<i>Performance of Artificial Intelligence-Driven Predictive Modeling for Sudden Cardiac De

A 2026 protocol for a systematic review evaluating the performance of AI models in predicting sudden cardiac death and ventricular arrhythmia in patients with non-ischemic cardiomyopathy. The review, authored by Rashi Bilgaiyan, will synthesize global research from medical and technical databases to assess diagnostic accuracy, prognostic performance, and clinical readiness of AI architectures like Convolutional Neural Networks and Random Forests.

Use Cases

• Benchmarking AI model performance (e.g., sensitivity, specificity) for cardiac risk prediction based on the review's synthesis of global research.
• Evaluating the implementation of explainable AI (XAI) techniques in clinical models based on the review's assessment of model interpretability.
• Identifying effective data modalities (e.g., ECG patterns, cardiac MRI) for arrhythmia risk modeling based on the review's analysis of input data types.
• Informing clinical guidelines for cardioverter-defibrillator implantation by assessing AI models as an alternative to traditional LVEF-based decision-making.

Strengths

• Protocol is licensed under CC-BY-4.0, allowing for broad reuse and adaptation.
• The review methodology includes independent data extraction by two reviewers to minimize bias.
• Focus is on a specific clinical population (non-ischemic cardiomyopathy) with clearly defined clinical endpoints for validation.

Limitations

• The dataset is a review protocol, not the underlying primary data; actual data quality and scale depend on the included studies.
• Column-level documentation for any extracted data is absent; field semantics must be inferred from the published review.
• Row count and specific file formats for the synthesized data are unknown.

Provenance

Source

figshare, authored by rashi bilgaiyan.

Collection Method

Describes a planned systematic review of literature from databases including PubMed, Cochrane Library, EMBASE, and IEEE Xplore.

Freshness

Last updated 2026-05-13 10:34:29.