Descriptions of Etiology, Clinical Course, and Prognosis of Patients Presenting with Pericardial Effusion.

Pericardial effusion (PE) presentation varies from an incidental finding to a life-threatening situation; thus, its etiology and clinical course remain unknown. The aim of the present study was to retrospectively investigate these factors.We analyzed 171 patients (0.4%) who presented with PE among 34,873 patients who underwent echocardiography between 2011 and 2021 at our hospital. Clinical and prognostic information was retrieved from electronic medical records. The primary endpoints were all-cause death, hospitalization due to heart failure (HF), and other cardiovascular events such as cardiovascular death, acute coronary syndrome, elective percutaneous coronary intervention, and stroke.The etiologies of PE were as follows: idiopathic (32%), HF-related (18%), iatrogenic (11%), cardiac surgery-related (10%), radiation therapy-related (9%), malignancy (8%), pericarditis/myocarditis (8%), myocardial infarction-related (2%), and acute aortic dissection (2%). Patients with idiopathic/HF etiology were more likely to be older than the others.During a mean follow-up period of 2.5 years, all-cause death occurred in 21 patients (12.3%), cardiovascular events in 10 patients (5.8%), and hospitalization for HF in 24 patients (14.0%). All-cause death was frequently observed in patients with malignancy (44% per person-year). Cardiovascular events were mostly observed in patients with radiation therapy-related and malignancy (8.6% and 7.3% per person-year, respectively).The annual incidence of hospitalization for HF was the highest in patients with HF-related (25.1% per person-year), followed by radiation therapy-related (10.4% per person-year).This retrospective study is the first, to the best of our knowledge, to reveal the contemporary prevalence of PE, its cause, and outcome in patients who visited a cardiovascular hospital in an urban area of Japan.

Evaluating convolutional neural network-enhanced electrocardiography for hypertrophic cardiomyopathy detection in a specialized cardiovascular setting.

The efficacy of convolutional neural network (CNN)-enhanced electrocardiography (ECG) in detecting hypertrophic cardiomyopathy (HCM) and dilated HCM (dHCM) remains uncertain in real-world applications. This retrospective study analyzed data from 19,170 patients (including 140 HCM or dHCM) in the Shinken Database (2010-2017). We evaluated the sensitivity, positive predictive rate (PPR), and F1 score of CNN-enhanced ECG in a ''basic diagnosis'' model (total disease label) and a ''comprehensive diagnosis'' model (including disease subtypes). Using all-lead ECG in the "basic diagnosis" model, we observed a sensitivity of 76%, PPR of 2.9%, and F1 score of 0.056. These metrics improved in cases with a diagnostic probability of ≥ 0.9 and left ventricular hypertrophy (LVH) on ECG: 100% sensitivity, 8.6% PPR, and 0.158 F1 score. The ''comprehensive diagnosis'' model further enhanced these figures to 100%, 13.0%, and 0.230, respectively. Performance was broadly consistent across CNN models using different lead configurations, particularly when including leads viewing the lateral walls. While the precision of CNN models in detecting HCM or dHCM in real-world settings is initially low, it improves by targeting specific patient groups and integrating disease subtype models. The use of ECGs with fewer leads, especially those involving the lateral walls, appears comparably effective.

Identifying patients with acute aortic dissection using an electrocardiogram with convolutional neural network.

Background: The potential of utilizing artificial intelligence with electrocardiography (ECG) for initial screening of aortic dissection (AD) is promising. However, achieving a high positive predictive rate (PPR) remains challenging. Methods and results: This retrospective analysis of a single-center, prospective cohort study (Shinken Database 2010-2017, N = 19,170) used digital 12-lead ECGs from initial patient visits. We assessed a convolutional neural network (CNN) model's performance for AD detection with eight-lead (I, II, and V1-6), single-lead, and double-lead (I, II) ECGs via five-fold cross-validation. The mean age was 63.5 ± 12.5 years for the AD group (n = 147) and 58.1 ± 15.7 years for the non-AD group (n = 19,023). The CNN model achieved an area under the curve (AUC) of 0.936 (standard deviation [SD]: 0.023) for AD detection with eight-lead ECGs. In the entire cohort, the PPR was 7 %, with 126 out of 147 AD cases correctly diagnosed (sensitivity 86 %). When applied to patients with D-dimer levels ≥1 µg/dL and a history of hypertension, the PPR increased to 35 %, with 113 AD cases correctly identified (sensitivity 86 %). The single V1 lead displayed the highest diagnostic performance (AUC: 0.933, SD: 0.03), with PPR improvement from 8 % to 38 % within the same population. Conclusions: Our CNN model using ECG data for AD detection achieved an over 30% PPR when applied to patients with elevated D-dimer levels and hypertension history while maintaining sensitivity. A similar level of performance was observed with a single-lead V1 ECG in the CNN model.

