Identification of a small thrombus in the left ventricle identified on iodine maps derived from dual-source photon-counting detector CT.

Transthoracic echocardiography is the main imaging modality to diagnose left ventricular thrombus (LVT), but its efficacy in certain cases is suboptimal. We report a patient in whom an LVT, initially unidentified by transthoracic echocardiography, was successfully diagnosed with iodine maps derived from dual-source photon-counting detector CT (DS-PCD-CT). The 64-year-old male was admitted to our institution following myocardial infarction. Although TTE failed to detect this small LVT, iodine maps derived from CT angiography (which was conducted to evaluate the coronary artery stenosis) revealed its presence. Iodine maps derived from DS-PCD-CT collecting data with high temporal resolution are beneficial to diagnose LVTs.

Utility of unsupervised deep learning using a 3D variational autoencoder in detecting inner ear abnormalities on CT images.

BACKGROUND AND PURPOSE: To examine the diagnostic performance of unsupervised deep learning using a 3D variational autoencoder (VAE) for detecting and localizing inner ear abnormalities on CT images. METHOD: Temporal bone CT images of 6663 normal inner ears and 113 malformations were analyzed. For unsupervised learning, 113 images from both the malformation and normal cases were used as test data. Other normal images were used for training. A colored difference map representing differences between input and output images of 3D-VAE and the ratio of colored to total pixel numbers were calculated. Supervised learning was also investigated using a 3D deep residual network and all data were classified as normal or malformation using 10-fold cross-validation. RESULTS: For unsupervised learning, a significant difference in the colored pixel ratio was seen between normal (0.00021 ± 0.00022) and malformation (0.00148 ± 0.00087) cases with an area under the curve of 0.99 (specificity = 92.0%, sensitivity = 99.1%). Upon evaluation of the difference map, abnormal regions were partially and not highlighted in 7% and 0% of the malformations, respectively. For supervised learning, which achieved 99.8% specificity and 90.3% sensitivity, a part of and no abnormal regions were highlighted on interpretation maps in 34% and 8% of the malformations, respectively. Abnormal regions were not highlighted in 4 malformation cases diagnosed as malformations and were highlighted in 6 cases misdiagnosed as normal. CONCLUSIONS: Unsupervised deep learning of 3D-VAE precisely detected inner ear malformations and localized abnormal regions. Supervised learning did not identify whole abnormal regions frequently and basis for diagnosis was sometimes unclear.

Ambient-task combined lighting to regulate autonomic and psychomotor arousal levels without compromising subjective comfort to lighting.

BACKGROUND: Although evidence of both beneficial and adverse biological effects of lighting has accumulated, biologically favorable lighting often does not match subjectively comfortable lighting. By controlling the correlated color temperature (CCT) of ambient lights, we investigated the feasibility of combined lighting that meets both biological requirements and subjective comfort. METHODS: Two types of combined lightings were compared; one consisted of a high-CCT (12000 K) light-emitting diode (LED) panel as the ambient light and a low-CCT (5000 K) LED stand light as the task light (high-low combined lighting), and the other consisted of a low-CCT (4500 K) LED panel as the ambient light and the same low-CCT (5000 K) stand light as the task light (low-low combined lighting) as control. Ten healthy subjects (5 young and 5 elderly) were exposed to the two types of lighting on separate days. Autonomic function by heart rate variability, psychomotor performances, and subjective comfort were compared. RESULTS: Both at sitting rest and during psychomotor workload, heart rate was higher and the parasympathetic index of heart rate variability was lower under the high-low combined lighting than the low-low combined lighting in both young and elderly subject groups. Increased psychomotor alertness in the elderly and improved sustainability of concentration work performance in both age groups were also observed under the high-low combined lighting. However, no significant difference was observed in the visual-analog-scale assessment of subjective comfort between the two types of lightings. CONCLUSIONS: High-CCT ambient lighting, even when used in combination with low-CCT task lighting, could increase autonomic and psychomotor arousal levels without compromising subjective comfort. This finding suggests the feasibility of independent control of ambient and task lighting as a way to achieve both biological function regulation and subjective comfort.

