Carter, Rickey E.

Background Wearable devices and sensor technology provide objective, unbiased range of motion measurements that help health care professionals overcome the hindrances of protractor-based goniometry. This review aims to analyze the accuracy of existing wearable sensor technologies for hand range of motion measurement and identify the most accurate one. Methods We performed a systematic review by searching PubMed, CINAHL, and Embase for studies evaluating wearable sensor technology in hand range of motion assessment. Keywords used for the inquiry were related to wearable devices and hand goniometry. Results Of the 71 studies, 11 met the inclusion criteria. Ten studies evaluated gloves and 1 evaluated a wristband. The most common types of sensors used were bend sensors, followed by inertial sensors, Hall effect sensors, and magnetometers. Most studies compared wearable devices with manual goniometry, achieving optimal accuracy. Although most of the devices reached adequate levels of measurement error, accuracy evaluation in the reviewed studies might be subject to bias owing to the use of poorly reliable measurement techniques for comparison of the devices. Conclusion Gloves using inertial sensors were the most accurate. Future studies should use different comparison techniques, such as infrared camera–based goniometry or virtual motion tracking, to evaluate the performance of wearable devices.

Abstract Aims Early detection of aortic stenosis (AS) is becoming increasingly important with a better outcome after aortic valve replacement in asymptomatic severe AS patients and a poor outcome in moderate AS. We aimed to develop artificial intelligence-enabled electrocardiogram (AI-ECG) using a convolutional neural network to identify patients with moderate to severe AS. Methods and results Between 1989 and 2019, 258 607 adults [mean age 63 ± 16.3 years; women 122 790 (48%)] with an echocardiography and an ECG performed within 180 days were identified from the Mayo Clinic database. Moderate to severe AS by echocardiography was present in 9723 (3.7%) patients. Artificial intelligence training was performed in 129 788 (50%), validation in 25 893 (10%), and testing in 102 926 (40%) randomly selected subjects. In the test group, the AI-ECG labelled 3833 (3.7%) patients as positive with the area under the curve (AUC) of 0.85. The sensitivity, specificity, and accuracy were 78%, 74%, and 74%, respectively. The sensitivity increased and the specificity decreased as age increased. Women had lower sensitivity but higher specificity compared with men at any age groups. The model performance increased when age and sex were added to the model (AUC 0.87), which further increased to 0.90 in patients without hypertension. Patients with false-positive AI-ECGs had twice the risk for developing moderate or severe AS in 15 years compared with true negative AI-ECGs (hazard ratio 2.18, 95% confidence interval 1.90–2.50). Conclusion An AI-ECG can identify patients with moderate or severe AS and may serve as a powerful screening tool for AS in the community.

Aims To investigate the prognostic value of diagnostic scores for heart failure (HF) with preserved ejection fraction (HFpEF). Methods and results Consecutive patients with HFpEF admitted for unequivocal decompensated HF treated with intravenous loop diuretics were evaluated ( n = 443; mean age 78 ± 12 years; 60% women). The HFA‐PEFF and H 2 FPEF scores were calculated for all patients with echocardiography data available within 1 year and the population was stratified according to HFA‐PEFF scores 2–4 ( n = 79), 5 ( n = 93), or 6 ( n = 271) and H 2 FPEF score probabilities <90% ( n = 80), 90–95% ( n = 61), and 96–100% ( n = 293). HF readmission rates (95% confidence intervals) increased from 28.9 (22.7–35.0) per 100 patient‐years in HFA‐PEFF 2–4 to 46.0 (38.5–53.5) in HFA‐PEFF 5 and 45.0 (40.1–49.8) in HFA‐PEFF 6. Similarly, HF readmission rates increased with increasing H 2 FPEF probability: <0.90 [31.8 (25.3–38.2) per 100 patient‐years], 0.90–0.95 [41.5 (32.9–50.1)], and 0.96–1.00 [45.9 (41.2–50.6]. Median survival was 65 months (36–89 months) in HFA‐PEFF score 2–4, 45 months (26–59 months) in HFA‐PEFF score 5, and 28 months (22–42 months) in HFA‐PEFF score 6 ( P < 0.001), while the hazard ratio (95% confidence interval) for all‐cause mortality was 1.16 (1.02–1.32) per 0.10 increase in H 2 FPEF probability. Conclusions Among patients hospitalized with HFpEF, higher HFpEF probability according to diagnostic scores is associated with increased risk of subsequent HF readmissions and all‐cause mortality.

Background: Heart rate–corrected QT interval (QTc) prolongation, whether secondary to drugs, genetics including congenital long QT syndrome, and/or systemic diseases including SARS-CoV-2–mediated coronavirus disease 2019 (COVID-19), can predispose to ventricular arrhythmias and sudden cardiac death. Currently, QTc assessment and monitoring relies largely on 12-lead electrocardiography. As such, we sought to train and validate an artificial intelligence (AI)–enabled 12-lead ECG algorithm to determine the QTc, and then prospectively test this algorithm on tracings acquired from a mobile ECG (mECG) device in a population enriched for repolarization abnormalities. Methods: Using >1.6 million 12-lead ECGs from 538 200 patients, a deep neural network (DNN) was derived (patients for training, n = 250 767; patients for testing, n = 107 920) and validated (n = 179 513 patients) to predict the QTc using cardiologist-overread QTc values as the “gold standard”. The ability of this DNN to detect clinically-relevant QTc prolongation (eg, QTc ≥500 ms) was then tested prospectively on 686 patients with genetic heart disease (50% with long QT syndrome) with QTc values obtained from both a 12-lead ECG and a prototype mECG device equivalent to the commercially-available AliveCor KardiaMobile 6L. Results: In the validation sample, strong agreement was observed between human over-read and DNN-predicted QTc values (−1.76±23.14 ms). Similarly, within the prospective, genetic heart disease–enriched dataset, the difference between DNN-predicted QTc values derived from mECG tracings and those annotated from 12-lead ECGs by a QT expert (−0.45±24.73 ms) and a commercial core ECG laboratory [10.52±25.64 ms] was nominal. When applied to mECG tracings, the DNN’s ability to detect a QTc value ≥500 ms yielded an area under the curve, sensitivity, and specificity of 0.97, 80.0%, and 94.4%, respectively. Conclusions: Using smartphone-enabled electrodes, an AI DNN can predict accurately the QTc of a standard 12-lead ECG. QTc estimation from an AI-enabled mECG device may provide a cost-effective means of screening for both acquired and congenital long QT syndrome in a variety of clinical settings where standard 12-lead electrocardiography is not accessible or cost-effective.

