Exercise Testing and Artificial Intelligence as Allies in Improving the Detection and Diagnosis of Long QT Syndrome
Bibliographic record
Abstract
Dehkordi et al,1Dehkordi N.R. Dehkordi N.R. Toudeshki K.K. Farjoo M.H. Artificial intelligence in diagnosis of long QT syndrome: a review of current state, challenges, and future perspectives.Mayo Clin Proc Digit Health. 2024; 2: 21-31https://doi.org/10.1016/j.mcpdig.2023.11.003Abstract Full Text Full Text PDF Google Scholar recently published an interesting review of current state, challenges, and future perspectives related to artificial intelligence (AI) in the diagnosis of long QT syndrome (LQTS). As LQTS is at the center of our current research, we very much enjoyed this innovative and enlightening read. Recent articles suggest that AI exhibits superior diagnostic performance compared with expert clinicians, particularly excelling in identifying cases of dangerous, concealed LQTS.1Dehkordi N.R. Dehkordi N.R. Toudeshki K.K. Farjoo M.H. Artificial intelligence in diagnosis of long QT syndrome: a review of current state, challenges, and future perspectives.Mayo Clin Proc Digit Health. 2024; 2: 21-31https://doi.org/10.1016/j.mcpdig.2023.11.003Abstract Full Text Full Text PDF Google Scholar We must, however, remind the scientific community about the major link between LQTS and exercise testing,2Harvey A. Curnier D. Dodin P. Abadir S. Jacquemet V. Caru M. The influence of exercise and postural changes on ventricular repolarization in the long QT syndrome: a systematic scoping review.Eur J Prev Cardiol. 2022; 29: 1633-1677https://doi.org/10.1093/eurjpc/zwac081Crossref Scopus (4) Google Scholar which may act as an ally to AI in the battle to improve LQTS detection and diagnosis. The main issue in LQTS diagnosis is the significant overlap of the corrected QT interval (QTc) range between LQTS (≥470 ms for males; ≥480 ms for females) and healthy individuals, as measured by the electrocardiogram (ECG).3Chattha I.S. Sy R.W. Yee R. et al.Utility of the recovery electrocardiogram after exercise: a novel indicator for the diagnosis and genotyping of long QT syndrome?.Heart Rhythm. 2010; 7: 906-911https://doi.org/10.1016/j.hrthm.2010.03.006Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar In fact, 25%-50% of LQTS patients have a resting QTc in the normal (<440 ms for males; <460 ms for females) or borderline (440-469 ms for males; 460-479 ms for females) range.4Sy R.W. van der Werf C. Chattha I.S. et al.Derivation and validation of a simple exercise-based algorithm for prediction of genetic testing in relatives of LQTS probands.Circulation. 2011; 124: 2187-2194https://doi.org/10.1161/CIRCULATIONAHA.111.028258Crossref PubMed Scopus (163) Google Scholar Exercise testing has been recognized as a key aspect in the identification and evaluation of patients at risk of congenital LQTS, especially for those who present with a dangerous, concealed QT interval prolongation at rest.2Harvey A. Curnier D. Dodin P. Abadir S. Jacquemet V. Caru M. The influence of exercise and postural changes on ventricular repolarization in the long QT syndrome: a systematic scoping review.Eur J Prev Cardiol. 2022; 29: 1633-1677https://doi.org/10.1093/eurjpc/zwac081Crossref Scopus (4) Google Scholar Specifically, LQTS causes abnormal QTc prolongation during exercise and/or recovery.2Harvey A. Curnier D. Dodin P. Abadir S. Jacquemet V. Caru M. The influence of exercise and postural changes on ventricular repolarization in the long QT syndrome: a systematic scoping review.Eur J Prev Cardiol. 2022; 29: 1633-1677https://doi.org/10.1093/eurjpc/zwac081Crossref Scopus (4) Google Scholar Hence, the 1993-2011 Schwartz LQTS Diagnostic Criteria suggest the combination of resting and recovery QTc data to increase diagnostic sensitivity and specificity.5Schwartz P.J. Crotti L. QTc Behavior during exercise and genetic testing for the long-QT syndrome.Circulation. 