LIRRN: Location-Independent Relative Radiometric Normalization of Bitemporal Remote-Sensing Images.

Relative radiometric normalization (RRN) is a critical pre-processing step that enables accurate comparisons of multitemporal remote-sensing (RS) images through unsupervised change detection. Although existing RRN methods generally have promising results in most cases, their effectiveness depends on specific conditions, especially in scenarios with land cover/land use (LULC) in image pairs in different locations. These methods often overlook these complexities, potentially introducing biases to RRN results, mainly because of the use of spatially aligned pseudo-invariant features (PIFs) for modeling. To address this, we introduce a location-independent RRN (LIRRN) method in this study that can automatically identify non-spatially matched PIFs based on brightness characteristics. Additionally, as a fast and coregistration-free model, LIRRN complements keypoint-based RRN for more accurate results in applications where coregistration is crucial. The LIRRN process starts with segmenting reference and subject images into dark, gray, and bright zones using the multi-Otsu threshold technique. PIFs are then efficiently extracted from each zone using nearest-distance-based image content matching without any spatial constraints. These PIFs construct a linear model during subject-image calibration on a band-by-band basis. The performance evaluation involved tests on five registered/unregistered bitemporal satellite images, comparing results from three conventional methods: histogram matching (HM), blockwise KAZE, and keypoint-based RRN algorithms. Experimental results consistently demonstrated LIRRN's superior performance, particularly in handling unregistered datasets. LIRRN also exhibited faster execution times than blockwise KAZE and keypoint-based approaches while yielding results comparable to those of HM in estimating normalization coefficients. Combining LIRRN and keypoint-based RRN models resulted in even more accurate and reliable results, albeit with a slight lengthening of the computational time. To investigate and further develop LIRRN, its code, and some sample datasets are available at link in Data Availability Statement.

Interferometric orbital angular momentum mode detection in turbulence with deep learning.

Orbital angular momentum (OAM) modes are topical due to their versatility, and they have been used in several applications including free-space optical communication systems. The classification of OAM modes is a common requirement, and there are several methods available for this. One such method makes use of deep learning, specifically convolutional neural networks, which distinguishes between modes using their intensities. However, OAM mode intensities are very similar if they have the same radius or if they have opposite topological charges, and as such, intensity-only approaches cannot be used exclusively for individual modes. Since the phase of each OAM mode is unique, deep learning can be used in conjugation with interferometry to distinguish between different modes. In this paper, we demonstrate a very high classification accuracy of a range of OAM modes in turbulence using a shear interferometer, which crucially removes the requirement of a reference beam. For comparison, we show only marginally higher accuracy with a more conventional Mach-Zehnder interferometer, making the technique a promising candidate towards real-time, low-cost modal decomposition in turbulence.

Machine learning and statistical methods for predicting mortality in heart failure.

Heart failure is a debilitating clinical syndrome associated with increased morbidity, mortality, and frequent hospitalization, leading to increased healthcare budget utilization. Despite the exponential growth in the introduction of pharmacological agents and medical devices that improve survival, many heart failure patients, particularly those with a left ventricular ejection fraction less than 40%, still experience persistent clinical symptoms that lead to an overall decreased quality of life. Clinical risk prediction is one of the strategies that has been implemented for the selection of high-risk patients and for guiding therapy. However, most risk predictive models have not been well-integrated into the clinical setting. This is partly due to inherent limitations, such as creating risk predicting models using static clinical data that does not consider the dynamic nature of heart failure. Another limiting factor preventing clinicians from utilizing risk prediction models is the lack of insight into how predictive models are built. This review article focuses on describing how predictive models for risk-stratification of patients with heart failure are built.

Using text mining techniques to extract prostate cancer predictive information (Gleason score) from semi-structured narrative laboratory reports in the Gauteng province, South Africa.

BACKGROUND: Prostate cancer (PCa) is the leading male neoplasm in South Africa with an age-standardised incidence rate of 68.0 per 100,000 population in 2018. The Gleason score (GS) is the strongest predictive factor for PCa treatment and is embedded within semi-structured prostate biopsy narrative reports. The manual extraction of the GS is labour-intensive. The objective of our study was to explore the use of text mining techniques to automate the extraction of the GS from irregularly reported text-intensive patient reports. METHODS: We used the associated Systematized Nomenclature of Medicine clinical terms morphology and topography codes to identify prostate biopsies with a PCa diagnosis for men aged > 30 years between 2006 and 2016 in the Gauteng Province, South Africa. We developed a text mining algorithm to extract the GS from 1000 biopsy reports with a PCa diagnosis from the National Health Laboratory Service database and validated the algorithm using 1000 biopsies from the private sector. The logical steps for the algorithm were data acquisition, pre-processing, feature extraction, feature value representation, feature selection, information extraction, classification, and discovered knowledge. We evaluated the algorithm using precision, recall and F-score. The GS was manually coded by two experts for both datasets. The top five GS were reported, with the remaining scores categorised as "Other" for both datasets. The percentage of biopsies with a high-risk GS (≥ 8) was also reported. RESULTS: The first output reported an F-score of 0.99 that improved to 1.00 after the algorithm was amended (the GS reported in clinical history was ignored). For the validation dataset, an F-score of 0.99 was reported. The most commonly reported GS were 5 + 4 = 9 (17.6%), 3 + 3 = 6 (17.5%), 4 + 3 = 7 (16.4%), 3 + 4 = 7 (14.7%) and 4 + 4 = 8 (14.2%). For the validation dataset, the most commonly reported GS were: (i) 3 + 3 = 6 (37.7%), (ii) 3 + 4 = 7 (19.4%), (iii) 4 + 3 = 7 (14.9%), (iv) 4 + 4 = 8 (10.0%) and (v) 4 + 5 = 9 (7.4%). A high-risk GS was reported for 31.8% compared to 17.4% for the validation dataset. CONCLUSIONS: We demonstrated reliable extraction of information about GS from narrative text-based patient reports using an in-house developed text mining algorithm. A secondary outcome was that late presentation could be assessed.

