GeneAI 3.0: powerful, novel, generalized hybrid and ensemble deep learning frameworks for miRNA species classification of stationary patterns from nucleotides

Due to the intricate relationship between the small non-coding ribonucleic acid (miRNA) sequences, the classification of miRNA species, namely Human, Gorilla, Rat, and Mouse is challenging. Previous methods are not robust and accurate. In this study, we present AtheroPoint's GeneAI 3.0, a powerful, novel, and generalized method for extracting features from the fixed patterns of purines and pyrimidines in each miRNA sequence in ensemble paradigms in machine learning (EML) and convolutional neural network (CNN)-based deep learning (EDL) frameworks. GeneAI 3.0 utilized five conventional (Entropy, Dissimilarity, Energy, Homogeneity, and Contrast), and three contemporary (Shannon entropy, Hurst exponent, Fractal dimension) features, to generate a composite feature set from given miRNA sequences which were then passed into our ML and DL classification framework. A set of 11 new classifiers was designed consisting of 5 EML and 6 EDL for binary/multiclass classification. It was benchmarked against 9 solo ML (SML), 6 solo DL (SDL), 12 hybrid DL (HDL) models, resulting in a total of 11 + 27 = 38 models were designed. Four hypotheses were formulated and validated using explainable AI (XAI) as well as reliability/statistical tests. The order of the mean performance using accuracy (ACC)/area-under-the-curve (AUC) of the 24 DL classifiers was: EDL > HDL > SDL. The mean performance of EDL models with CNN layers was superior to that without CNN layers by 0.73%/0.92%. Mean performance of EML models was superior to SML models with improvements of ACC/AUC by 6.24%/6.46%. EDL models performed significantly better than EML models, with a mean increase in ACC/AUC of 7.09%/6.96%. The GeneAI 3.0 tool produced expected XAI feature plots, and the statistical tests showed significant p-values. Ensemble models with composite features are highly effective and generalized models for effectively classifying miRNA sequences.

The Canadian Women's Heart Health Alliance ATLAS on the Epidemiology, Diagnosis, and Management of Cardiovascular Disease in Women - Chapter 8: Knowledge Gaps and Status of Existing Research Programs in Canada

Despite significant progress in medical research and public health efforts, gaps in knowledge of women's heart health remain across epidemiology, presentation, management, outcomes, education, research, and publications. Historically, heart disease was viewed primarily as a condition in men and male individuals, leading to limited understanding of the unique risks and symptoms that women experience. These knowledge gaps are particularly problematic because globally heart disease is the leading cause of death for women. Until recently, sex and gender have not been addressed in cardiovascular research, including in preclinical and clinical research. Recruitment was often limited to male participants and individuals identifying as men, and data analysis according to sex or gender was not conducted, leading to a lack of data on how treatments and interventions might affect female patients and individuals who identify as women differently. This lack of data has led to suboptimal treatment and limitations in our understanding of the underlying mechanisms of heart disease in women, and is directly related to limited awareness and knowledge gaps in professional training and public education. Women are often unaware of their risk factors for heart disease or symptoms they might experience, leading to delays in diagnosis and treatments. Additionally, health care providers might not receive adequate training to diagnose and treat heart disease in women, leading to misdiagnosis or undertreatment. Addressing these knowledge gaps requires a multipronged approach, including education and policy change, built on evidence-based research. In this chapter we review the current state of existing cardiovascular research in Canada with a specific focus on women.

Deep learning artificial intelligence framework for multiclass coronary artery disease prediction using combination of conventional risk factors, carotid ultrasound, and intraplaque neovascularization

OBJECTIVE

Cardiovascular disease (CVD) is a major healthcare challenge and therefore early risk assessment is vital. Previous assessment techniques use either "conventional CVD risk calculators (CCVRC)" or machine learning (ML) paradigms. These techniques are ad-hoc, unreliable, not fully automated, and have variabilities. We, therefore, introduce AtheroEdge-MCDLAI (AE3.0DL) windows-based platform using multiclass Deep Learning (DL) system.

