Automatic 3D left atrial strain extraction framework on cardiac computed tomography.

BACKGROUND AND OBJECTIVE: Strain analysis provides insights into myocardial function and cardiac condition evaluation. However, the anatomical characteristics of left atrium (LA) inherently limit LA strain analysis when using echocardiography. Cardiac computed tomography (CT) with its superior spatial resolution, has become critical for in-depth evaluation of LA function. Recent studies have explored the feasibility of CT-derived strain; however, they relied on manually selected regions of interest (ROIs) and mainly focused on left ventricle (LV). This study aimed to propose a first-of-its-kind fully automatic deep learning (DL)-based framework for three-dimensional (3D) LA strain extraction on cardiac CT. METHODS: A total of 111 patients undergoing ECG-gated contrast-enhanced CT for evaluating subclinical atrial fibrillation (AF) were enrolled in this study. We developed a 3D strain extraction framework on cardiac CT images, containing a 2.5D GN-U-Net network for LA segmentation, axis-oriented 3D view extraction, and LA strain measure. The segmentation accuracy was evaluated using Dice similarity coefficient (DSC). The model-extracted LA volumes and emptying fraction (EF) were compared with ground-truth measurements using intraclass correlation coefficient (ICC), correlation coefficient (r), and Bland-Altman plot (B-A). The automatically extracted LA strains were evaluated against the LA strains measured from 2D echocardiograms. We utilized this framework to gauge the effect of AF burden on LA strain, employing the atrial high rate episode (AHRE) burden as the measurement parameter. RESULTS: The GN-U-Net LA segmentation network achieved a DSC score of 0.9603 on the test set. The framework-extracted LA estimates demonstrated excellent ICCs of 0.949 (95 % CI: 0.93-0.97) for minimal LA volume, 0.904 (95 % CI: 0.86-0.93) for maximal LA volume, and 0.902 (95 % CI: 0.86-0.93) for EF, compared with expert measurements. The framework-extracted LA strains demonstrated moderate agreement with the LA strains based on 2D echocardiography (ICCs >0.703). Patients with AHRE > 6 min had significantly lower global strain and LAEF, as extracted by the framework than those with AHRE ≤ 6 min. CONCLUSION: The promising results highlighted the feasibility and clinical usefulness of automatically extracting 3D LA strain from CT images using a DL-based framework. This tool could provide a 3D-based alternative to echocardiography for assessing LA function.

A multi-label transformer-based deep learning approach to predict focal visual field progression.

PURPOSE: Tracking functional changes in visual fields (VFs) through standard automated perimetry remains a clinical standard for glaucoma diagnosis. This study aims to develop and evaluate a deep learning (DL) model to predict regional VF progression, which has not been explored in prior studies. METHODS: The study included 2430 eyes of 1283 patients with four or more consecutive VF examinations from the baseline. A multi-label transformer-based network (MTN) using longitudinal VF data was developed to predict progression in six VF regions mapped to the optic disc. Progression was defined using the mean deviation (MD) slope and calculated for all six VF regions, referred to as clusters. Separate MTN models, trained for focal progression detection and forecasting on various numbers of VFs as model input, were tested on a held-out test set. RESULTS: The MTNs overall demonstrated excellent macro-average AUCs above 0.884 in detecting focal VF progression given five or more VFs. With a minimum of 6 VFs, the model demonstrated superior and more stable overall and per-cluster performance, compared to 5 VFs. The MTN given 6 VFs achieved a macro-average AUC of 0.848 for forecasting progression across 8 VF tests. The MTN also achieved excellent performance (AUCs ≥ 0.86, 1.0 sensitivity, and specificity ≥ 0.70) in four out of six clusters for the eyes already with severe VF loss (baseline MD ≤ - 12 dB). CONCLUSION: The high prediction accuracy suggested that multi-label DL networks trained with longitudinal VF results may assist in identifying and forecasting progression in VF regions.

Machine-learning models are superior to severity scoring systems for the prediction of the mortality of critically ill patients in a tertiary medical center.