Lead-Specific Performance for Atrial Fibrillation Detection in Convolutional Neural Network Models Using Sinus Rhythm Electrocardiography.

Background: We developed a convolutional neural network (CNN) model to detect atrial fibrillation (AF) using the sinus rhythm ECG (SR-ECG). However, the diagnostic performance of the CNN model based on different ECG leads remains unclear. Methods and Results: In this retrospective analysis of a single-center, prospective cohort study, we identified 616 AF cases and 3,412 SR cases for the modeling dataset among new patients (n=19,170). The modeling dataset included SR-ECGs obtained within 31 days from AF-ECGs in AF cases and SR cases with follow-up ≥1,095 days. We evaluated the CNN model's performance for AF detection using 8-lead (I, II, and V1-6), single-lead, and double-lead ECGs through 5-fold cross-validation. The CNN model achieved an area under the curve (AUC) of 0.872 (95% confidence interval (CI): 0.856-0.888) and an odds ratio of 15.24 (95% CI: 12.42-18.72) for AF detection using the eight-lead ECG. Among the single-lead and double-lead ECGs, the double-lead ECG using leads I and V1 yielded an AUC of 0.871 (95% CI: 0.856-0.886) with an odds ratio of 14.34 (95% CI: 11.64-17.67). Conclusions: We assessed the performance of a CNN model for detecting AF using eight-lead, single-lead, and double-lead SR-ECGs. The model's performance with a double-lead (I, V1) ECG was comparable to that of the 8-lead ECG, suggesting its potential as an alternative for AF screening using SR-ECG.

Identification of patients with dilated phase of hypertrophic cardiomyopathy using a convolutional neural network applied to multiple, dual, and single lead electrocardiograms.

Background: This study sought to develop an artificial intelligence-derived model to detect the dilated phase of hypertrophic cardiomyopathy (dHCM) on digital electrocardiography (ECG) and to evaluate the performance of the model applied to multiple-lead or single-lead ECG. Methods: This is a retrospective analysis using a single-center prospective cohort study (Shinken Database 2010-2017, n = 19,170). After excluding those without a normal P wave on index ECG (n = 1,831) and adding dHCM patients registered before 2009 (n = 39), 17,378 digital ECGs were used. Totally 54 dHCM patients were identified of which 11 diagnosed at baseline, 4 developed during the time course, and 39 registered before 2009. The performance of the convolutional neural network (CNN) model for detecting dHCM was evaluated using eight-lead (I, II, and V1-6), single-lead, and double-lead (I, II) ECGs with the five-fold cross validation method. Results: The area under the curve (AUC) of the CNN model to detect dHCM (n = 54) with eight-lead ECG was 0.929 (standard deviation [SD]: 0.025) and the odds ratio was 38.64 (SD 9.10). Among the single-lead and double-lead ECGs, the AUC was highest with the single lead of V5 (0.953 [SD: 0.038]), with an odds ratio of 58.89 (SD:68.56). Conclusion: Compared with the performance of eight-lead ECG, the most similar performance was achieved with the model with a single V5 lead, suggesting that this single-lead ECG can be an alternative to eight-lead ECG for the screening of dHCM.

Cardiovascular events and artificial intelligence-predicted age using 12-lead electrocardiograms.

