European heart journal. Digital health最新文献

Aims: Discharge education reduces recurrent cardiac events in people after acute coronary syndrome (ACS). This trial investigates the effectiveness of a self-administered avatar-based discharge education application (app) on knowledge and clinical outcomes among inpatients compared with usual care.

Methods and results: Single-centre randomized controlled trial of adults hospitalized with ACS who were being discharged home. The app addressed heart disease diagnosis, treatment, risk factors, symptoms, and secondary prevention. Primary outcome was heart disease knowledge at three months. Secondary outcomes were quality of life, cardiac rehabilitation attendance, hospital re-presentations, symptom beliefs, physical activity, and smoking status. Satisfaction and app costs were also evaluated. Participants (n = 84) were 86% male and aged 60 ± 11 years. Both groups had improved knowledge and quality of life. There was no difference in knowledge between groups at three months after adjusting for baseline scores [0.88 points, 95% confidence interval (CI) -5.00, 6.76]. Cardiac rehabilitation attendance was 74% and 64% in the intervention and control groups, with no differences between groups (relative risk 1.15, 95% CI 0.87, 1.51). Ninety-two per cent found the app easy to use, but only 50% used the app as anticipated. Economic analysis showed that the intervention was dominant.

Conclusion: In this sample of people with ACS with high cardiac rehabilitation attendance, the app was highly acceptable but did not improve knowledge compared with usual care. Knowledge improved in both groups and may have potential to reduce cost to the health service with the app. Further work should explore the most appropriate target audience for app-based education.

Clinical trial registration: ACTRN12622001436763.

Aims: Autonomic nervous system activation plays an important role in the pathophysiology of atrial fibrillation (AF). It can be determined using heart rate variability (HRV). We aimed to evaluate the feasibility of using photoplethysmography (PPG) recordings for the assessment of the ultra-short-term HRV.

Methods and results: TeleCheck-AF is a structured mobile health approach, comprising the on-demand use of a PPG-based smartphone application prior to a scheduled teleconsultation to ensure comprehensive remote AF management. Participants with at least one PPG recording in sinus rhythm were included to assess resting heart rate, root mean square of successive differences (RMSSD), patient compliance and data consistency. In total, 855 patients [39.4% women] with 13 465 recordings were included. Patient compliance was 95.2% (IQR 76.2-114.3%). Median heart rate per patient was 66.5 (IQR 60.0-74.0) b.p.m., median RMSSD per patient was 40 (IQR 33-50) ms and median recording consistency was ±5.2 (IQR 3.8-7.0) b.p.m. and ±14.8 (IQR 9.3-21.1) ms, respectively. RMSSD was lower in men than women, in patients with CHA₂D₂-VA-Score 0, with a history of AF, and following ablation of AF. Older age and lower body mass index were associated with higher RMSSD.

Conclusion: The ultra-short-term HRV can be determined in 1-min PPG recordings with high user compliance and high inter-recording consistency within a structured mobile health AF management approach. The strategy used in this study may also be feasible for the management of other conditions in which the HRV plays a role for diagnostics and therapy.

Aims: Artificial intelligence (AI)-enhanced 12-lead electrocardiogram (ECG) can detect a range of structural heart diseases (SHDs); however, it has a limited role in community-based screening. We developed and externally validated a noise-resilient single-lead AI-ECG algorithm that can detect SHDs and predict the risk of their development using wearable/portable devices.

Methods and results: Using 266 740 ECGs from 99 205 patients with paired echocardiographic data at Yale New Haven Hospital, we developed AI Deep learning for Adapting Portable Technology in HEART disease detection (ADAPT-HEART), a noise-resilient, deep learning algorithm, to detect SHDs using lead I ECG. SHD was defined as a composite of having a left ventricular ejection fraction of < 40%, moderate or severe left-sided valvular disease, and severe left ventricular hypertrophy. ADAPT-HEART was validated in four community hospitals in USA, and the population-based cohort of ELSA-Brasil. We assessed the model's performance as a predictive biomarker among those without baseline SHD across hospital-based sites and the UK Biobank. The development population had a median age of 66 [interquartile range, 54-77] years and included 49 947 (50.3%) women, with 18 896 (19.0%) having any SHD. ADAPT-HEART had an area under the receiver operating characteristics curve (AUROC) of 0.879 (95% confidence interval, 0.870-0.888) with good calibration for detecting SHD in the test set, and consistent performance in hospital-based external sites (AUROC: 0.852-0.891) and ELSA-Brasil (AUROC: 0.859). Among individuals without baseline SHD, high vs. low ADAPT-HEART probability conferred a 2.8- to 5.7-fold increase in the risk of future SHD across data sources (all P < 0.05).

