A Novel Contact-Free Atrial Fibrillation Monitor- A Pilot Study

Aims Atrial fibrillation (AF) is a major cause of morbidity and mortality. Current guidelines support performing ECG screenings to spot atrial fibrillation in high-risk patients. The purpose of this study was to validate a new algorithm aimed to identify AF in patients measured with a recent FDA-cleared contact-free optical device. Methods and results Study participants were measured simultaneously using two devices: a contact-free optical system that measures chest motion-vibrations (investigational device, “Gili”) and a standard reference bed-side ECG monitor (Mindray®). Each reference ECG was evaluated by two board certified cardiologists that defined each trace as: regular rhythm, atrial fibrillation, other irregular rhythm or indecipherable/missing. A total of 3582, 30-sec intervals, pertaining to 444 patients (41.9% with a history of AF) were made available for analysis. Distribution of patients with active AF, other irregular rhythm and regular rhythm was 16.9%, 29.5% and 53.6% respectively. Following application of cross-validated machine learning approach, the observed sensitivity and specificity were 0.92 (95% CI: 0.91-0.93) and 0.96 (95% CI: 0.95-0.96) respectively. Conclusions This study demonstrates for the first time the efficacy of a contact-free optical device for detecting atrial fibrillation.

Feasibility of remote technical support for electrophysiological ablation procedures during the current CoVid-19 pandemic

The treatment of heart rhythm disorders has been significantly impacted by direct consequences of the current CoVid-19 pandemic, as well as by restrictions aimed towards constraining viral spread. Usually, catheter ablations of cardiac arrhythmias are guided by electro-anatomic mapping systems. Technical staff with medical training, or medical staff with technical training, is needed to assist the operator. Travel restrictions due to the current COVID-19 pandemic have limited the in-person availability of technical support staff. To overcome these limitations, we explored the feasibility of a trans-atlantic remote technical support for electro-anatomic mapping, with an internet based communication platform, for complex electrophysiological ablation procedures. Our first experience, based on nine ablation procedures of different arrythmias, highlights the feasibility of this approach. Remote support for electro-anatomic mapping might therefore facilitate continuous care for patients with arrhythmias during the COVID-19 pandemic, particularly in insular settings. Beyond COVID-19 related challenges, this approach will likely play a greater role in the cardiology field in years to come, due to its significant advantages.

Predictors of non-participation in a cardiac telerehabilitation programme: a prospective analysis

Introduction Current cardiac telerehabilitation (CTR) interventions are insufficiently tailored to the preferences and competences of individual patients, which raises the question whether their implementation will increase overall participation and adherence to cardiac rehabilitation. However, research on patient-specific factors that influence participation and adoption of CTR interventions is scarce. Objective The aim of this study was to evaluate which patient-related characteristics influence participation in a novel CTR intervention in patients with coronary artery disease. Methods This prospective observational substudy of the SmartCare-CAD randomised controlled trial evaluated patient characteristics of study participants as proxy for participation in a CTR intervention. We compared demographic, geographic and health-related characteristics between trial participants and non-participants to determine which characteristics influenced trial participation. Results A total of 699 patients (300 participants and 399 non-participants; 84% male, mean age 64.3 ± 10.5 years) were included. Most of the non-participants refused participation because of insufficient technical skills or lack of interest in digital health (26%), or preferred centre-based cardiac rehabilitation (CR) (21%). Variables independently associated with non-participation included: higher age, lower educational level, shorter traveling distance, smoking, positive family history for cardiovascular disease, having undergone coronary artery bypass grafting; and a higher blood pressure, worse exercise capacity and higher risk of depression before the start of CR. Conclusion Participation in CTR is strongly influenced by demographic and health-related factors such as age, educational level, smoking status and both physical and mental functioning. CTR interventions should therefore be redesigned with the involvement of these currently underrepresented patient subgroups.