Background: There is increasing evidence that 12-lead electrocardiograms (ECG) can be used to predict biological age, which is associated with cardiovascular events. However, the utility of artificial intelligence (AI)-predicted age using ECGs remains unclear. Methods: Using a single-center database, we developed an AI-enabled ECG using 17 042 sinus rhythm ECGs (SR-ECG) to predict chronological age (CA) with a convolutional neural network that yields AI-predicted age. Using the 5-fold cross validation method, AI-predicted age deriving from the test dataset was yielded for all ECGs. The incidence by AgeDiff and the areas under the curve by receiver operating characteristic curve with AI-predicted age for cardiovascular events were analyzed. Results: During the mean follow-up period of 460.1 days, there were 543 cardiovascular events. The annualized incidence of cardiovascular events was 2.24 %, 2.44 %, and 3.01 %/year for patients with AgeDiff < -6, -6 to ≤6, and >6 years, respectively. The areas under the curve for cardiovascular events with CA and AI-predicted age, respectively, were 0.673 and 0.679 (Delong's test, P = 0.388) for all patients; 0.642 and 0.700 (P = 0.003) for younger patients (CA < 60 years); and 0.584 and 0.570 (P = 0.268) for older patients (CA ≥ 60 years). Conclusions: AI-predicted age using 12-lead ECGs showed superiority in predicting cardiovascular events compared with CA in younger patients, but not in older patients.

Identifying patients with atrial fibrillation during sinus rhythm on ECG: Significance of the labeling in the artificial intelligence algorithm.

BACKGROUND: This study aimed to increase the knowledge on how to enhance the performance of artificial intelligence (AI)-enabled electrocardiography (ECG) to detect atrial fibrillation (AF) on sinus rhythm ECG (SR-ECG). METHODS: It is a retrospective analysis of a single-center, prospective cohort study (Shinken Database). We developed AI-enabled ECG using SR-ECG to predict AF with a convolutional neural network (CNN). Among new patients in our hospital (n = 19,170), 276 AF label (having ECG on AF [AF-ECG] in the ECG database) and 1896 SR label with following three conditions were identified in the derivation dataset: (1) without structural heart disease, (2) in AF label, SR-ECG was taken within 31 days from AF-ECG, and (3) in SR label, follow-up ≥ 1,095 days. Three patterns of AF label were analyzed by timing of SR-ECG to AF-ECG (before/after/before-or-after, CNN algorithm 1 to 3). The outcome measurement was area under the curve (AUC), sensitivity, specificity, accuracy, and F1 score. As an extra-testing dataset, the performance of AI-enabled ECG was tested in patients with structural heart disease. RESULTS: The AUC of AI-enabled ECG with CNN algorithm 1, 2, and 3 in the derivation dataset was 0.83, 0.88, and 0.86, respectively; when tested in patients with structural heart disease, 0.75, 0.81, and 0.78, respectively. CONCLUSION: We confirmed high performance of AI-enabled ECG to detect AF on SR-ECG in patients without structural heart disease. The performance enhanced especially when SR-ECG after index AF-ECG was included in the algorithm, which was consistent in patients with structural heart disease.

Prediction of biological age and all-cause mortality by 12-lead electrocardiogram in patients without structural heart disease.

BACKGROUND: There is a well-established relationship between 12-lead electrocardiogram (ECG) and age and mortality. Furthermore, there is increasing evidence that ECG can be used to predict biological age. However, the utility of biological age from ECG for predicting mortality remains unclear. METHODS: This was a single-center cohort study from a cardiology specialized hospital. A total of 19,170 patients registered in this study from February 2010 to March 2018. ECG was analyzed in a final 12,837 patients after excluding those with structural heart disease or with pacing beats, atrial or ventricular tachyarrhythmia, or an indeterminate axis (R axis > 180°) on index ECG. The models for biological age were developed by principal component analysis (BA) and the Klemera and Doubal's method (not adjusted for age [BAE] and adjusted for age [BAEC]) using 438 ECG parameters. The predictive capability for all-cause death and cardiovascular death by chronological age (CA) and biological age using the three algorithms were evaluated by receiver operating characteristic analysis. RESULTS: During the mean follow-up period of 320.4 days, there were 55 all-cause deaths and 23 cardiovascular deaths. The predictive capabilities for all-cause death by BA, BAE, and BAEC using area under the curves were 0.731, 0.657, and 0.685, respectively, which were comparable to 0.725 for CA (p = 0.760, 0.141, and 0.308, respectively). The predictive capabilities for cardiovascular death by BA, BAE, and BAEC were 0.682, 0.685, and 0.692, respectively, which were also comparable to 0.674 for CA (p = 0.775, 0.839, and 0.706, respectively). In patients aged 60-74 years old, the area under the curves for all-cause death by BA, BAE, and BAEC were 0.619, 0.702, and 0.697, respectively, which tended to be or were significantly higher than 0.482 for CA (p = 0.064, 0.006, and 0.005, respectively). CONCLUSION: Biological age by 12-lead ECG showed a similar predictive capability for mortality compared to CA among total patients, but partially showed a significant increase in predictive capability among patients aged 60-74 years old.