Risk stratification after acute myocardial infarction by amplitude-frequency mapping of cyclic variation of heart rate.

BACKGROUND: Blunted cyclic variation of heart rate (CVHR), measured as a decrease in CVHR amplitude (Acv), predicts mortality risk after acute myocardial infarction (AMI). However, Acv also can be reduced in mild sleep apnea with mild O2 desaturation. We investigated whether Acv's predictive power for post-AMI mortality could be improved by considering the effect of sleep apnea severity. METHODS: In 24-hr ECG in 265,291 participants of the Allostatic State Mapping by Ambulatory ECG Repository project, sleep apnea severity was estimated by the frequency of CVHR (Fcv) measured by an automated algorithm for auto-correlated wave detection by adaptive threshold (ACAT). The distribution of Acv on the Acv-Fcv relation map was modeled by percentile regression, and a function converting Acv into percentile value was developed. In the retrospective cohort of the Enhancing Recovery in Coronary Heart Disease (ENRICHD) study, consisting of 673 survivors and 44 non-survivors after AMI, the mortality predictive power of percentile Acv calculated by the function was compared with that of unadjusted Acv. RESULTS: Among the ALLSTAR ECG data, low Acv values appeared more likely when Fcv was low. The logistic regression analysis for mortality in the ENRICHD cohort showed c-statistics of 0.667 (SE, 0.041), 0.817 (0.035), and 0.843 (0.030) for Fcv, unadjusted Acv, and the percentile Acv, respectively. Compared with unadjusted Acv, the percentile Acv showed a significant net reclassification improvement of 0.90 (95% CI, 0.51-1.42). CONCLUSIONS: The predictive power of Acv for post-AMI mortality is improved by considering its relation to sleep apnea severity estimated by Fcv.

Survival Predictors of Heart Rate Variability After Myocardial Infarction With and Without Low Left Ventricular Ejection Fraction.

BACKGROUND: Heart rate variability (HRV) and heart rate (HR) dynamics are used to predict the survival probability of patients after acute myocardial infarction (AMI), but the association has been established in patients with mixed levels of left ventricular ejection fraction (LVEF). OBJECTIVE: We investigated whether the survival predictors of HRV and HR dynamics depend on LVEF after AMI. METHODS: We studied 687 post-AMI patients including 147 with LVEF ≤35% and 540 with LVEF >35%, of which 23 (16%) and 22 (4%) died during the 25 month follow-up period, respectively. None had an implanted cardioverter-defibrillator. From baseline 24 h ECG, the standard deviation (SDNN), root mean square of successive difference (rMSSD), percentage of successive difference >50 ms (pNN50) of normal-to-normal R-R interval, ultra-low (ULF), very-low (VLF), low (LF), and high (HF) frequency power, deceleration capacity (DC), short-term scaling exponent (α1), non-Gaussianity index (λ25 s), and the amplitude of cyclic variation of HR (Acv) were calculated. RESULTS: The predictors were categorized into three clusters; DC, SDNN, α1, ULF, VLF, LF, and Acv as Cluster 1, λ25 s independently as Cluster 2, and rMSSD, pNN50, and HF as Cluster 3. In univariate analyses, mortality was best predicted by indices belonging to Cluster 1 regardless of LVEF. In multivariate analyses, however, mortality in patients with low LVEF was best predicted by the combinations of Cluster 1 predictors or Cluster 1 and 3 predictors, whereas in patients without low LVEF, it was best predicted by the combinations of Cluster 1 and 2 predictors. CONCLUSION: The mortality risk in post-AMI patients with low LVEF is predicted by indices reflecting decreased HRV or HR responsiveness and cardiac parasympathetic dysfunction, whereas in patients without low LVEF, the risk is predicted by a combination of indices that reflect decreased HRV or HR responsiveness and indicator that reflects abrupt large HR changes suggesting sympathetic involvement.