Conclusion: We propose a novel model that detects and predicts a range of SHDs from noisy single-lead ECGs obtainable on portable/wearable devices, providing a scalable strategy for community-based screening and risk stratification for SHD.

Aims: The classification of physiological patterns of coronary artery disease (CAD) is crucial for clinical decision-making, significantly affecting the planning and success of percutaneous coronary interventions (PCIs). This study aimed to develop a novel index to reliably interpret and classify physiological CAD patterns based on virtual pullbacks from single-view Murray's law-based quantitative flow ratio (μFR) analysis.

Methods and results: The pullback pressure gradient index (PPGi) was used to classify CAD patterns, with a cut-off value of PPGi = 0.78 distinguishing focal from diffuse and non-focal disease. The machine learning methods using penalized logistic regression and random forest were proposed to assess CAD patterns. Scores derived from multivariate functional principal component analysis of μFR and quantitative coronary analysis improved model performance. Expert panel interpretations served as the reference. A total of 343 vessels (291 patients) underwent classification. The PPGi cut-off of 0.78 achieved 67% accuracy [95% confidence interval (CI): 66-68%] for focal vs. diffuse and 76% accuracy (95% CI: 75-76%) for focal vs. non-focal classification. The penalized logistic regression model, including PPGi as a feature, provided superior accuracy: 88% (95% CI: 87-88%) for focal vs. diffuse and 81% (95% CI: 80-81%) for focal vs. non-focal classification. Moreover, the random forest model with PPGi as one of the features was applied for multiclass classification, providing an accuracy of 73% (95% CI: 73-73%).

Conclusion: The machine learning models for physiological patterns of CAD classification outperformed the binary PPGi method, providing robust and generalizable classification across different study populations.

Aims: Atrial fibrillation (AF) is characterized by heterogeneity in presentation, comorbidity profile and prognosis, with different AF subphenotypes having been previously suggested. Mental health disorders are common in the AF population. The current classification of AF, based on episode duration, fails to capture the complexity of the condition. Machine learning (ML) techniques and utilization of information on mental health disorders might improve identification of different and actionable AF subphenotypes.

Methods and results: We utilized Nationwide UK data from the Clinical Practice Research Datalink (199 308 AF patients; age 75.4 ± 12.6; 49.2% women) and unsupervised ML for clustering (k-means). Twenty-five clinical features were used in the model, including the presence of mental health disorders (anxiety, depression, and psychosis). Outcomes were assessed at 5 years across different clusters. We identified five different clusters of AF patients with specific characteristics and behaviour. Clusters were labelled based on the most prevalent features: (i) elderly and cardiopaths; (ii) young age and mental health disease; (iii) elderly and hypertensive; (iv) middle age and depression; and (v) very elderly. Mental health disorders were present in 18% at baseline. When comparing across the different clusters, significant differences were observed for the rates of the different assessed outcomes: higher mortality, heart failure and dementia in cluster (v), cancer and anxiety or depression in cluster (iii).

Conclusion: Using unsupervised clustering we identified five distinct clinically actionable AF subphenotypes. The differences in outcomes and event rates at 5 years, suggests the possibility of specific tailored therapy and interventions requiring further investigation. Management of mental health should be part of holistic or integrated care management in this population.

Aims: Rich data in cardiovascular diagnostic testing are often sequestered in unstructured reports, limiting their use.

Methods and results: We sequentially deployed generative and interpretative open-source large language models (LLMs; Llama2-70b, Llama2-13b). Using Llama2-70b, we generated varying formats of transthoracic echocardiogram (TTE) reports from 3000 real-world reports with paired structured elements. Using prompt-based supervised training, we fine-tuned Llama2-13b using sequentially larger batches of generated TTE reports as inputs, to extract data across 18 clinically-relevant echocardiographic fields. We evaluated the fine-tuned model, HeartDX-LM, on distinct datasets: (i) different time periods and formats at Yale New Haven Health System (YNHHS), (ii) Medical Information Mart for Intensive Care (MIMIC) III, and (iii) MIMIC IV. We used accuracy and Cohen's kappa as evaluation metrics and have publicly released the HeartDX-LM model. HeartDX-LM was trained on 2000 synthetic reports with varying formats and paired structured labels. We identified a lower threshold of 500 unstructured reports-structured data pairs required for fine-tuning to achieve consistent performance. At YNHHS, HeartDX-LM accurately extracted 69 144 of 70 032 values (98.7%) across 18 fields in the contemporary test set where paired structured data were available. In 100 older YNHHS reports, HeartDX-LM achieved 87.1% accuracy against expert annotations. In external validation sets from MIMIC-III and MIMIC-IV, HeartDX-LM correctly extracted 615 of 707 available values (87.9%) and 201 of 220 available values (91.3%), from 100 random, expert-annotated reports from each set.