BACKGROUND: Intensive care unit (ICU) mortality prediction helps to guide therapeutic decision making for critically ill patients. Several scoring systems based on statistical techniques have been developed for this purpose. In this study, we developed a machine-learning model to predict patient mortality in the very early stage of ICU admission. METHODS: This study was performed with data from all patients admitted to the intensive care units of a tertiary medical center in Taiwan from 2009 to 2018. The patients' comorbidities, co-medications, vital signs, and laboratory data on the day of ICU admission were obtained from electronic medical records. We constructed random forest and extreme gradient boosting (XGBoost) models to predict ICU mortality, and compared their performance with that of traditional scoring systems. RESULTS: Data from 12,377 patients was allocated to training (n = 9901) and testing (n = 2476) datasets. The median patient age was 70.0 years; 9210 (74.41%) patients were under mechanical ventilation in the ICU. The areas under receiver operating characteristic curves for the random forest and XGBoost models (0.876 and 0.880, respectively) were larger than those for the Acute Physiology and Chronic Health Evaluation II score (0.738), Sequential Organ Failure Assessment score (0.747), and Simplified Acute Physiology Score II (0.743). The fraction of inspired oxygen on ICU admission was the most important predictive feature across all models. CONCLUSION: The XGBoost model most accurately predicted ICU mortality and was superior to traditional scoring systems. Our results highlight the utility of machine learning for ICU mortality prediction in the Asian population.

Monitoring the Progression of Clinically Suspected Microbial Keratitis Using Convolutional Neural Networks.

Purpose: For this study, we aimed to determine whether a convolutional neural network (CNN)-based method (based on a feature extractor and an identifier) can be applied to monitor the progression of keratitis while managing suspected microbial keratitis (MK). Methods: This multicenter longitudinal cohort study included patients with suspected MK undergoing serial external eye photography at the 5 branches of Chang Gung Memorial Hospital from August 20, 2000, to August 19, 2020. Data were primarily analyzed from January 1 to March 25, 2022. The CNN-based model was evaluated via F1 score and accuracy. The area under the receiver operating characteristic curve (AUROC) was used to measure the precision-recall trade-off. Results: The model was trained using 1456 image pairs from 468 patients. In comparing models via only training the identifier, statistically significant higher accuracy (P < 0.05) in models via training both the identifier and feature extractor (full training) was verified, with 408 image pairs from 117 patients. The full training EfficientNet b3-based model showed 90.2% (getting better) and 82.1% (becoming worse) F1 scores, 87.3% accuracy, and 94.2% AUROC for 505 getting better and 272 becoming worse test image pairs from 452 patients. Conclusions: A CNN-based approach via deep learning applied in suspected MK can monitor the progress/regress during treatment by comparing external eye image pairs. Translational Relevance: The study bridges the gap between the investigation of the state-of-the-art CNN-based deep learning algorithm applied in ocular image analysis and the clinical care of suspected patients with MK.

PoEMS: Policy Network-Based Early Warning Monitoring System for Sepsis in Intensive Care Units.

Sepsis is among the leading causes of morbidity and mortality in modern intensive care units (ICU). Due to accurate and early warning, the in-time antibiotic treatment of sepsis is critical for improving sepsis outcomes, contributing to saving lives, and reducing medical costs. However, the earlier prediction of sepsis onset is made, the fewer monitoring measurements can be processed, causing a lower prediction accuracy. In contrast, a more accurate prediction can be expected by analyzing more data but leading to the delayed warning associated with life-threatening events. In this study, we propose a novel deep reinforcement learning framework for solving early prediction of sepsis, called the Policy Network-based Early Warning Monitoring System (PoEMS). The proposed PoEMS provides accurate and early prediction results for sepsis onset based on analyzing varied-length electronic medical records (EMR). Furthermore, the system serves by monitoring the patient's health status consistently and provides an early warning only when a high risk of sepsis is detected. Additionally, a controlling parameter is designed for users to adjust the trade-off between earliness and accuracy, providing the adaptability of the model to meet various medical requirements in practical scenarios. Through a series of experiments on real-world medical data, the results demonstrate that our proposed PoEMS achieves a high AUROC result of more than 91% for early prediction, and predicts sepsis onset earlier and more accurately compared to other state-of-the-art competing methods.

Deep learning-based automatic left atrial appendage filling defects assessment on cardiac computed tomography for clinical and subclinical atrial fibrillation patients.