Left atrial dimension and ischemic stroke in patients with and without atrial fibrillation.

The incidence of ischemic stroke (IS) increases in patients with enlarged left atrium (LA) irrespective of whether or not the existence of atrial fibrillation (AF). In such situation, it is unclear whether the impact of LA on incidence of IS still significant in young, non-AF patients with enlarged LA who are primarily unconcerned on anticoagulation therapy. The study population consisted of 18,511 consecutive patients not receiving oral anticoagulants and undergoing echocardiography with measurement of LAD at baseline. The incidence rate of ischemic stroke was calculated in 3 groups according to left atrial dimension (LAD; < 30, 30-45 and ≥ 45 mm) in AF and non-AF patients. Further subgroup analysis was performed in stratification by elderly and young (aged ≥ 65 and < 65 years, respectively). The incidences of IS (per 100 patient-years) were 0.11 and 0.71 in non-AF and AF patients with LAD < 30 mm, respectively, which increased to 0.58 and 1.35 in LAD ≥ 45 mm (adjusted hazard ratios [HRs]; 1.95 [95% confidence intervals, CIs: 0.76-5.01] and 1.22 [95% CIs: 0.27-5.58], interaction P was 0.246). In non-AF patients, the incidences of IS were 0.30 and 0.04 in elderly and young patients with LAD < 30 mm, which increased to 0.67 and 0.48 in LAD ≥ 45 mm (adjusted HRs; 1.34 [95% CIs: 0.43-4.15] and 4.21 [95% CIs: 0.77-23.12], interaction P was 0.158). The incidence of IS significantly increased with increase of LAD in non-AF, especially in non-AF and young patients, although the difference was not independent of other clinical factors. The impact of LAD on IS was numerically larger in non-AF than in AF, and larger in young and non-AF than in elderly counterpart, although a significant interaction was not observed in this small population. Further studies with large population are necessary to judge whether these population with enlarged LA need antithrombotic therapy.

Relationship between resting 12-lead electrocardiogram and all-cause death in patients without structural heart disease: Shinken Database analysis.

BACKGROUND: Resting 12-lead electrocardiography is widely used for the detection of cardiac diseases. Electrocardiogram readings have been reported to be affected by aging and, therefore, can predict patient mortality. METHODS: A total of 12,837 patients without structural heart disease who underwent electrocardiography at baseline were identified in the Shinken Database among those registered between 2010 and 2017 (n = 19,170). Using 438 electrocardiography parameters, predictive models for all-cause death and cardiovascular (CV) death were developed by a support vector machine (SVM) algorithm. RESULTS: During the observation period of 320.4 days, 55 all-cause deaths and 23 CV deaths were observed. In the SVM prediction model, the mean c-statistics of 10 cross-validation models with training and testing datasets were 0.881 ± 0.027 and 0.927 ± 0.101, respectively, for all-cause death and 0.862 ± 0.029 and 0.897 ± 0.069, respectively for CV death. For both all-cause and CV death, high values of permutation importance in the ECG parameters were concentrated in the QRS complex and ST-T segment. CONCLUSIONS: Parameters acquired from 12-lead resting electrocardiography could be applied to predict the all-cause and CV deaths of patients without structural heart disease. The ECG parameters that greatly contributed to the prediction were concentrated in the QRS complex and ST-T segment.

Prediction of current and new development of atrial fibrillation on electrocardiogram with sinus rhythm in patients without structural heart disease.