Preventing unnecessary imaging in patients suspect of coronary artery disease through machine learning of electronic health records

Introduction With the aging European population, the incidence of coronary artery disease (CAD) is expected to rise. This will likely result in an increased imaging use. Symptom recognition can be complicated, as symptoms caused by CAD can be atypical, particularly in women. Early CAD exclusion may help to optimize use of diagnostic resources and thus improve the sustainability of the healthcare system. Objective To develop sex-stratified algorithms, trained on routinely available electronic health records, raw electrocardiograms, and hematology data to exclude CAD in patients upfront. Methods We trained XGBoost algorithms on data from patients from the Utrecht Patient-Oriented Database, who underwent coronary computed tomography angiography (CCTA), and/or stress cardiac magnetic resonance (CMR) imaging or stress single-photon emission computerized tomography (SPECT) in the UMC Utrecht. Outcomes were extracted from radiology reports. We aimed to maximize negative predictive value (NPV) to minimize the false negative risk with acceptable specificity. Results Of 6,808 CCTA patients (31% female), 1029 females (48%) and 1908 males (45%) had no diagnosis of CAD. Of 3,053 CMR/SPECT patients (45% female), 650 females (47%) and 881 males (48%) had no diagnosis of CAD. On the train and test set, the CCTA models achieved NPVs and specificities of 0.95 and 0.19 (females) and 0.96 and 0.09 (males). The CMR/SPECT models achieved NPVs and specificities of 0.75 and 0.041 (females) and 0.92 and 0.026 (males). Conclusion CAD can be excluded from EHRs with high NPV. Our study demonstrates new possibilities to reduce unnecessary imaging in women and men suspected of CAD.

Artificial intelligence–based detection of aortic stenosis from chest radiographs

Aims We aimed to develop models to detect aortic stenosis (AS) from chest radiographs—one of the most basic imaging tests—with artificial intelligence. Methods and Results We used 10433 retrospectively collected digital chest radiographs from 5638 patients to train, validate, and test three deep learning models. Chest radiographs were collected from patients who had also undergone echocardiography at a single institution between July 2016 and May 2019. These were labelled from the corresponding echocardiography assessments as AS-positive or AS-negative. The radiographs were separated on a patient basis into training (8327 images from 4512 patients, mean age 65 ± [SD] 15 years), validation (1041 images from 563 patients, mean age 65 ± 14 years), and test (1065 images from 563 patients, mean age 65 ± 14 years) datasets. The soft voting-based ensemble of the three developed models had the best overall performance for predicting AS with an AUC, sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of 0.83 (95% CI 0.77–0.88), 0.78 (0.67–0.86), 0.71 (0.68–0.73), 0.71 (0.68–0.74), 0.18 (0.14–0.23), and 0.97 (0.96–0.98), respectively, in the validation dataset and 0.83 (0.78–0.88), 0.83 (0.74–0.90), 0.69 (0.66–0.72), 0.71 (0.68–0.73), 0.23 (0.19–0.28), and 0.97 (0.96–0.98), respectively, in the test dataset. Conclusion Deep learning models using chest radiographs have the potential to differentiate between radiographs of patients with and without AS. Lay summary We created AI models using deep learning to identify aortic stenosis from chest radiographs. Three AI models were developed and evaluated with 10433 retrospectively collected radiographs and labelled from echocardiography reports. The ensemble AI model could detect aortic stenosis in a test dataset with an AUC of 0.83 (95% CI 0.78–0.88). Since chest radiography is a cost effective and widely available imaging test, our model can provide an additive resource for the detection of aortic stenosis.

Atrial fibrillation screening: feasible approaches and implementation challenges across Europe

Background Atrial fibrillation (AF) screening has the potential to increase early detection and possibly reduce complications of AF. Guidelines recommend screening, but the most appropriate approaches are unknown. Purpose We aimed to explore the views of stakeholders across Europe about the opportunities and challenges of implementing four different AF screening scenarios. Method This qualitative study included 21 semi-structured interviews with healthcare professionals and regulators potentially involved in AF screening implementation in nine European countries. Data were analysed using thematic analysis. Results Three themes evolved. 1) Current approaches to screening: there are no national AF screening programmes, with most AF detected in symptomatic patients. Patient-led screening exists via personal devices, creating screening inequity by the reach of screening programmes being limited to those who access healthcare services. 2) Feasibility of screening approaches: single time point opportunistic screening in primary care using single lead ECG devices was considered the most feasible approach and AF screening may be possible in previously unexplored settings such as dentists and podiatrists. Software algorithms may aid identification of patients suitable for screening and telehealth services have the potential to support diagnosis. However, there is a need for advocacy to encourage the use of telehealth to aid AF diagnosis, and training for screening familiarisation and troubleshooting. 3) Implementation requirements: sufficient evidence of benefit is required. National rather than pan-European screening processes must be developed due to different payment mechanisms and health service regulations. There is concern that the rapid spread of wearable devices for heart rate monitoring may increase workload due to false positives in low risk populations for AF. Data security and inclusivity for those without access to primary care or personal devices must be addressed. Conclusions There is an overall awareness of AF screening. Opportunistic screening appears to be most feasible across Europe. Challenges that need to be addressed concern health inequalities, identification of best target groups for screening, streamlined processes, the need for evidence of benefit, and a tailored approach adapted to national realities. Funding Acknowledgement Type of funding sources: Public grant(s) – EU funding. Main funding source(s): H2020 Screening Scenarios Graphical abstract