Rationale and objectives: Selecting region of interest (ROI) for left atrial appendage (LAA) filling defects assessment can be time consuming and prone to subjectivity. This study aimed to develop and validate a novel artificial intelligence (AI), deep learning (DL) based framework for automatic filling defects assessment on CT images for clinical and subclinical atrial fibrillation (AF) patients. Materials and methods: A total of 443,053 CT images were used for DL model development and testing. Images were analyzed by the AI framework and expert cardiologists/radiologists. The LAA segmentation performance was evaluated using Dice coefficient. The agreement between manual and automatic LAA ROI selections was evaluated using intraclass correlation coefficient (ICC) analysis. Receiver operating characteristic (ROC) curve analysis was used to assess filling defects based on the computed LAA to ascending aorta Hounsfield unit (HU) ratios. Results: A total of 210 patients (Group 1: subclinical AF, n = 105; Group 2: clinical AF with stroke, n = 35; Group 3: AF for catheter ablation, n = 70) were enrolled. The LAA volume segmentation achieved 0.931-0.945 Dice scores. The LAA ROI selection demonstrated excellent agreement (ICC ≥0.895, p < 0.001) with manual selection on the test sets. The automatic framework achieved an excellent AUC score of 0.979 in filling defects assessment. The ROC-derived optimal HU ratio threshold for filling defects detection was 0.561. Conclusion: The novel AI-based framework could accurately segment the LAA region and select ROIs while effectively avoiding trabeculae for filling defects assessment, achieving close-to-expert performance. This technique may help preemptively detect the potential thromboembolic risk for AF patients.

Identification of Coronary Culprit Lesion in ST Elevation Myocardial Infarction by Using Deep Learning.

OBJECTIVE: Early revascularization of the occluded coronary artery in patients with ST elevation myocardial infarction (STEMI) has been demonstrated to decrease mortality and morbidity. Currently, physicians rely on features of electrocardiograms (ECGs) to identify the most likely location of coronary arteries related to an infarct. We sought to predict these culprit arteries more accurately by using deep learning. METHODS: A deep learning model with a convolutional neural network (CNN) that incorporated ECG signals was trained on 384 patients with STEMI who underwent primary percutaneous coronary intervention (PCI) at a medical center. The performances of various signal preprocessing methods (short-time Fourier transform [STFT] and continuous wavelet transform [CWT]) with different lengths of input ECG signals were compared. The sensitivity and specificity for predicting each infarct-related artery and the overall accuracy were evaluated. RESULTS: ECG signal preprocessing with STFT achieved fair overall prediction accuracy (79.3%). The sensitivity and specificity for predicting the left anterior descending artery (LAD) as the culprit vessel were 85.7% and 88.4%, respectively. The sensitivity and specificity for predicting the left circumflex artery (LCX) were 37% and 99%, respectively, and the sensitivity and specificity for predicting the right coronary artery (RCA) were 88.4% and 82.4%, respectively. Using CWT (Morlet wavelet) for signal preprocessing resulted in better overall accuracy (83.7%) compared with STFT preprocessing. The sensitivity and specificity were 93.46% and 80.39% for LAD, 56% and 99.7% for LCX, and 85.9% and 92.9% for RCA, respectively. CONCLUSION: Our study demonstrated that deep learning with a CNN could facilitate the identification of the culprit coronary artery in patients with STEMI. Preprocessing ECG signals with CWT was demonstrated to be superior to doing so with STFT.

Continual learning with attentive recurrent neural networks for temporal data classification.