METHODS

Data was collected on 500 patients having both carotid ultrasound and corresponding coronary angiography scores (CAS), measured as stenosis in coronary arteries and considered as the gold standard. A total of 39 covariates were used, clubbed into three clusters, namely (i) Office-based: age, gender, body mass index, smoker, hypertension, systolic blood pressure, and diastolic blood pressure; (ii) Laboratory-based: Hyperlipidemia, hemoglobin A1c, and estimated glomerular filtration rate; and (iii) Carotid ultrasound image phenotypes: maximum plaque height, total plaque area, and intra-plaque neovascularization. Baseline characteristics for four classes (target labels) having significant (p < 0.0001) values were calculated using Chi-square and ANOVA. For handling the cohort's imbalance in the risk classes, AE3.0DL used the synthetic minority over-sampling technique (SMOTE). AE3.0DL used Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) DL models and the performance (accuracy and area-under-the-curve) was computed using 10-fold cross-validation (90% training, 10% testing) frameworks. AE3.0DL was validated and benchmarked.

RESULTS

The AE3.0DL using RNN and LSTM showed an accuracy and AUC (p < 0.0001) pairs as (95.00% and 0.98), and (95.34% and 0.99), respectively, and showed an improvement of 32.93% and 9.94% against CCVRC and ML, respectively. AE3.0DL runs in <1 s.

CONCLUSION

DL algorithms are a powerful paradigm for coronary artery disease (CAD) risk prediction and CVD risk stratification.

Vascular imaging of atherosclerosis: Strengths and weaknesses

Atherosclerosis is an inflammatory disease that can lead to several complications such as ischemic heart disease, stroke, and peripheral vascular disease. Therefore, researchers and clinicians rely heavily on the use of imaging modalities to identify, and more recently, quantify the burden of atherosclerosis in the aorta, carotid arteries, coronary arteries, and peripheral vasculature. These imaging techniques vary in invasiveness, cost, resolution, radiation exposure, and presence of artifacts. Consequently, a detailed understanding of the risks and benefits of each technique is crucial prior to their introduction into routine cardiovascular screening. Additionally, recent research in the field of microvascular imaging has proven to be important in the field of atherosclerosis. Using techniques such as contrast-enhanced ultrasound and superb microvascular imaging, researchers have been able to detect blood vessels within a plaque lesion that may contribute to vulnerability and rupture. This paper will review the strengths and weaknesses of the various imaging techniques used to measure atherosclerotic burden. Furthermore, it will discuss the future of advanced imaging modalities as potential biomarkers for atherosclerosis.

Carotid intraplaque neovascularization predicts coronary artery disease and cardiovascular events

AIMS

It is thought that the majority of cardiovascular (CV) events are caused by vulnerable plaque. Such lesions are rupture prone, in part due to neovascularization. It is postulated that plaque vulnerability may be a systemic process and that vulnerable lesions may co-exist at multiple sites in the vascular bed. This study sought to examine whether carotid plaque vulnerability, characterized by contrast-enhanced ultrasound (CEUS)-assessed intraplaque neovascularization (IPN), was associated with significant coronary artery disease (CAD) and future CV events.

METHODS AND RESULTS

We investigated carotid IPN using carotid CEUS in 459 consecutive stable patients referred for coronary angiography. IPN was graded based on the presence and location of microbubbles within each plaque (0, not visible; 1, peri-adventitial; and 2, plaque core). The grades of each plaque were averaged to obtain an overall score per patient. Coronary plaque severity and complexity was also determined angiographically. Patients were followed for 30 days following their angiogram. This study found that a higher CEUS-assessed carotid IPN score was associated with significant CAD (≥50% stenosis) (1.8 ± 0.4 vs. 0.5 ± 0.6, P < 0.0001) and greater complexity of coronary lesions (1.7 ± 0.5 vs. 1.3 ± 0.8, P < 0.0001). Furthermore, an IPN score ≥1.25 could predict significant CAD with a high sensitivity (92%) and specificity (89%). The Kaplan-Meier analysis demonstrated a significantly higher proportion of participants having CV events with an IPN score ≥1.25 (P = 0.004).

CONCLUSION

Carotid plaque neovascularization was found to be predictive of significant and complex CAD and future CV events. CEUS-assessed carotid IPN is a clinically useful tool for CV risk stratification in high-risk cardiac patients.

The use of ultrasound to assess aortic biomechanics: Implications for aneurysm and dissection

Arterial stiffening, which occurs when conduit arteries thicken and lose elasticity, has been associated with cardiovascular disease and increased risk for future cardiovascular events. Specifically, aortic stiffening plays a large role in the pathogenesis of vascular diseases, such as aneurysm formation and dissection. Current parameters used to assess risk of aortic rupture include absolute diameter and growth rate. However, these properties lack the reliability required to accurately risk-stratify patients. As with any elastic conduit, it is important to assess the biomechanical properties of the aorta in order to assess cardiovascular risk and prevent disease progression. There are several invasive and noninvasive methods by which stiffness of the large arteries can be assessed. Of particular interest are ultrasound-based methods, such as tissue Doppler imaging and speckle-tracking echocardiography, due to their noninvasive and feasible nature. In this review, we summarize studies demonstrating utility of noninvasive ultrasound imaging methods for measuring aortic biomechanics for the assessment and management of aortic aneurysms.