AbstractBackgroundUsing an objective mathematical algorithm to assess corticolimbic involvement of neurofibrillary tangle pathology, we identified three AD subtypes: hippocampal sparing (HpSp) AD, typical AD, and limbic predominant AD. Typical AD brains were representative of the expected patterns of hippocampal and cortical involvement as outlined by Braak tangle stage. In contrast, we discovered an extreme phenotype that exists outside of the Braakian‐concept of neurofibrillary tangle patterns. Hippocampal sparing AD cases demonstrate unexpected sparing of the hippocampus relative to severely involved association cortices and limbic predominant AD cases demonstrate inundation of the hippocampus relative to mildly involved association cortices. Using these AD subtypes, our objective was to leverage our understanding of disease spectrum to uncover transcriptomic changes that underlie selective vulnerability of the hippocampus in AD.MethodWe performed RNA‐sequencing in 40 neuropathologically diagnosed AD cases (10 HpSp, 20 Typical and 10 Limbic) and 15 controls to uncover gene expression changes associated with phenotype differences allowing us to prioritize 44 genes for further exploration. Validation with NanoString and quantification with digital pathology were performed in an expanded cohort of 158 AD cases (36 HpSp, 79 Typical and 35 Limbic) and 32 controls. We quantified digital pathology measures of early tangle pathology (CP13), mature tangle pathology (Ab39), amyloid‐β (33.1.1). In addition, we measured cellular markers for microglia (CD68), endothelia (CD34), and astroglia (GFAP). Deep learning based prediction models were employed, which nominated the SLC38A2 gene as a predictor of AD.ResultThe SLC38A2 gene was originally prioritized in our RNA‐sequencing data based upon differential expression between controls and typical AD. We observed upregulation of SLC38A2 in all AD subtypes. RNA‐seq gene expression measures validated well with NanoString for SLC38A2 (R=0.97, p<0.001). Regression analysis identified early tangle (p=0.011) and late tangle measures (p=0.026) as significant predictors of gene expression, but not amyloid measures (p=0.74). Microgliosis approached significance (p=0.052), but with no contribution from endothelial cell burden (p=0.17) or astrogliosis (p=0.58). Age approached significance (p=0.06), but not sex (p=0.88) or APOE‐ε4 status (p=0.10).ConclusionOur data supports consideration of intra‐disease divergence with regard to case stratification and may reveal genes previously masked by heterogeneity of cohorts.

Background: Myocardial injury results in release of cardiac troponin (cTn) into the bloodstream, readily detected by high-sensitivity cTn (hs-cTn) assays. A noninvasive, rapid, home-based test to exclude acute myocardial injury at the time of the test and for the next 7 hours would help manage patients with suspected myocardial infarction. Since myocyte injury is associated with ECG changes, we hypothesized an artificial intelligence ECG (AI-ECG) could identify absence of injury. Objective: To train and test an AI-ECG convolutional neural network (CNN) to identify patients using a single ECG who are suspected of myocardial infarction who are low risk, with hs-cTn levels below the 99 th percentile at test time and for the subsequent 7 hours. Methods: A CNN tuned to identify the absence of a hs-cTnT (5 th Gen cTnT Roche Diagnostics) >15ng/L for men and >10ng/L for women was developed. All ECGs were recorded within one hour of the hs-cTnT assay. We used 73,012 ECGs and hs-cTnT pairs from 47,542 unique patients to train the network, 9031 ECGs from 5,811 patients for internal validation to optimize hyperparameters, and 11,904 ECGs with 21,191 hs-cTnT measurements up to 7 hours after the ECG, from 11,904 different patients as a holdout test set. Results: The mean age was 63.9±17.5 years, and 30,348 of the 59,446 patients (51%) were male. 5,852 patients (49.1%) had no elevation of hs-cTnT and 6,052 (50.9%) had an hs-cTnT above the 99 th percentile at baseline or within 7 hours of the test. Of the 11,904 patients in the test set, using a sensitive threshold, the 12 lead AI ECG identified 1037 patients (8.7%) likely to have a low risk for subsequent hs-cTnT increases > 99 th % (AUC 0.86), and the single lead ECG identified 685 patients. Of the 1037 low risk pts, 59 had an hs-cTnT > 99 th percentile within 7 hours. The mean maximum hs-cTnT among the 59 low-risk patients was 53 ng/L ± 92 vs 184 ng/L ± 1474 in the others. None of these low risk patients died within 14 days of the test. Conclusion: Use of the AI ECG to identify patients without cardiac injury at the time of the test and for 7 subsequent hours is feasible. This may permit rapid, point of care, home-based, bloodless, painless testing to exclude cardiac injury, facilitating care and management of patients with suspected ischemia.