Continual learning is an emerging research branch of deep learning, which aims to learn a model for a series of tasks continually without forgetting knowledge obtained from previous tasks. Despite receiving a lot of attention in the research community, temporal-based continual learning techniques are still underutilized. In this paper, we address the problem of temporal-based continual learning by allowing a model to continuously learn on temporal data. To solve the catastrophic forgetting problem of learning temporal data in task incremental scenarios, in this research, we propose a novel method based on attentive recurrent neural networks, called Temporal Teacher Distillation (TTD). TTD solves the catastrophic forgetting problem in an attentive recurrent neural network based on three hypotheses, namely Rotation Hypothesis, Redundant Hypothesis, and Recover Hypothesis. Rotation Hypothesis and Redundant hypotheses could cause the attention shift phenomenon, which degrades the model performance on the learned tasks. Moreover, not considering the Recover Hypothesis increases extra memory usage in continuously training different tasks. Therefore, the proposed TTD based on the above hypotheses complements the inadequacy of the existing methods for temporal-based continual learning. For evaluating the performance of our proposed method in task incremental setting, we use a public dataset, WIreless Sensor Data Mining (WISDM), and a synthetic dataset, Split-QuickDraw-100. According to experimental results, the proposed TTD significantly outperforms state-of-the-art methods by up to 14.6% and 45.1% in terms of accuracy and forgetting measures, respectively. To the best of our knowledge, this is the first work that studies continual learning in real-world incremental categories for temporal data classification with attentive recurrent neural networks and provides the proper application-oriented scenario.

Deep Learning Approach in Image Diagnosis of Pseudomonas Keratitis.

This investigation aimed to explore deep learning (DL) models' potential for diagnosing Pseudomonas keratitis using external eye images. In the retrospective research, the images of bacterial keratitis (BK, n = 929), classified as Pseudomonas (n = 618) and non-Pseudomonas (n = 311) keratitis, were collected. Eight DL algorithms, including ResNet50, DenseNet121, ResNeXt50, SE-ResNet50, and EfficientNets B0 to B3, were adopted as backbone models to train and obtain the best ensemble 2-, 3-, 4-, and 5-DL models. Five-fold cross-validation was used to determine the ability of single and ensemble models to diagnose Pseudomonas keratitis. The EfficientNet B2 model had the highest accuracy (71.2%) of the eight single-DL models, while the best ensemble 4-DL model showed the highest accuracy (72.1%) among the ensemble models. However, no statistical difference was shown in the area under the receiver operating characteristic curve and diagnostic accuracy among these single-DL models and among the four best ensemble models. As a proof of concept, the DL approach, via external eye photos, could assist in identifying Pseudomonas keratitis from BK patients. All the best ensemble models can enhance the performance of constituent DL models in diagnosing Pseudomonas keratitis, but the enhancement effect appears to be limited.

Quantitative Prediction of SYNTAX Score for Cardiovascular Artery Disease Patients via the Inverse Problem Algorithm Technique as Artificial Intelligence Assessment in Diagnostics.

The quantitative prediction of the SYNTAX score for cardiovascular artery disease patients using the inverse problem algorithm (IPA) technique in artificial intelligence was explored in this study. A 29-term semi-empirical formula was defined according to seven risk factors: (1) age, (2) mean arterial pressure, (3) body surface area, (4) pre-prandial blood glucose, (5) low-density-lipoprotein cholesterol, (6) Troponin I, and (7) C-reactive protein. Then, the formula was computed via the STATISTICA 7.0 program to obtain a compromised solution for a 405-patient dataset with a specific loss function [actual-predicted]2 as low as 3.177, whereas 0.0 implies a 100% match between the prediction and observation via "the lower, the better" principle. The IPA technique first created a data matrix [405 × 29] from the included patients' data and then attempted to derive a compromised solution of the column matrix of 29-term coefficients [29 × 1]. The correlation coefficient, r2, of the regression line for the actual versus predicted SYNTAX score was 0.8958, showing a high coincidence among the dataset. The follow-up verification based on another 105 patients' data from the same group also had a high correlation coefficient of r2 = 0.8304. Nevertheless, the verified group's low derived average AT (agreement) (ATavg = 0.308 ± 0.193) also revealed a slight deviation between the theoretical prediction from the STATISTICA 7.0 program and the grades assigned by clinical cardiologists or interventionists. The predicted SYNTAX scores were compared with earlier reported findings based on a single-factor statistical analysis or scanned images obtained by sonography or cardiac catheterization. Cardiologists can obtain the SYNTAX score from the semi-empirical formula for an instant referral before performing a cardiac examination.

Artificial Intelligence-Enabled Model for Early Detection of Left Ventricular Hypertrophy and Mortality Prediction in Young to Middle-Aged Adults.

