Key result
AI CVD detection achieves high accuracy, but only ~23% of systems undergo external validation.
Why the study?
Artificial intelligence integrated with physiological signals offers pathways for scalable cardiovascular assessment, yet a fragmented evidence base has impeded the identification of validated approaches and clinical translation pathways.
Does artificial intelligence integrated with physiological signals (ECG, PPG, PCG) improve the detection and clinical translation of cardiovascular disease assessment across 35 peer-reviewed studies?
Systematic Review
Does artificial intelligence integrated with physiological signals (ECG, PPG, PCG) improve the detection and clinical translation of cardiovascular disease assessment across 35 peer-reviewed studies?
While AI-driven physiological signal analysis shows high reported accuracy (unweighted mean 93.7%) for cardiovascular disease detection, clinical translation is severely limited by a lack of external validation (23%), clinical deployment (14%), and multi-signal fusion.
Answers are grounded in this paper and cite their sources.
High reported AI accuracies for CVD detection should not yet change clinical practice; leaves open the need for external validation and prospective deployment trials.
Cardiovascular disease exacts an extraordinary global toll, with annual mortality exceeding 17.9 million lives and representing nearly one-third of all deaths worldwide. Reducing diagnostic delays, particularly in resource-limited settings, requires innovative, scalable, and cost-effective solutions. Artificial intelligence integrated with physiological signals, such as electrocardiogram (ECG), photoplethysmogram (PPG) and phonocardiogram (PCG), offers promising pathways for scalable, real-time cardiovascular assessment. Despite rapid advances, the evidence base remains fragmented, impeding identification of validated approaches and clinical translation pathways. A PRISMA-guided search across five databases (Scopus, Web of Science, PubMed, Embase, IEEE Xplore) identified 35 eligible peer-reviewed studies. Quality and risk of bias were appraised using a QUADAS-2 adapted framework. Most systems utilised ECG (71.4%) with deep learning architectures, including convolutional and recurrent neural networks. Critical gaps were quantified: only 23% of studies conducted external validation, only 14% achieved clinical-scale deployment, no study integrated principled multi-signal fusion (ECG + PPG + PCG), and only one system (1/35, 2.9%) implemented a patient-facing conversational interface. Reported accuracies ranged from 70.5 to 99.9% (n = 21 reporting; unweighted mean 93.7%), but inconsistent metric reporting and limited external validation tempered translational confidence. A three-directional framework is proposed, addressing generalisability through multi-signal (ECG + PPG + PCG) models, interpretability through LLM-driven clinical narratives, and accessibility through edge-cloud deployment, to guide future research toward clinically deployable systems.
No takes yet. Share an insight, caveat, or question.
Alam et al. (2026) conducted a systematic review in Cardiovascular disease. AI-driven software systems and mobile applications using physiological signals (ECG, PPG, PCG) was evaluated on Diagnostic classification accuracy and clinical deployment readiness across 30 included studies. AI-driven CVD detection systems achieved 70.5% to 99.9% accuracy across 30 studies, though only 23% performed external validation and 10% reached clinical deployment.
Synapse has enriched 10 related papers. Consider them for comparative context: