{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,2]],"date-time":"2026-10-02T15:13:07Z","timestamp":1790953987759,"version":"4.1.0"},"reference-count":29,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JMIR Form Res"],"abstract":"Abstract<\/jats:title>\n \n Background<\/jats:title>\n Electrocardiograms (ECGs) are commonly stored in PDF, particularly as vector-based files generated by ECG management systems. Previous studies have demonstrated that ECG signals can be extracted through PDF-to-Scalable Vector Graphics (SVG) conversion, highlighting the potential to reconstruct waveform signals from vector graphics. These reconstructed signals further enable the derivation and prediction of clinically relevant ECG parameters.<\/jats:p>\n <\/jats:sec>\n \n Objective<\/jats:title>\n This study aimed to develop an integrated framework for direct reconstruction of high-fidelity ECG signals from vector-based PDF files and for simultaneous estimation of multiple clinically relevant ECG parameters using deep learning.<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n In this retrospective methodological study, 50,000 twelve-lead ECG PDFs generated by a MUSE system (2015\u20102024) were analyzed. A direct PDF parsing pipeline was developed to extract vector path objects and reconstruct time-series signals without intermediate format conversion. Reconstruction accuracy was evaluated against the original system-exported signals. Two deep learning models, DualECGFormer and DualResNetECG, were developed to estimate 8 specific ECG parameters. The reference standards for these parameters\u2014including ventricular rate; PR interval; QRS duration; QT interval; corrected QT interval (QTc); and the electrical axes of the P wave, QRS complex, and T wave\u2014were derived from machine-generated values within the MUSE system database. Performance was compared with a rule-based approach (NeuroKit2).<\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n \n The proposed method achieved high reconstruction fidelity, with mean absolute errors (MAEs) below 1\u00d710\n \u20133<\/jats:sup>\n mV across all leads. Compared with an SVG-based workflow, the direct parsing approach reduced processing time by approximately 4.6-fold. For parameter estimation, deep learning models outperformed the rule-based method for most parameters. DualResNetECG achieved the best overall performance, with MAEs of 1.11 bpm for ventricular rate, 7.27 milliseconds for PR interval, 9.89 milliseconds for QRS duration, 11.87 milliseconds for QT, and 13.71 milliseconds for QTc. For electrical axis estimation, MAEs ranged from 8.0\u00b0 to 16.31\u00b0. The model also demonstrated reliable detection of physiologically undefined parameters (PR interval and P-wave axis), achieving an area under the receiver operating characteristic curve of up to 0.978.\n <\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n This study presents an efficient and scalable framework for direct extraction of ECG signals from MUSE-generated vector-based PDFs and integrated multiparameter estimation using deep learning. The approach achieves high reconstruction accuracy and competitive predictive performance, supporting its potential utility for large-scale retrospective MUSE ECG analysis.<\/jats:p>\n <\/jats:sec>","DOI":"10.2196\/80597","type":"journal-article","created":{"date-parts":[[2026,5,15]],"date-time":"2026-05-15T00:25:06Z","timestamp":1778804706000},"page":"e80597-e80597","source":"Crossref","is-referenced-by-count":1,"title":["Direct Reconstruction of High-Fidelity Electrocardiogram Signals From Vector-Based PDF Files With Integrated Deep Learning for Multiparameter Estimation: Retrospective Methodological Study"],"prefix":"10.2196","volume":"10","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6891-5205","authenticated-orcid":false,"given":"Shian-Sen","family":"Shie","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4314-6021","authenticated-orcid":false,"given":"Po-Yen","family":"Huang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0056-7985","authenticated-orcid":false,"given":"Ming-Shien","family":"Wen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1546-7298","authenticated-orcid":false,"given":"Chao-Yung","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2026,7,24]]},"reference":[{"issue":"6","key":"R1","doi-asserted-by":"publisher","first-page":"776","DOI":"10.1016\/j.jelectrocard.2017.08.001","article-title":"PDF-ECG in clinical practice: a model for long-term preservation of digital 12-lead ECG data","volume":"50","author":"Sassi","journal-title":"J Electrocardiol"},{"key":"R2","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2020.104077","article-title":"Deep learning for digitizing highly noisy paper-based ECG records","volume":"127","author":"Li","journal-title":"Comput Biol Med"},{"key":"R3","doi-asserted-by":"publisher","DOI":"10.1016\/j.cmpb.2022.106890","article-title":"Digitizing ECG image: a new method and open-source software code","volume":"221","author":"Fortune","journal-title":"Comput Methods Programs Biomed"},{"key":"R4","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-022-25284-1","article-title":"A fully-automated paper ECG digitisation algorithm using deep learning","volume":"12","author":"Wu","journal-title":"Sci Rep"},{"key":"R5","doi-asserted-by":"publisher","DOI":"10.1016\/j.jelectrocard.2023.05.009","article-title":"Automatic digitization of paper electrocardiograms - a systematic review","volume":"80","author":"Lence","journal-title":"J Electrocardiol"},{"issue":"17","key":"R6","doi-asserted-by":"publisher","first-page":"1910","DOI":"10.3390\/diagnostics14171910","article-title":"Paper-recorded ECG digitization method with automatic reference voltage selection for telemonitoring and diagnosis","volume":"14","author":"Wang","journal-title":"Diagnostics (Basel)"},{"issue":"19","key":"R7","doi-asserted-by":"publisher","first-page":"7138","DOI":"10.3390\/s22197138","article-title":"Development and validation of an algorithm for the digitization of ECG paper images","volume":"22","author":"Randazzo","journal-title":"Sensors (Basel)"},{"issue":"1","key":"R8","doi-asserted-by":"publisher","first-page":"284","DOI":"10.1016\/j.cmpb.2013.09.014","article-title":"Raw data extraction from electrocardiograms with portable document format","volume":"113","author":"Ortigosa","journal-title":"Comput Methods Programs Biomed"},{"issue":"4","key":"R9","doi-asserted-by":"publisher","first-page":"1689","DOI":"10.3758\/s13428-020-01516-y","article-title":"NeuroKit2: a Python toolbox for neurophysiological signal processing","volume":"53","author":"Makowski","journal-title":"Behav Res Methods"},{"issue":"8","key":"R10","doi-asserted-by":"publisher","DOI":"10.2196\/38454","article-title":"State-of-the-art deep learning methods on electrocardiogram data: systematic review","volume":"10","author":"Petmezas","journal-title":"JMIR Med Inform"},{"issue":"1","key":"R11","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-020-79512-7","article-title":"Delineation of the electrocardiogram with a mixed-quality-annotations dataset using convolutional neural networks","volume":"11","author":"Jimenez-Perez","journal-title":"Sci Rep"},{"key":"R12","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2020.113911","article-title":"DENS-ECG: a deep learning approach for ECG signal delineation","volume":"165","author":"Peimankar","journal-title":"Expert Syst Appl"},{"key":"R13","doi-asserted-by":"publisher","DOI":"10.1109\/EMBC40787.2023.10341204","article-title":"QTNet: deep learning for estimating QT intervals using a single lead ECG","volume":"2023","author":"Alam","journal-title":"Annu Int Conf IEEE Eng Med Biol Soc"},{"issue":"12","key":"R14","doi-asserted-by":"publisher","first-page":"euaf274","DOI":"10.1093\/europace\/euaf274","article-title":"QTcNet: a deep learning model for direct heart rate corrected QT interval estimation","volume":"27","author":"Plagwitz","journal-title":"Europace"},{"key":"R15","unstructured":"Uzi-Granot\/PdfFileAnalyzer. GitHub. URL: https:\/\/github.com\/Uzi-Granot\/PdfFileAnalyzer [Accessed 09-07-2026]"},{"issue":"3","key":"R16","doi-asserted-by":"publisher","first-page":"230","DOI":"10.1109\/TBME.1985.325532","article-title":"A real-time QRS detection algorithm","volume":"32","author":"Pan","journal-title":"IEEE Trans Biomed