Redundancy among risk predictors derived from heart rate variability and dynamics: ALLSTAR big data analysis.

BACKGROUND: Many indices of heart rate variability (HRV) and heart rate dynamics have been proposed as cardiovascular mortality risk predictors, but the redundancy between their predictive powers is unknown. METHODS: From the Allostatic State Mapping by Ambulatory ECG Repository project database, 24-hr ECG data showing continuous sinus rhythm were extracted and SD of normal-to-normal R-R interval (SDNN), very-low-frequency power (VLF), scaling exponent α1 , deceleration capacity (DC), and non-Gaussianity λ25s were calculated. The values were dichotomized into high-risk and low-risk values using the cutoffs reported in previous studies to predict mortality after acute myocardial infarction. The rate of multiple high-risk predictors accumulating in the same person was examined and was compared with the rate expected under the assumption that these predictors are independent of each other. RESULTS: Among 265,291 ECG data from the ALLSTAR database, the rates of subjects with high-risk SDNN, DC, VLF, α1 , and λ25s values were 2.95, 2.75, 5.89, 15.75, and 18.82%, respectively. The observed rate of subjects without any high-risk value was 66.68%, which was 1.10 times the expected rate (60.74%). The ratios of observed rate to the expected rate at which one, two, three, four, and five high-risk values accumulate in the same person were 0.73 times (24.10 and 32.82%), 1.10 times (6.56 and 5.99%), 4.26 times (1.87 and 0.44%), 47.66 times (0.63 and 0.013%), and 1,140.66 times (0.16 and 0.00014%), respectively. CONCLUSIONS: High-risk predictors of HRV and heart rate dynamics tend to cluster in the same person, indicating a high degree of redundancy between them.

Optimal length of R-R interval segment window for Lorenz plot detection of paroxysmal atrial fibrillation by machine learning.

BACKGROUND: Heartbeat interval Lorenz plot (LP) imaging is a promising method for detecting atrial fibrillation (AF) in long-term monitoring, but the optimal segment window length for the LP images is unknown. We examined the performance of AF detection by LP images with different segment window lengths by machine learning with convolutional neural network (CNN). LP images with a 32 × 32-pixel resolution of non-overlapping segments with lengths between 10 and 500 beats were created from R-R intervals of 24-h ECG in 52 patients with chronic AF and 58 non-AF controls as training data and in 53 patients with paroxysmal AF and 52 non-AF controls as test data. For each segment window length, discriminant models were made by fivefold cross-validation subsets of the training data and its classification performance was examined with the test data. RESULTS: In machine learning with the training data, the averages of cross-validation scores were 0.995 and 0.999 for 10 and 20-beat LP images, respectively, and > 0.999 for 50 to 500-beat images. The classification of test data showed good performance for all segment window lengths with an accuracy from 0.970 to 0.988. Positive likelihood ratio for detecting AF segments, however, showed a convex parabolic curve linear relationship to log segment window length and peaked at 85 beats, while negative likelihood ratio showed monotonous increase with increasing segment window length. CONCLUSIONS: This study suggests that the optimal segment window length that maximizes the positive likelihood ratio for detecting paroxysmal AF with 32 × 32-pixel LP image is 85 beats.

Association of heart rate variability with regional difference in senility death ratio: ALLSTAR big data analysis.

OBJECTIVES: Senility death is defined as natural death in the elderly who do not have a cause of death to be described otherwise and, if human life is finite, it may be one of the ultimate goals of medicine and healthcare. A recent survey in Japan reports that municipalities with a high senility death ratio have lower healthcare costs per late-elderly person. However, the causes of regional differences in senility death ratio and their biomedical determinants were unknown. In this study, we examined the relationships of the regional difference in senility death ratio with the regional differences in heart rate variability and physical activity. METHODS: We compared the age-adjusted senility death ratio of all Japanese prefectures with the regional averages of heart rate variability and actigraphic physical activity obtained from a physiological big data of Allostatic State Mapping by Ambulatory ECG Repository (ALLSTAR). RESULTS: The age-adjusted senility death ratio of 47 Japanese prefectures in 2015 ranged from 1.2% to 3.6% in men and from 3.5% to 7.8% in women. We compared these ratios with the age-adjusted indices of heart rate variability in 108,865 men and 136,536 women and of physical activity level in 16,661 men and 21,961 women. Heart rate variability indices and physical activity levels that are known to be associated with low mortality risk were higher in prefectures with higher senility death ratio. CONCLUSION: The regional senility death ratio in Japan may be associated with regional health status as reflected in heart rate variability and physical activity levels.

