2018/11/19 by Kahkashan Afrin, Parikshit Verma, Afrin, Kahkashan +6
Computer Science · Engineering · Mathematics · Medicine · Neuroscience · #Artificial intelligence #Benchmark (surveying) #Cardiac monitoring #Cardiology #Cartography #Computer science #ECG Monitoring and Analysis #EEG and Brain-Computer Interfaces #Embedded system #FOS: Computer and information sciences #FOS: Electrical engineering #Lead (geology) #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Medicine #Mobile device #Signal Processing (eess.SP) #Wearable computer #cs.LG #eess.SP #electronic engineering #information engineering #stat.ML
paper · pdf · doi:10.48550/arxiv.1811.08035
openalex publication_date 2018/11/19 · arxiv created 2018/11/20 · arxiv updated 2018/11/21 · openalex created_date 2022/08/02 · openalex updated_date 2026/08/06
Recent introduction of wearable single-lead ECG devices of diverse\nconfigurations has caught the intrigue of the medical community. While these\ndevices provide a highly affordable support tool for the caregivers for\ncontinuous monitoring and to detect acute conditions, such as arrhythmia, their\nutility for cardiac diagnostics remains limited. This is because clinical\ndiagnosis of many cardiac pathologies is rooted in gleaning patterns from\nsynchronous 12-lead ECG. If synchronous 12-lead signals of clinical quality can\nbe synthesized from these single-lead devices, it can transform cardiac care by\nsubstantially reducing the costs and enhancing access to cardiac diagnostics.\nHowever, prior attempts to synthesize synchronous 12-lead ECG have not been\nsuccessful. Vectorcardiography (VCG) analysis suggests that cardiac axis\nsynthesized from earlier attempts deviates significantly from that estimated\nfrom 12-lead and/or Frank lead measurements. This work is perhaps the first\nsuccessful attempt to synthesize clinically equivalent synchronous 12-lead ECG\nfrom single-lead ECG. Our method employs a random forest machine learning model\nthat uses a subject's historical 12-lead recordings to estimate the morphology\nincluding the actual timing of various ECG events (relative to the measured\nsingle-lead ECG) for all 11 missing leads of the subject. Our method was\nvalidated on two benchmark datasets as well as paper ECG and AliveCor-Kardia\ndata obtained from the Heart, Artery, and Vein Center of Fresno, California.\nResults suggest that this approach can synthesize synchronous ECG with\naccuracies (R2) exceeding 90%. Accurate synthesis of 12-lead ECG from a\nsingle-lead device can ultimately enable its wider application and improved\npoint-of-care (POC) diagnostics.\n