Conclusion: We developed and validated a novel approach using paired large and moderate-sized LLMs to transform free-text echocardiographic reports into tabular datasets.

Cardiovascular disease (CVD) remains the leading cause of morbidity and mortality worldwide. Large language models (LLMs) offer potential solutions for enhancing patient education and supporting clinical decision-making. This study aimed to evaluate LLMs' applications in CVD and explore their current implementation, from prevention to treatment. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, this systematic review assessed LLM applications in CVD. A comprehensive PubMed search identified relevant studies. The review prioritized pragmatic and practical applications of LLMs. Key applications, benefits, and limitations of LLMs in CVD prevention were summarized. Thirty-five observational studies met the eligibility criteria. Of these, 54% addressed primary prevention and risk factor management, while 46% focused on established CVD. Commercial LLMs were evaluated in all but one study, with 91% (32 studies) assessing ChatGPT. The LLM applications were categorized as follows: 72% addressed patient education, 17% clinical decision support, and 11% both. In 68% of studies, the primary objective was to evaluate LLMs' performance in answering frequently asked patient questions, with results indicating accurate, comprehensive, and generally safe responses. However, occasional misinformation and hallucinated references were noted. Additional applications included patient guidance on CVD, first aid, and lifestyle recommendations. Large language models were assessed for medical questions, diagnostic support, and treatment recommendations in clinical decision support. Large language models hold significant potential in CVD prevention and treatment. Evidence supports their potential as an alternative source of information for addressing patients' questions about common CVD. However, further validation is needed for their application in individualized care, from diagnosis to treatment.

Aims: Accurate detection of coronary artery stenosis (CAS) on coronary computed tomography angiography is vital for saving lives, as timely diagnosis can prevent severe cardiac events. However, this task remains challenging due to data complexity and variability in imaging protocols. Deep learning offers promising potential to automate detection, but robust methods are essential to address real-world challenges effectively and enhance patient outcomes.

Methods and results: A total of 900 cases with curved multiplanar reformations, pre-generated during routine clinical workflows, were used to train a multi-instance learning (MIL) model for detecting significant CAS (≥50% luminal obstruction) in the left anterior descending (LAD), right coronary artery (RCA), and left circumflex (LCX), comprising 776 LAD, 694 RCA, and 600 LCX reconstructions. Patient-level predictions utilized attention scores to quantify each slice's contribution, ensuring a robust and interpretable diagnostic approach. The model achieved the best performance for LAD [area under the curve (AUC): 0.92, 95% confidence interval (CI): 0.87-0.96; Brier score: 0.11], followed by RCA (AUC: 0.91, 95% CI: 0.82-0.999; Brier score: 0.09) and LCX (AUC: 0.92, 95% CI: 0.84-0.99; Brier score: 0.07). Calibration was good for LAD but less precise for RCA and LCX. Attention scores enhanced diagnostic precision by focusing on the most relevant slices.

Conclusion: This study highlights the potential of MIL models for CAS detection, with remarkable performance in the LAD. By using attention scores, the model effectively identifies key slices from real-world data, seamlessly integrating with routine clinical workflows. Multi-range pre-processing addresses data complexity, enhancing diagnostic accuracy and supporting clinical decision-making.

Aims: The increasing need for diagnostic echocardiography tests presents challenges in preserving the quality and promptness of reports. While Large Language Models (LLMs) have proven effective in summarizing clinical texts, their application in echo remains underexplored.

Methods and results: Adult echocardiography studies, conducted at the Mayo Clinic from 1 January 2017 to 31 December 2017, were categorized into two groups: development (all Mayo locations except Arizona) and Arizona validation sets. We adapted open-source LLMs (Llama-2, MedAlpaca, Zephyr, and Flan-T5) using In-Context Learning and Quantized Low-Rank Adaptation fine-tuning (FT) for echo report summarization from 'Findings' to 'Impressions.' Against cardiologist-generated Impressions, the models' performance was assessed both quantitatively with automatic metrics and qualitatively by cardiologists. The development dataset included 97 506 reports from 71 717 unique patients, predominantly male (55.4%), with an average age of 64.3 ± 15.8 years. EchoGPT, a fine-tuned Llama-2 model, outperformed other models with win rates ranging from 87% to 99% in various automatic metrics, and produced reports comparable to cardiologists in qualitative review (significantly preferred in conciseness (P < 0.001), with no significant preference in completeness, correctness, and clinical utility). Correlations between automatic and human metrics were fair to modest, with the best being RadGraph F1 scores vs. clinical utility (r = 0.42) and automatic metrics showed insensitivity (0-5% drop) to changes in measurement numbers.

Conclusion: EchoGPT can generate draft reports for human review and approval, helping to streamline the workflow. However, scalable evaluation approaches dedicated to echo reports remains necessary.