Machine learning with convolutional neural networks for clinical cardiologists

Patisiran for the Treatment of Transthyretin-mediated Amyloidosis with Cardiomyopathy

Transthyretin (TTR) is a tetrameric protein, synthesized primarily by the liver, that acts as a physiological transport protein for retinol and thyroxine. TTR can misfold into pathogenic amyloid fibrils that deposit in the heart and nerves, causing a life-threatening transthyretin amyloidosis cardiomyopathy (ATTR-CM), and a progressive and debilitating polyneuropathy (ATTR-PN). Recent therapeutic advances have resulted in the development of drugs that reduce TTR production. Patisiran is a small interfering RNA that disrupts the complimentary mRNA and inhibits TTR synthesis, and is the first gene-silencing medication licensed for the treatment of ATTR amyloidosis. After encouraging results following the use of patisiran for the treatment of patients with ATTR-PN, there has been increasing interest in the use of patisiran for the treatment of ATTR-CM. Various studies have demonstrated improvements across a wide range of cardiac biomarkers following treatment with patisiran, and have changed the perception of ATTR-CM from being thought of as a terminal disease process, to now being regarded as a treatable disease. These successes represent a huge milestone and have the potential to revolutionize the landscape of treatment for ATTR-CM. However, the long-term safety of patisiran and how best to monitor cardiac response to treatment remain to be determined.

Artificial intelligence-enabled electrocardiogram to distinguish atrioventricular re-entrant tachycardia from atrioventricular nodal re-entrant tachycardia

Background

Accurately determining arrhythmia mechanism from a 12-lead electrocardiogram (ECG) of supraventricular tachycardia can be challenging. We hypothesized a convolutional neural network (CNN) can be trained to classify atrioventricular re-entrant tachycardia (AVRT) vs atrioventricular nodal re-entrant tachycardia (AVNRT) from the 12-lead ECG, when using findings from the invasive electrophysiology (EP) study as the gold standard.

Methods

We trained a CNN on data from 124 patients undergoing EP studies with a final diagnosis of AVRT or AVNRT. A total of 4962 5-second 12-lead ECG segments were used for training. Each case was labeled AVRT or AVNRT based on the findings of the EP study. The model performance was evaluated against a hold-out test set of 31 patients and compared to an existing manual algorithm.

Results

The model had an accuracy of 77.4% in distinguishing between AVRT and AVNRT. The area under the receiver operating characteristic curve was 0.80. In comparison, the existing manual algorithm achieved an accuracy of 67.7% on the same test set. Saliency mapping demonstrated the network used the expected sections of the ECGs for diagnoses; these were the QRS complexes that may contain retrograde P waves.

Conclusion

We describe the first neural network trained to differentiate AVRT from AVNRT. Accurate diagnosis of arrhythmia mechanism from a 12-lead ECG could aid preprocedural counseling, consent, and procedure planning. The current accuracy from our neural network is modest but may be improved with a larger training dataset.

RNA Targeting and Gene Editing Strategies for Transthyretin Amyloidosis

Transthyretin (TTR) is a tetrameric protein synthesized primarily by the liver. TTR can misfold into pathogenic ATTR amyloid fibrils that deposit in the nerves and heart, causing a progressive and debilitating polyneuropathy (PN) and life-threatening cardiomyopathy (CM). Therapeutic strategies, which are aimed at reducing ongoing ATTR amyloid fibrillogenesis, include stabilization of the circulating TTR tetramer or reduction of TTR synthesis. Small interfering RNA (siRNA) or antisense oligonucleotide (ASO) drugs are highly effective at disrupting the complementary mRNA and inhibiting TTR synthesis. Since their development, patisiran (siRNA), vutrisiran (siRNA) and inotersen (ASO) have all been licensed for treatment of ATTR-PN, and early data suggest these drugs may have efficacy in treating ATTR-CM. An ongoing phase 3 clinical trial will evaluate the efficacy of eplontersen (ASO) in the treatment of both ATTR-PN and ATTR-CM, and a recent phase 1 trial demonstrated the safety of novel in vivo CRISPR-Cas9 gene-editing therapy in patients with ATTR amyloidosis. Recent results from trials of gene silencer and gene-editing therapies suggest these novel therapeutic agents have the potential to substantially alter the landscape of treatment for ATTR amyloidosis. Their success has already changed the perception of ATTR amyloidosis from a universally progressive and fatal disease to one that is treatable through availability of highly specific and effective disease-modifying therapies. However, important questions remain including long-term safety of these drugs, potential for off-target gene editing, and how best to monitor the cardiac response to treatment.Kindly check and confirm the processed running title.This is correct.