Design, synthesis, and molecular docking of novel 3,5-disubstituted-1,3,4-oxadiazole derivatives as iNOS inhibitors.

To obtain new anti-inflammatory agents, recent studies have aimed to replace the carboxylate functionality of nonsteroidal anti-inflammatory drugs with less acidic heterocyclic bioisosteres like 1,3,4-oxadiazole to protect the gastric mucosa from free carboxylate moieties. In view of these observations, we designed and synthesized a series of 3,5-disubstituted-1,3,4-oxadiazole derivatives as inhibitors of prostaglandin E2 (PGE2 ) and NO production with an improved activity profile. As initial screening, and to examine the anti-inflammatory activities of the compounds, the inhibitions of the productions of lipopolysaccharide-induced NO and PGE2 in RAW 264.7 macrophages were evaluated. The biological assays showed that, compared with indomethacin, compounds 5a, 5g, and 5h significantly inhibited NO production with 12.61 ± 1.16, 12.61 ± 1.16, and 18.95 ± 3.57 µM, respectively. Consequently, the three compounds were evaluated for their in vivo anti-inflammatory activities. Compounds 5a, 5g, and 5h showed a potent anti-inflammatory activity profile almost equivalent to indomethacin at the same dose in the carrageenan-induced paw edema test. Moreover, the treatment with 40 mg/kg of 5h produced significant anti-inflammatory activity data. Furthermore, docking studies were performed to reveal possible interactions with the inducible nitric oxide synthase enzyme. Docking results were able to rationalize the biological activity data of the studied inhibitors. In summary, our data suggest that compound 5h is identified as a promising candidate for further anti-inflammatory drug development with an extended safety profile.

Entropy-Based Image Fusion with Joint Sparse Representation and Rolling Guidance Filter.

Image fusion is a very practical technology that can be applied in many fields, such as medicine, remote sensing and surveillance. An image fusion method using multi-scale decomposition and joint sparse representation is introduced in this paper. First, joint sparse representation is applied to decompose two source images into a common image and two innovation images. Second, two initial weight maps are generated by filtering the two source images separately. Final weight maps are obtained by joint bilateral filtering according to the initial weight maps. Then, the multi-scale decomposition of the innovation images is performed through the rolling guide filter. Finally, the final weight maps are used to generate the fused innovation image. The fused innovation image and the common image are combined to generate the ultimate fused image. The experimental results show that our method's average metrics are: mutual information ( M I )-5.3377, feature mutual information ( F M I )-0.5600, normalized weighted edge preservation value ( Q A B / F )-0.6978 and nonlinear correlation information entropy ( N C I E )-0.8226. Our method can achieve better performance compared to the state-of-the-art methods in visual perception and objective quantification.

Behçet's disease and risk of vascular events.

PURPOSE OF REVIEW: We discuss the relationship between several factors and the risk of vascular events in patients with Behçet's disease. RECENT FINDINGS: Behçet's disease, a systemic, chronic relapsing vasculitis, is mainly seen in the Mediterranean area and is typically characterized by recurrent oro-genital ulcers, ocular inflammation, and skin manifestations, including articular, vascular, gastroenteric, and neurological involvement. It is a chronic inflammatory disease with relapses and remissions. The prognosis varies. Behçet's disease can cause venous or arterial lesions. Vascular involvement contributes to the mortality and morbidity associated with Behçet's disease. SUMMARY: The cause of thrombosis or vascular events in Behçet's disease remains incompletely understood; several factors have been studied with conflicting results. Vasculitis is considered to underlie several clinical manifestations of Behçet's disease.

A New Artificial Chordae Implantation Method in Mitral Valve Surgery.

A novel method is described for artificial chordae replacement with expanded polytetrafluoroethylene suture in mitral valve repair procedures. The technique does not involve knots over or beneath the free edge of the mitral valve leaflets. Artificial chords suspend the exact free margin of leaflets as if it were a continuation of the free margin, such that the smooth zone of the coapting area can be preserved. This technique is simple, reproducible, and applicable to both anterior and posterior leaflets. Moreover, the length of the artificial chords can be adjusted rapidly and accurately at the first attempt.