On-device artificial intelligence: mobile solution for detecting severe aortic valve stenosis based on heart sounds

Background Aortic stenosis is still one of the major causes of sudden cardiac death in the elderly. Noninvasive screening for severe aortic valve stenosis (AS) may result in early cardiac diagnostic leading to an appropriate and timely medical intervention. Purpose The aims of this study were 1) to develop an artificial intelligence to detect severe AS based on heart sounds and 2) to build an application to screen patients using electronic stethoscope and smartphones, which will provide an efficient diagnostic workflow for screening as a complementary tool in daily clinical practice. Methods We enrolled 100 patients diagnosed with severe AS and 200 patients without severe AS (no echocardiographic sign of AS [n=100], mild AS [n=50], moderate AS [n=50]). The heart sounds were recorded in 4000 Hz waveform audio format at the following 3 sites of each patient; the 2nd intercostal right sternal border, the Erb's area and the apex. Each record was divided into multiple data of 4 seconds duration, which built 10800 sound records in total. We developed multiple convolutional neural networks (CNN) designed to recognize severe AS in heart sounds according to the recorded 3 sites. We adopted a stratified 4-fold cross-validation method by which the CNN was trained with 60% of the whole data, validated with 20% data and tested with the remaining 20% data not used during training and validation. As performance metrics we adopted the accuracy, F1 value and the area under the curve (AUC) calculated as the average of all cross-validation folds. For the smartphone application, we combined the best CNN-models from each recorded site for the best performance. Further 40 patients were newly enrolled for its clinical validation (no AS [n=10], mild AS [n=10], moderate AS [n=10], severe AS [n=10]). Results The accuracy, F1 value and AUC of each model were 88.9±5.7%, 0.888±0.006 and 0.953±0.008, respectively. The sensitivity and specificity were 87.9±2.2% and 89.9±2.4%. The recognition accuracy of moderate AS was significantly lower as compared to the other AS grades (moderate AS 74.1±6.1% vs no AS 98.0±1.4%, mild AS 97.6±1.2%, severe AS 87.9±2.2%, respectively, P<0.05). Our smartphone application showed a sensitivity of 100% (10/10), a specificity of 73.3% (22/30), and an accuracy of 80.0% (32/40), which implicated a good utility for screening. In the detailed analysis of 8 mistaken decisions, these were highly affected by the presence of severe mitral or tricuspid valve regurgitation despite of non-severe AS (7/8 [87.5%]). Conclusions This study demonstrated the promising possibility of an end-to-end screening for severe aortic valve stenosis using an electronic stethoscope and a smartphone application. This technology may improve the efficacy of daily medicine particularly where the human resource is limited or support a remote medical consultation. Further investigations are necessary to increase accuracy. Funding Acknowledgement Type of funding sources: None.

Inter operator variability of machine learning researchers predicting all-cause mortality in patients admitted to intensive care unit