BACKGROUND: Concealed left ventricular hypertrophy (LVH) is a prevalent condition that is correlated with a substantial risk of cardiovascular events and mortality, especially in young to middle-aged adults. Early identification of LVH is warranted. In this work, we aimed to develop an artificial intelligence (AI)-enabled model for early detection and risk stratification of LVH using 12-lead ECGs. METHODS: By deep learning techniques on the ECG recordings from 28 745 patients (20-60 years old), the AI model was developed to detect verified LVH from transthoracic echocardiography and evaluated on an independent cohort. Two hundred twenty-five patients from Japan were externally validated. Cardiologists' diagnosis of LVH was based on conventional ECG criteria. The area under the curve (AUC), sensitivity, and specificity were applied to evaluate the model performance. A Cox regression model estimated the independent effects of AI-predicted LVH on cardiovascular or all-cause death. RESULTS: The AUC of the AI model in diagnosing LVH was 0.89 (sensitivity: 90.3%, specificity: 69.3%), which was significantly better than that of the cardiologists' diagnosis (AUC, 0.64). In the second independent cohort, the diagnostic performance of the AI model was consistent (AUC, 0.86; sensitivity: 85.4%, specificity: 67.0%). After a follow-up of 6 years, AI-predicted LVH was independently associated with higher cardiovascular or all-cause mortality (hazard ratio, 1.91 [1.04-3.49] and 1.54 [1.20-1.97], respectively). The predictive power of the AI model for mortality was consistently valid among patients of different ages, sexes, and comorbidities, including hypertension, diabetes, stroke, heart failure, and myocardial infarction. Last, we also validated the model in the international independent cohort from Japan (AUC, 0.83). CONCLUSIONS: The AI model improved the detection of LVH and mortality prediction in the young to middle-aged population and represented an attractive tool for risk stratification. Early identification by the AI model gives every chance for timely treatment to reverse adverse outcomes.

A Novel Constraint-Based Knee- Guided Neuroevolutionary Algorithm for Context-Specific ECG Early Classification.

Cardiovascular diseases (CVDs) are considered the greatest threat to human life according to World Health Organization. Early classification of CVDs and the appropriate follow-up treatment are crucial for preventing sudden deaths. Electrocardiogram (ECG) is one of the most common non-invasive tools used to evaluate the state of the heart, which can be exploited to automatically diagnose as well. However, the importance of diagnosing CVDs is varying in different context-specific scenarios. For example, ST-segment elevation (STE) is an acute myocardial infarction indicator for patients associated with chest pain and cardiac biomarker. In in-hospital healthcare, STE should be diagnosed with a higher priority than the other phenotypes of ECG. Hence, the context-specific requirements should be considered in ECG early classification problems. We formalize the ECG early classification problem as the context-specific time series classification problem. We propose a novel Constraint-based Knee-guided Neuroevolutionary Algorithm (CKNA) based on the Snippet Policy Networks V2 to solve this problem. To validate the proposed method, we perform a series of experiments on two public ECG datasets under various context-specific simulated scenarios after consulting with physicians specializing in the area. Experimental results show that CKNA significantly improves the average recall of disease classification by 5.5% compared to the competing baseline under user-specified requirements. Moreover, experimental results prove that CKNA presents a feasible solution for the early classifying of cardiac arrhythmias under different user-specified scenarios.

Snippet Policy Network V2: Knee-Guided Neuroevolution for Multi-Lead ECG Early Classification.

Early time series classification predicts the class label of a given time series before it is completely observed. In time-critical applications, such as arrhythmia monitoring in ICU, early treatment contributes to the patient's fast recovery, and early warning could even save lives. Hence, in these cases, it is worthy of trading, to some extent, classification accuracy in favor of earlier decisions when the time series data are collected over time. In this article, we propose a novel deep reinforcement learning-based framework, snippet policy network V2 (SPN-V2), for long and varied-length multi-lead electrocardiogram (ECG) early classification. The proposed SNP-V2 contains two main components: snippet representation learning (SRL) and early classification timing learning (ECTL). The SRL is proposed to encode inner-snippet spatial correlations and inter-snippet temporal correlations into the hidden representations of the subsegment (snippet) of the input ECG. ECTL aims to learn a decision agent to classify the time series early and accurately. To optimize the proposed framework, we design a novel knee-guided neuroevolution algorithm (KGNA) to solve cardiovascular diseases' early classification problem, automatically optimizing the proposed SPN-V2 regarding the tradeoff between accuracy and earliness. In addition, we conduct a series of experiments on two real-world ECG datasets. The experimental results show the superiority of the proposed algorithm over the state-of-the-art competing methods.

Effect of the Nintendo Ring Fit Adventure Exergame on Running Completion Time and Psychological Factors Among University Students Engaging in Distance Learning During the COVID-19 Pandemic: Randomized Controlled Trial.

BACKGROUND: The COVID-19 outbreak has not only changed the lifestyles of people globally but has also resulted in other challenges, such as the requirement of self-isolation and distance learning. Moreover, people are unable to venture out to exercise, leading to reduced movement, and therefore, the demand for exercise at home has increased. OBJECTIVE: We intended to investigate the relationships between a Nintendo Ring Fit Adventure (RFA) intervention and improvements in running time, cardiac force index (CFI), sleep quality (Chinese version of the Pittsburgh Sleep Quality Index score), and mood disorders (5-item Brief Symptom Rating Scale score). METHODS: This was a randomized prospective study and included 80 students who were required to complete a 1600-meter outdoor run before and after the intervention, the completion times of which were recorded in seconds. They were also required to fill out a lifestyle questionnaire. During the study, 40 participants (16 males and 24 females, with an average age of 23.75 years) were assigned to the RFA group and were required to exercise for 30 minutes 3 times per week (in the adventure mode) over 4 weeks. The exercise intensity was set according to the instructions given by the virtual coach during the first game. The remaining 40 participants (30 males and 10 females, with an average age of 22.65 years) were assigned to the control group and maintained their regular habits during the study period. RESULTS: The study was completed by 80 participants aged 20 to 36 years (mean 23.20, SD 2.96 years). The results showed that the running time in the RFA group was significantly reduced. After 4 weeks of physical training, it took females in the RFA group 19.79 seconds (P=.03) and males 22.56 seconds (P=.03) less than the baseline to complete the 1600-meter run. In contrast, there were no significant differences in the performance of the control group in the run before and after the fourth week of intervention. In terms of mood disorders, the average score of the RFA group increased from 1.81 to 3.31 for males (difference=1.50, P=.04) and from 3.17 to 4.54 for females (difference=1.38, P=.06). In addition, no significant differences between the RFA and control groups were observed for the CFI peak acceleration (CFIPA)_walk, CFIPA_run, or sleep quality. CONCLUSIONS: RFA could either maintain or improve an individual's physical fitness, thereby providing a good solution for people involved in distance learning or those who have not exercised for an extended period. TRIAL REGISTRATION: ClinicalTrials.gov NCT05227040; https://clinicaltrials.gov/ct2/show/NCT05227040.

Hierarchy-aware contrastive learning with late fusion for skin lesion classification.

BACKGROUND AND OBJECTIVE: The incidence rate of skin cancers is increasing worldwide annually. Using machine learning and deep learning for skin lesion classification is one of the essential research topics. In this study, we formulate a major-type misclassification problem that previous studies did not consider in the multi-class skin lesion classification. Moreover, addressing the major-type misclassification problem is significant for real-world computer-aided diagnosis. METHODS: This study presents a novel method, namely Hierarchy-Aware Contrastive Learning with Late Fusion (HAC-LF), to improve the overall performance of multi-class skin classification. In HAC-LF, we design a new loss function, Hierarchy-Aware Contrastive Loss (HAC Loss), to reduce the impact of the major-type misclassification problem. The late fusion method is applied to balance the major-type and multi-class classification performance. RESULTS: We conduct a series of experiments with the ISIC 2019 Challenges dataset, which consists of three skin lesion datasets, to verify the performance of our methods. The results show that our proposed method surpasses the representative deep learning methods for skin lesion classification in all evaluation metrics used in this study. HAC-LF achieves 0.871, 0.842, 0.889 for accuracy, sensitivity, and specificity in the major-type classification, respectively. With the imbalanced class distribution, HAC-LF outperforms the baseline model regarding the sensitivity of minority classes. CONCLUSIONS: This research formulates a major-type misclassification problem. We propose HAC-LF to deal with it and boost the multi-class skin lesion classification performance. According to the results, the advantage of HAC-LF is that the proposed HAC Loss can beneficially reduce the impact of the major-type misclassification by decreasing the major-type error rate. Besides the medical field HAC-LF is promising to be applied to other domains possessing the data with the hierarchical structure.