2011; 124: 2181-2184https://doi.org/10.1161/CIRCULATIONAHA.111.062182Crossref PubMed Scopus (250) Google Scholar Interestingly, studies have also identified a genotype-specific repolarization response to exercise. Thus, exercise testing may serve as a phenotypic enhancer of silent congenital LQTS.2Harvey A. Curnier D. Dodin P. Abadir S. Jacquemet V. Caru M. The influence of exercise and postural changes on ventricular repolarization in the long QT syndrome: a systematic scoping review.Eur J Prev Cardiol. 2022; 29: 1633-1677https://doi.org/10.1093/eurjpc/zwac081Crossref Scopus (4) Google Scholar As described in the review, the potential of AI-driven analysis of ECG data to accurately identify and anticipate LQTS diagnosis and distinguish genetic subtypes is promising.1Dehkordi N.R. Dehkordi N.R. Toudeshki K.K. Farjoo M.H. Artificial intelligence in diagnosis of long QT syndrome: a review of current state, challenges, and future perspectives.Mayo Clin Proc Digit Health. 2024; 2: 21-31https://doi.org/10.1016/j.mcpdig.2023.11.003Abstract Full Text Full Text PDF Google Scholar However, the authors transparently discuss several important challenges, including uninterpretable AI decision-making processes and the potential misclassification of healthy controls as LQTS based on resting ECG values.1Dehkordi N.R. Dehkordi N.R. Toudeshki K.K. Farjoo M.H. Artificial intelligence in diagnosis of long QT syndrome: a review of current state, challenges, and future perspectives.Mayo Clin Proc Digit Health. 2024; 2: 21-31https://doi.org/10.1016/j.mcpdig.2023.11.003Abstract Full Text Full Text PDF Google Scholar Specifically, if resting ECG data is incorrect or not sufficiently robust, the AI model will lack power and, therefore, will be insufficient to make it a strong element of clinical diagnosis. To address this issue, exercise ECG data should be added to resting ECG data to optimize AI, machine learning, and neural network analysis systems. This may ultimately help us decipher AI analytics, all the while limiting unnecessary genetic testing, health care costs, and unwanted psychological distress due to false-positive results. In particular, the exercise dataset would allow more congenitally affected individuals who may go undetected at rest to be identified and properly managed. In return, AI may be advantageous for the much-needed standardization of exercise testing in LQTS screening and follow-up. Manual and automatic ECG analysis is very time consuming, especially when using continuous ECG recording. As mentioned by Dehkordi et al,1Dehkordi N.R. Dehkordi N.R. Toudeshki K.K. Farjoo M.H. Artificial intelligence in diagnosis of long QT syndrome: a review of current state, challenges, and future perspectives.Mayo Clin Proc Digit Health. 2024; 2: 21-31https://doi.org/10.1016/j.mcpdig.2023.11.003Abstract Full Text Full Text PDF Google Scholar AI offers promising solutions by enhancing the accuracy and efficiency of ECG interpretation. Notably, AI algorithms can process ECG data more rapidly than human experts, provide real-time analysis, and reduce interobserver variability.1Dehkordi N.R. Dehkordi N.R. Toudeshki K.K. Farjoo M.H. Artificial intelligence in diagnosis of long QT syndrome: a review of current state, challenges, and future perspectives.Mayo Clin Proc Digit Health. 2024; 2: 21-31https://doi.org/10.1016/j.mcpdig.2023.11.003Abstract Full Text Full Text PDF Google Scholar Therefore, there could be a 2-way advantage of using exercise in conjunction with AI for the improvement of LQTS identification and diagnosis. The combination of resting ECG, continuous exercise ECG, and AI analytics could be the key recipe to ensure that individuals with a concealed form of LQTS do not go undetected, and consequently, dangerously unmanaged.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".