Background The number of publications using machine learning (ML) to predict cardiovascular outcomes and identify clusters of patients at greater risk has risen dramatically in recent years. However, research papers which use ML often fail to provide sufficient information about their algorithms to enable results to be replicated by others in the same or different datasets. Aim To test the reproducibility of results from ML algorithms given three different levels of information commonly found in publications: model type alone, a description of the model, and complete algorithm. Methods MIMIC-III is a healthcare dataset comprising detailed information from over 60,000 intensive care unit (ICU) admissions from the Beth Israel Deaconess Medical Centre between 2001 and 2012. Access is available to everyone pending approval and completion of a short training course. Using this dataset, three models for predicting all-cause in-hospital mortality were created, two from a PhD student working in ML, and one from an existing research paper which used the same dataset and provided complete model information. A second researcher (a PhD student in ML and cardiology) was given the same dataset and was tasked with reproducing their results. Initially, this second researcher was told what type of model was created in each case, followed by a brief description of the algorithms. Finally, the complete algorithms from each participant were provided. In all three scenarios, recreated models were compared to original models using Area Under the Receiver Operating Characteristic Curve (AUC). Results After excluding those younger than 18 years and events with missing or invalid entries, 21,139 ICU admissions remained from 18,094 patients between 2001 and 2012, including 2,797 in-hospital deaths. Three models were produced: two Recurrent Neural Networks (RNNs) which differed significantly in internal weights and variables, and a Boosted Tree Classifier (BTC). The AUC of the first reproduced RNN matched that of the original RNN (Figure 1), however the second RNN and the BTC could not be reproduced given model type alone. As more information was provided about these algorithms, the results from the reproduced models matched the original results more closely. Conclusions In order to create clinically useful ML tools with results that are reproducible and consistent, it is vital that researchers share enough detail about their models. Model type alone is not enough to guarantee reproducibility. Although some models can be recreated with limited information, this is not always the case, and the best results are found when the complete algorithm is shared. These findings have huge relevance when trying to apply ML in clinical practice. Funding Acknowledgement Type of funding sources: None.

Hybrid telerehabilitation in patients with hypertrophic cardiomyopathy without left ventricular outflow tract obstruction and preserved left ventricular ejection fraction-the randomised clinical trial

Background Hypertrophic cardiomyopathy (HCM) is the most common hereditary heart disease, and its diagnosis is often associated with limited physical activity. Little is known about cardiac rehabilitation programs for patients with HCM. Therefore the novel hybrid cardiac telerehabilitation (HCTR) model consisting of hospital-based rehabilitation and home-based telemonitored rehabilitation might be an option to improve physical capacity in patients with HCM. Purpose To evaluate the safety, effectiveness and adherence to HCTR in patients with HCM without the left ventricle (LV) outflow tract obstruction and preserved LV ejection fraction. Methods The study group comprised 60 patients with HCM (51.1±13.3 years; NYHA II-III; LV ejection fraction 66.1±6.9%). Patients were randomised (1:1) to either HCTR program (hospital-based rehabilitation [1 month] based on cycloergometer training and home-based telemonitored rehabilitation [2 months] based on Nordic walking, five times a week, at 40–70% of maximal estimated heart rate) - training group (TG), or to a control group (CG). All patients had implantable cardioverter-defibrillator. In order to perform home-based telemonitored rehabilitation, a special device was used which enabled patients to: (1) do Nordic walking training according to a preprogrammed plan, (2) record and send electrocardiograms (ECGs) via mobile phone network to the monitoring centre. The moments of automatic ECGs registration were pre-set and coordinated with exercise training. The effectiveness of HCTR was assessed by changes - delta (Δ) in duration (t) of the workload, peak oxygen consumption (pVO2) in cardiopulmonary exercise test, 6-minute walking test distance (6-MWT) as a result of comparing t (s), pVO2 (ml/kg/min), 6-MWT (m) from the beginning and the end of the program. Results Safety of HCTR. Neither death nor other serious adverse events occurred during HCTR. We did not observe any ICDs intervention during the HCTR. Effectiveness of HCTR: Within-group analysis: t, pVO2, 6-MWT increased significantly in TG: t 657±183 vs 766±181 (p<0.001), pVO2 19.2±5.0 vs 20.6±4.9 (p=0.007), 6-MWT 445±88 vs 551±77 (p<0.001). In the untrained CG, the unfavourable results were observed: 695±198 vs 717±187 (p=0.114), pVO2 21.2±5.1 vs 21.1±5.6 (p=0.723), 6-MWT 512±83 vs 536±84 (p=0.061). Between-group analysis: The differences between TG and CG were statistically significant: in Δt (p<0.001); ΔpVO2 (p=0.012); Δ6-MWT (p<0.001). Adherence to HCTR: In TG 28 patients (93%) completed the HCTR program. Two patients did no undergo HCTR because of personal issues. Conclusion Hybrid cardiac telerehabilitation in patients with HCM without the left ventricle (LV) outflow tract obstruction and preserved LV ejection fraction is safe and effective. The adherence to HCTR is high. Funding Acknowledgement Type of funding sources: Public Institution(s). Main funding source(s): Statutory work in The Cardinal Stefan Wyszyński National Institute of Cardiology in Warsaw, Poland