A Deep Learning-Enabled Electrocardiogram Model for the Identification of a Rare Inherited Arrhythmia: Brugada Syndrome.

BACKGROUND: Brugada syndrome is a major cause of sudden cardiac death in young people and has distinctive electrocardiographic (ECG) features. We aimed to develop a deep learning-enabled ECG model for automatic screening for Brugada syndrome to identify these patients at an early point in time, thus allowing for life-saving therapy. METHODS: A total of 276 ECGs with a type 1 Brugada ECG pattern (276 type 1 Brugada ECGs and another randomly retrieved 276 non-Brugada type ECGs for 1:1 allocation) were extracted from the hospital-based ECG database for a 2-stage analysis with a deep learning model. After trained network for identifying right bundle branch block pattern, we transferred the first-stage learning to the second task to diagnose the type 1 Brugada ECG pattern. The diagnostic performance of the deep learning model was compared with that of board-certified practicing cardiologists. The model was further validated in an independent ECG data set collected from hospitals in Taiwan and Japan. RESULTS: The diagnoses by the deep learning model (area under the receiver operating characteristic curve [AUC] 0.96, sensitivity 88.4%, specificity 89.1%) were highly consistent with the standard diagnoses (kappa coefficient 0.78). However, the diagnoses by the cardiologists were significantly different from the standard diagnoses, with only moderate consistency (kappa coefficient 0.63). In the independent ECG cohort, the deep learning model still reached a satisfactory diagnostic performance (AUC 0.89, sensitivity 86.0%, specificity 90.0%). CONCLUSIONS: We present the first deep learning-enabled ECG model for diagnosing Brugada syndrome, which appears to be a robust screening tool with a diagnostic potential rivalling trained physicians.

Comparisons of deep learning algorithms for diagnosing bacterial keratitis via external eye photographs.

Bacterial keratitis (BK), a painful and fulminant bacterial infection of the cornea, is the most common type of vision-threatening infectious keratitis (IK). A rapid clinical diagnosis by an ophthalmologist may often help prevent BK patients from progression to corneal melting or even perforation, but many rural areas cannot afford an ophthalmologist. Thanks to the rapid development of deep learning (DL) algorithms, artificial intelligence via image could provide an immediate screening and recommendation for patients with red and painful eyes. Therefore, this study aims to elucidate the potentials of different DL algorithms for diagnosing BK via external eye photos. External eye photos of clinically suspected IK were consecutively collected from five referral centers. The candidate DL frameworks, including ResNet50, ResNeXt50, DenseNet121, SE-ResNet50, EfficientNets B0, B1, B2, and B3, were trained to recognize BK from the photo toward the target with the greatest area under the receiver operating characteristic curve (AUROC). Via five-cross validation, EfficientNet B3 showed the most excellent average AUROC, in which the average percentage of sensitivity, specificity, positive predictive value, and negative predictive value was 74, 64, 77, and 61. There was no statistical difference in diagnostic accuracy and AUROC between any two of these DL frameworks. The diagnostic accuracy of these models (ranged from 69 to 72%) is comparable to that of the ophthalmologist (66% to 74%). Therefore, all these models are promising tools for diagnosing BK in first-line medical care units without ophthalmologists.

Deep Learning for Automated Diabetic Retinopathy Screening Fused With Heterogeneous Data From EHRs Can Lead to Earlier Referral Decisions.