Artificial intelligence-enabled electrocardiogram to distinguish cavotricuspid isthmus dependence from other atrial tachycardia mechanisms. EUROPEAN HEART JOURNAL. DIGITAL HEALTH 2022;3:405-414. [PMID: 36712163 PMCID: PMC9708023 DOI: 10.1093/ehjdh/ztac042]

Aims

Accurately determining atrial arrhythmia mechanisms from a 12-lead electrocardiogram (ECG) can be challenging. Given the high success rate of cavotricuspid isthmus (CTI) ablation, identification of CTI-dependent typical atrial flutter (AFL) is important for treatment decisions and procedure planning. We sought to train a convolutional neural network (CNN) to classify CTI-dependent AFL vs. non-CTI dependent atrial tachycardia (AT), using data from the invasive electrophysiology (EP) study as the gold standard.

Methods and results

We trained a CNN on data from 231 patients undergoing EP studies for atrial tachyarrhythmia. A total of 13 500 five-second 12-lead ECG segments were used for training. Each case was labelled CTI-dependent AFL or non-CTI-dependent AT based on the findings of the EP study. The model performance was evaluated against a test set of 57 patients. A survey of electrophysiologists in Europe was undertaken on the same 57 ECGs. The model had an accuracy of 86% (95% CI 0.77-0.95) compared to median expert electrophysiologist accuracy of 79% (range 70-84%). In the two thirds of test set cases (38/57) where both the model and electrophysiologist consensus were in agreement, the prediction accuracy was 100%. Saliency mapping demonstrated atrial activation was the most important segment of the ECG for determining model output.

Conclusion

We describe the first CNN trained to differentiate CTI-dependent AFL from other AT using the ECG. Our model matched and complemented expert electrophysiologist performance. Automated artificial intelligence-enhanced ECG analysis could help guide treatment decisions and plan ablation procedures for patients with organized atrial arrhythmias.

Imaging-Guided Treatment for Cardiac Amyloidosis

Purpose of Review

This review will explore the role of cardiac imaging in guiding treatment in the two most commonly encountered subtypes of cardiac amyloidosis (immunoglobulin light-chain amyloidosis [AL] and transthyretin amyloidosis [ATTR]).

Recent Findings

Advances in multi-parametric cardiac imaging involving a combination of bone scintigraphy, echocardiography and cardiac magnetic resonance imaging have resulted in earlier diagnosis and initiation of treatment, while the evolution of techniques such as longitudinal strain and extracellular volume quantification allow clinicians to track individuals’ response to treatment. Imaging developments have led to a deeper understanding of the disease process and treatment mechanisms, which in combination result in improved patient outcomes.

Summary

The rapidly expanding treatment regimens for cardiac amyloidosis have led to an even greater reliance on cardiac imaging to help establish an accurate diagnosis, monitor treatment response and aid the adjustment of treatment strategies accordingly.

Current State and Future Perspectives of Artificial Intelligence for Automated Coronary Angiography Imaging Analysis in Patients with Ischemic Heart Disease

Purpose of Review

Artificial intelligence (AI) applications in (interventional) cardiology continue to emerge. This review summarizes the current state and future perspectives of AI for automated imaging analysis in invasive coronary angiography (ICA).

Recent Findings

Recently, 12 studies on AI for automated imaging analysis In ICA have been published. In these studies, machine learning (ML) models have been developed for frame selection, segmentation, lesion assessment, and functional assessment of coronary flow. These ML models have been developed on monocenter datasets (in range 31–14,509 patients) and showed moderate to good performance. However, only three ML models were externally validated.

Summary

Given the current pace of AI developments for the analysis of ICA, less-invasive, objective, and automated diagnosis of CAD can be expected in the near future. Further research on this technology in the catheterization laboratory may assist and improve treatment allocation, risk stratification, and cath lab logistics by integrating ICA analysis with other clinical characteristics.