Multi-scale heart rate dynamics detected by phase-rectified signal averaging predicts mortality after acute myocardial infarction.

AIMS: Acceleration and deceleration capacity (AC and DC) for beat-to-beat short-term heart rate dynamics are powerful predictors of mortality after acute myocardial infarction (AMI). We examined if AC and DC for minute-order long-term heart rate dynamics also have independent predictive value. METHODS AND RESULTS: We studied 24-hr Holter electrcardiograms in 708 post-AMI patients who were followed up for up to 30 months thereafter. Acceleration capacity and DC was calculated with the time scales of T (window size defining heart rate) and s (wavelet scale) from 1 to 500 s and compared their prognostic values with conventional measures (AC(conv) and DC(conv)) that were calculated with (T,s) = [1,2 (beat)]. During the follow-up, 47 patients died. Both increased AC(conv) and decreased DC(conv) predicted mortality (C statistic, 0.792 and 0.797). Concordantly, sharp peaks of C statistics were observed at (T,s) = [2,7 (sec)] for both increased AC and decreased DC (0.762 and 0.768), but there were larger peaks of C statistics at around [30,60 (sec)] for both (0.783 and 0.796). The C statistic was greater for DC than AC at (30,60) (P = 0.0012). Deceleration capacity at (30,60) was a significant predictor even after adjusted for AC(conv) (P = 0.020) and DC(conv) (P = 0.028), but the predictive power of AC at (30,60) was no longer significant. CONCLUSION: A decrease in DC for minute-order long-term heart rate dynamics is a strong predictor for post-AMI mortality and the predictive power is independent of AC(conv) and DC(conv) for beat-to-beat short-term heart rate dynamics.

Nonlinear measures of heart rate variability and mortality risk in hemodialysis patients.

BACKGROUND AND OBJECTIVES: Nonlinear measures of heart rate variability (HRV) have gained recent interest as powerful risk predictors in various clinical settings. This study examined whether they improve risk stratification in hemodialysis patients. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS: To assess heart rate turbulence, deceleration capacity, fractal scaling exponent (α(1)), and other conventional HRV measures, 281 hemodialysis patients underwent 24-hour electrocardiography between January 2002 and May 2004 and were subsequently followed up. RESULTS: During a median 87-month follow-up, 77 patients (27%) died. Age, left ventricular ejection fraction, serum albumin, C-reactive protein, and calcium × phosphate independently predicted mortality. Whereas all nonlinear HRV measures predicted mortality, only decreased scaling exponent α(1) remained significant after adjusting for clinical risk factors (hazard ratio per a 0.25 decrement, 1.46; 95% confidence interval [95% CI], 1.16-1.85). The inclusion of α(1) into a prediction model composed of clinical risk factors increased the C statistic from 0.84 to 0.87 (P=0.03), with 50.8% (95% CI, 20.2-83.7) continuous net reclassification improvement for 5-year mortality. The predictive power of α(1) showed an interaction with age (P=0.02) and was particularly strong in patients aged <70 years (n=208; hazard ratio, 1.87; 95% CI, 1.38-2.53), among whom α(1) increased the C statistic from 0.85 to 0.89 (P=0.01), with a 93.1% (95% CI, 59.3-142.0) continuous net reclassification improvement. CONCLUSIONS: Scaling exponent α(1) that reflects fractal organization of short-term HRV improves risk stratification for mortality when added to the prediction model by conventional risk factors in hemodialysis patients, particularly those aged <70 years.