Purpose: Fundus images are typically used as the sole training input for automated diabetic retinopathy (DR) classification. In this study, we considered several well-known DR risk factors and attempted to improve the accuracy of DR screening. Metphods: Fusing nonimage data (e.g., age, gender, smoking status, International Classification of Disease code, and laboratory tests) with data from fundus images can enable an end-to-end deep learning architecture for DR screening. We propose a neural network that simultaneously trains heterogeneous data and increases the performance of DR classification in terms of sensitivity and specificity. In the current retrospective study, 13,410 fundus images and their corresponding nonimage data were collected from the Chung Shan Medical University Hospital in Taiwan. The images were classified as either nonreferable or referable for DR by a panel of ophthalmologists. Cross-validation was used for the training models and to evaluate the classification performance. Results: The proposed fusion model achieved 97.96% area under the curve with 96.84% sensitivity and 89.44% specificity for determining referable DR from multimodal data, and significantly outperformed the models that used image or nonimage information separately. Conclusions: The fusion model with heterogeneous data has the potential to improve referable DR screening performance for earlier referral decisions. Translational Relevance: Artificial intelligence fused with heterogeneous data from electronic health records could provide earlier referral decisions from DR screening.

Convolutional neural network based automatic screening tool for cardiovascular diseases using different intervals of ECG signals.

BACKGROUND AND OBJECTIVE: Automatic screening tools can be applied to detect cardiovascular diseases (CVDs), which are the leading cause of death worldwide. As an effective and non-invasive method, electrocardiogram (ECG) based approaches are widely used to identify CVDs. Hence, this paper proposes a deep convolutional neural network (CNN) to classify five CVDs using standard 12-lead ECG signals. METHODS: The Physiobank (PTB) ECG database is used in this study. Firstly, ECG signals are segmented into different intervals (one-second, two-seconds and three-seconds), without any wave detection, and three datasets are obtained. Secondly, as an alternative to any complex preprocessing, durations of raw ECG signals have been considered as input with simple min-max normalization. Lastly, a ten-fold cross-validation method is employed for one-second ECG signals and also tested on other two datasets (two-seconds and three-seconds). RESULTS: Comparing to the competing approaches, the proposed CNN acquires the highest performance, having an accuracy, sensitivity, and specificity of 99.59%, 99.04%, and 99.87%, respectively, with one-second ECG signals. The overall accuracy, sensitivity, and specificity obtained are 99.80%, 99.48%, and 99.93%, respectively, using two-seconds of signals with pre-trained proposed models. The accuracy, sensitivity, and specificity of segmented ECG tested by three-seconds signals are 99.84%, 99.52%, and 99.95%, respectively. CONCLUSION: The results of this study indicate that the proposed system accomplishes high performance and keeps the characterizations in brief with flexibility at the same time, which means that it has the potential for implementation in a practical, real-time medical environment.

Prenatal exposure to air pollutants and childhood atopic dermatitis and allergic rhinitis adopting machine learning approaches: 14-year follow-up birth cohort study.

The incidence of childhood atopic dermatitis (AD) and allergic rhinitis (AR) is increasing. This warrants development of measures to predict and prevent these conditions. We aimed to investigate the predictive ability of a spectrum of data mining methods to predict childhood AD and AR using longitudinal birth cohort data. We conducted a 14-year follow-up of infants born to pregnant women who had undergone maternal examinations at nine selected maternity hospitals across Taiwan during 2000-2005. The subjects were interviewed using structured questionnaires to record data on basic demographics, socioeconomic status, lifestyle, medical history, and 24-h dietary recall. Hourly concentrations of air pollutants within 1 year before childbirth were obtained from 76 national air quality monitoring stations in Taiwan. We utilized weighted K-nearest neighbour method (k = 3) to infer the personalized air pollution exposure. Machine learning methods were performed on the heterogeneous attributes set to predict allergic diseases in children. A total of 1439 mother-infant pairs were recruited in machine learning analysis. The prevalence of AD and AR in children up to 14 years of age were 6.8% and 15.9%, respectively. Overall, tree-based models achieved higher sensitivity and specificity than other methods, with areas under receiver operating characteristic curve of 83% for AD and 84% for AR, respectively. Our findings confirmed that prenatal air quality is an important factor affecting the predictive ability. Moreover, different air quality indices were better predicted, in combination than separately. Combining heterogeneous attributes including environmental exposures, demographic information, and allergens is the key to a better prediction of children allergies in the general population. Prenatal exposure to nitrogen dioxide (NO2) and its concatenation changes with time were significant predictors for AD and AR till adolescent.