:90%:10%) was applied for all 10 models for training and predicting (segmenting) the plaque region, which was then quantified to compute the plaque area in mm². Further, the data augmentation effect was analyzed. The database consisted of 970 ICA B-mode US scans taken from 99 moderate to high-risk patients. Using the difference area threshold of 10 mm² between ground truth (GT) and artificial intelligence (AI), the area under the curve (AUC) values were 0.91, 0.911, 0.908, 0.905, and 0.898, all with a p-value of <0.001 (for CE-loss models) and 0.883, 0.889, 0.905, 0.889, and 0.907, all with a p-value of <0.001 (for DSC-loss models). The correlations between the AI-based plaque area and GT plaque area were 0.98, 0.96, 0.97, 0.98, and 0.97, all with a p-value of <0.001 (for CE-loss models) and 0.98, 0.98, 0.97, 0.98, and 0.98 (for DSC-loss models). Overall, the online system performs plaque segmentation in less than 1 s. We validate our hypothesis that HDL and SDL models demonstrate comparable performance. SegNet-UNet was the best-performing hybrid architecture.