Handheld versus conventional vascular ultrasound for assessing carotid artery plaque

BACKGROUND

Coronary Artery Disease (CAD) is the leading cause of death worldwide. Ultrasound-detected carotid plaque has been previously shown to predict significant CAD. Despite this evidence, carotid plaque assessment has not been implemented in cardiac risk screening, likely due to the cost associated with a formal carotid ultrasound examination. This study sought to determine whether a handheld vascular ultrasound device could be used as an accurate point-of-care imaging tool for the assessment of carotid artery plaque.

METHODS

We performed a focused vascular ultrasound of the carotid arteries of 200 patients referred for coronary angiography using a handheld ultrasound device as well as a full-size conventional ultrasound system. For each participant, the maximum plaque height (MPH) and total plaque area (TPA) of the carotid artery bulbs were measured.

RESULTS

Carotid plaque assessment using the handheld device was comparable to that of the conventional ultrasound system. We found a good correlation and no relevant bias between handheld and conventional ultrasound systems in measuring MPH (r = 0.84, p < 0.0001) and TPA (r = 0.94, p < 0.0001). Furthermore, there was good inter-rater reliability for online and offline measurements of MPH and offline measurements of TPA by handheld ultrasound (0.79, 0.76 and 0.85, respectively).

CONCLUSIONS

This study demonstrates that a focused ultrasound of the carotid artery using a handheld device can be used to accurately measure MPH and TPA. This protocol has the potential to provide an expedited point-of-care assessment of carotid plaque.

Parents perspectives on whole genome sequencing for their children: qualified enthusiasm?

OBJECTIVE

To better understand the consequences of returning whole genome sequencing (WGS) results in paediatrics and facilitate its evidence-based clinical implementation, we studied parents' experiences with WGS and their preferences for the return of adult-onset secondary variants (SVs)-medically actionable genomic variants unrelated to their child's current medical condition that predict adult-onset disease.

METHODS

We conducted qualitative interviews with parents whose children were undergoing WGS as part of the SickKids Genome Clinic, a research project that studies the impact of clinical WGS on patients, families, and the healthcare system. Interviews probed parents' experience with and motivation for WGS as well as their preferences related to SVs. Interviews were analysed thematically.

RESULTS

Of 83 invited, 23 parents from 18 families participated. These parents supported WGS as a diagnostic test, perceiving clear intrinsic and instrumental value. However, many parents were ambivalent about receiving SVs, conveying a sense of self-imposed obligation to take on the 'weight' of knowing their child's SVs, however unpleasant. Some parents chose to learn about adult-onset SVs for their child but not for themselves.

CONCLUSIONS

Despite general enthusiasm for WGS as a diagnostic test, many parents felt a duty to learn adult-onset SVs. Analogous to 'inflicted insight', we call this phenomenon 'inflicted ought'. Importantly, not all parents of children undergoing WGS view the best interests of their child in relational terms, thereby challenging an underlying justification for current ACMG guidelines for reporting incidental secondary findings from whole exome and WGS.

The essential autophagy gene ATG7 modulates organ fibrosis via regulation of endothelial-to-mesenchymal transition

Pulmonary fibrosis is a progressive disease characterized by fibroblast proliferation and excess deposition of collagen and other extracellular matrix components. Although the origin of fibroblasts is multifactorial, recent data implicate endothelial-to-mesenchymal transition as an important source of fibroblasts. We report herein that loss of the essential autophagy gene ATG7 in endothelial cells (ECs) leads to impaired autophagic flux accompanied by marked changes in EC architecture, loss of endothelial, and gain of mesenchymal markers consistent with endothelial-to-mesenchymal transition. Loss of ATG7 also up-regulates TGFβ signaling and key pro-fibrotic genes in vitro. In vivo, EC-specific ATG7 knock-out mice exhibit a basal reduction in endothelial-specific markers and demonstrate an increased susceptibility to bleomycin-induced pulmonary fibrosis and collagen accumulation. Our findings help define the role of endothelial autophagy as a potential therapeutic target to limit organ fibrosis, a condition for which presently there are no effective available treatments.