Eng"},{"key":"R17","doi-asserted-by":"publisher","DOI":"10.1109\/JTEHM.2018.2878000","article-title":"An open-source feature extraction tool for the analysis of peripheral physiological data","volume":"6","author":"Nabian","journal-title":"IEEE J Transl Eng Health Med"},{"key":"R18","unstructured":"Vaswani A Shazeer N Parmar N et al. Attention is all you need. Presented at: 31st Conference on Neural Information Processing Systems; Dec 4-9, 2017; Long Beach, CA. URL: https:\/\/proceedings.neurips.cc\/paper_files\/paper\/2017\/file\/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf [Accessed 09-07-2026]"},{"key":"R19","doi-asserted-by":"crossref","unstructured":"He K Zhang X Ren S Sun J . Deep residual learning for image recognition. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE; 2016. [doi: 10.1109\/CVPR.2016.90]","DOI":"10.1109\/CVPR.2016.90"},{"issue":"8","key":"R20","doi-asserted-by":"publisher","first-page":"2011","DOI":"10.1109\/TPAMI.2019.2913372","article-title":"Squeeze-and-excitation networks","volume":"42","author":"Hu","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"issue":"11","key":"R21","doi-asserted-by":"publisher","first-page":"1003","DOI":"10.1016\/j.jacc.2008.12.016","article-title":"AHA\/ACCF\/HRS recommendations for the standardization and interpretation of the electrocardiogram: part VI: acute ischemia\/infarction: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society. Endorsed by the International Society for Computerized Electrocardiology","volume":"53","author":"Wagner","journal-title":"J Am Coll Cardiol"},{"issue":"9","key":"R22","doi-asserted-by":"publisher","DOI":"10.3390\/diagnostics11091678","article-title":"Potential of rule-based methods and deep learning architectures for ECG diagnostics","volume":"11","author":"Bortolan","journal-title":"Diagnostics (Basel)"},{"issue":"1","key":"R23","doi-asserted-by":"publisher","first-page":"65","DOI":"10.1038\/s41591-018-0268-3","article-title":"Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network","volume":"25","author":"Hannun","journal-title":"Nat Med"},{"issue":"1","key":"R24","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-020-15432-4","article-title":"Automatic diagnosis of the 12-lead ECG using a deep neural network","volume":"11","author":"Ribeiro","journal-title":"Nat Commun"},{"key":"R25","doi-asserted-by":"crossref","unstructured":"Ji J Chen X Luo C Li P . A deep multi-task learning approach for ECG data analysis. In: 2018 IEEE EMBS International Conference on Biomedical & Health Informatics (BHI). IEEE; 2018. [doi: 10.1109\/BHI.2018.8333385]","DOI":"10.1109\/BHI.2018.8333385"},{"issue":"12","key":"R26","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6579\/ad02da","article-title":"SEResUTer: a deep learning approach for accurate ECG signal delineation and atrial fibrillation detection","volume":"44","author":"Li","journal-title":"Physiol Meas"},{"issue":"6","key":"R27","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pdig.0000539","article-title":"Detecting QT prolongation from a single-lead ECG with deep learning","volume":"3","author":"Alam","journal-title":"PLOS Digit Health"},{"issue":"14","key":"R28","doi-asserted-by":"publisher","first-page":"5414","DOI":"10.3390\/s22145414","article-title":"Deep-learning-based estimation of the spatial QRS-T angle from reduced-lead ECGs","volume":"22","author":"Santos Rodrigues","journal-title":"Sensors (Basel)"},{"key":"R29","unstructured":"shiansen\/ECG-PDF-DeepParamExtractor. GitHub. URL: https:\/\/github.com\/shiansen\/ECG-PDF-DeepParamExtractor [Accessed 20-07-2026]"}],"container-title":["JMIR Formative Research"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/formative.jmir.org\/2026\/1\/e80597","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T22:26:40Z","timestamp":1785536800000},"score":1,"resource":{"primary":{"URL":"https:\/\/formative.jmir.org\/2026\/1\/e80597"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,24]]},"references-count":29,"URL":"https:\/\/doi.org\/10.2196\/80597","relation":{},"ISSN":["2561-326X"],"issn-type":[{"value":"2561-326X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,24]]},"article-number":"v10i6e80597"}}