BACKGROUND: Diagnosis of atrial fibrillation (AF) based on electrocardiogram (ECG) with sinus rhythm remains a major challenge. Obtaining a panoramic view with hundreds of automatically measured ECG parameters at sinus rhythm on the predictive capability for AF would be informative. METHODS: We used a single-center database of a specialist cardiovascular hospital (Shinken Database 2010-2017; n = 19,170). We analyzed 12,863 index ECGs with sinus rhythm after excluding those showing AF rhythm, other atrial tachyarrhythmia, pacing beat, or indeterminate axis, and those of patients with structural heart diseases. We used 438 automatically measured ECG parameters in the MUSE data management system. The predictive models were developed using random forest algorithm with the 10-fold cross-validation method. RESULTS: In 12,863 index ECGs with sinus rhythm, a predictive capability for current paroxysmal AF (n = 1131) by c-statistics was 0.99981 ± 0.00037 for training dataset and 0.91337 ± 0.00087 for testing dataset, respectively. Excluding AF at baseline (n = 11,732), a predictive capability for newly developed AF (n = 98) by c-statistics was 0.99973 ± 0.00086 for training dataset and 0.99160 ± 0.00038 for testing dataset, respectively. The distribution of parameter importance was mostly similar among P, QRS, and ST-T segment for both current and newly developed AF. CONCLUSIONS: This study intended to provide panoramic information in relation between ECG parameters and AF. The parameter importance of ECG parameters for predicting AF was mostly similar in P, QRS, and ST-T segment in models for both current and future AF.

Analysis of bioMARKer Distribution and Individual Reproducibility Under Rivaroxaban Treatment in Japanese Patients with Non-Valvular Atrial Fibrillation (R-MARK Study, CVI ARO2).

The "on-therapy range" of direct oral anticoagulants is the 90% interval of drug concentration. Previously, we reported the on-therapy range of rivaroxaban in a single-center cohort. The present study aimed to confirm the range and intraindividual reproducibility in a multicenter cohort.Eligible patients with non-valvular atrial fibrillation under rivaroxaban treatment for prevention of ischemic stroke were enrolled from nine institutes in Tokyo, Japan, between June 2016 and May 2017 (n = 324). The first and second (three months later) blood samples both taken within 1-5 hours after rivaroxaban intake were analyzed (n = 219). Plasma concentration of rivaroxaban (PC-Riv) and prothrombin time (PT) with five reagents were measured.The 90% interval of PC-Riv was 47.3-532.9 ng/mL. The 90% interval of PT measured with RecombiPlasTin 2G was 11.8-22.3 seconds, the widest range among the five reagents examined. PC-Riv reproducibility within a 90% interval was evaluated bidirectionally (first-to-second and second-to-first), and 92.4% of samples were reproducible. The change rate (CR) of PC-Riv between two samplings ranged widely, and high CR (≥54.3%, cutoff for predicting non-reproducibility) was predicted by concomitant drugs (non-dihydropyridine calcium antagonist and thiazide) and mitral regurgitation.We reported the on-therapy range of rivaroxaban in a multicenter cohort. This range was consistent with that of a single-center cohort and was highly reproducible within three months in daily clinical practice. However, caution is necessary regarding several factors that may affect the intraindividual variation of PC-Riv.

[Hemiparkinsonism due to a solitary infarction of the right external segment of the globus pallidus: a case report].

Three months prior to presentation, a 76-year-old woman suffered from insomnia and was prescribed some antidepressants and hypnotics. At that time, brain MRI showed no cerebral infarcts. Having developed an action tremor of the left hand, bradykinesia, and unstable gait, she visited our hospital. Neurological examination revealed rigidity of the neck and left limbs, clumsiness of the left hand, action tremor, and decreasing swing of the left arm while walking. 123I-metaiodobenzylguanidine scintigraphy showed no decrease of the heart/mediastinum ratio. The second MRI showed an old cerebral infarct located just in the right external segment of the globus pallidus. Since drug-induced parkinsonism was suspected, paroxetine and trazodone were discontinued, but her symptoms did not improve. We concluded that her hemiparkinsonism was due to the cerebral infarct in the right external segment of the globus pallidus, because her symptoms did not respond to dopamine agonist and L-dopa therapy, and the onset of symptoms corresponded with the time of appearance of the cerebral infarct. This is a rare case that is important for understanding the mechanism of parkinsonism.