The relation between monocyte to HDL ratio and no-reflow phenomenon in the patients with acute ST-segment elevation myocardial infarction.

BACKGROUND: No-reflow phenomenon is a prognostic value in ST-segment elevation myocardial infarction (STEMI). Monocyte to high density lipoprotein ratio (MHR) has recently emerged as a marker of inflammation and oxidative stress in the cardiovascular disease. PURPOSE: In this study, we aimed to investigate the relation between MHR and no-reflow phenomenon in patients with STEMI undergoing primary percutaneous coronary intervention (pPCI). MATERIAL AND METHODS: A total of 600 patients with STEMI (470 men; mean age, 62 ± 12 years) admitted within 12 hours from symptom onset were included into this study. Patients were classified into 2 groups based on postintervention Thrombolysis in Myocardial Infarction (TIMI) flow grade: no-reflow-TIMI flow grade 0, 1, or 2 (group 1); angiographic success-TIMI flow grade 3 (group 2). RESULTS: According to admission whole-blood cell count results, the patients in the no-reflow group had significantly higher monocyte count and MHR values when compared with those of the reflow patients. After multivariate backward logistic regression, MHR remained independent predictors of no reflow after pPCI. Adjusted odds ratios were calculated as 1.09 for MHR (P< .001; confidence interval [CI], 1.07-1.12). Receiver operating characteristic curve analysis suggested that the optimum MHR level cutoff point for patients with no-reflow was 22.5, with a sensitivity and specificity of 70.2% and 73.3%, respectively (area under curve, 0.768; 95% CI, 0.725-0.811). CONCLUSION: In conclusion, MHR levels are one of the independent predictors of no reflow in patients with STEMI after pPCI.

Endocan, a novel marker of endothelial dysfunction in patients with essential hypertension: comparative effects of amlodipine and valsartan.

Vascular inflammation plays an important role in the pathophysiology of hypertension and high levels of endocan may reflect ongoing vascular inflammation in hypertensive patients. In the present hypothesis-generating study, we aimed at investigating the comparative effects of amlodipine and valsartan on endocan levels in newly diagnosed hypertensive patients. The study population consisted of 37 untreated hypertensive patients who were randomized to the two treatment arms. After baseline assessment, each patient was randomly allocated to either 10 mg daily of amlodipine (n = 18, 7 males) or 160 mg daily of valsartan (n = 19, 3 males) and treated for a 3-month period. Sphygmomanometric blood pressure (BP) and serum endocan were measured before and every 2 weeks during drug treatment. There was no statistically significant difference between the two treatment arms as far as baseline socio-demographic and clinical characteristics are concerned. After a 3-month treatment period, systolic and diastolic BP values significantly reduced by antihypertensive treatment (p < 0.001). Furthermore, endocan levels were significantly decreased in both treatment arms (p < 0.05). However, amlodipine caused a greater percent decrease in circulating endocan levels compared with valsartan at the end of the treatment period. Both drugs reduced high sensitivity C-reactive protein values. However, the statistical significant difference vs baseline was achieved only in the group treated with amlodipine. No correlation was found between endocan plasma levels and BP reduction. The results of this hypothesis-generating study suggest that amlodipine and valsartan decrease endocan levels in newly diagnosed hypertensive patients. The effects, which are more evident with amlodipine, may contribute to the anti-inflammatory effects exerted by the two drugs on the vascular target.

Increased frequency of mitral valve prolapse in patients with deviated nasal septum.

Any abnormality of collagen may affect the tissues with higher collagen content, e.g., joints, heart valves, and great arteries. Mitral valve prolapse (MVP) is a characteristic of generalized collagen abnormality. Nasal septum (NS) is constituted by osseous and cartilaginous septums that are highly rich in collagen. We evaluated the co-existence of deviation of NS (DNS) in patients with MVP. We retrospectively evaluated the recordings of echocardiographic and nasal examinations of subjects with MVP and DNS. We analyzed the features of MVP and anatomical classification of DNS among subjects. Totally, 74 patients with DNS and 38 subjects with normal nasal passage were enrolled to the study. Presence of MVP was significantly higher in patients with DNS compared to normal subjects (63 vs 26%, p < 0.001). Prolapse of anterior, posterior and both leaflets was higher in patients with DNS. Thickness of anterior mitral leaflet was significantly increased in patients with DNS (3.57 ± 0.68 vs 4.59 ± 1.1 mm, p < 0.001) compared to normal subjects. Type I, II, and III, IV DNS were higher in frequency in patients with MVP while type V and VI were higher in normal subjects. DNS is highly co-existent with MVP and increased thickness of mitral anterior leaflet. Generalized abnormality of collagen which is the main component of mitral valves and nasal septum may be accounted for co-existence of MVP and DNS. Also co-existence of them may exaggerate the symptoms of patients with MVP due to limited airflow through the nasal passage.

Serum endocan levels as a marker of disease activity in patients with Behçet disease.

BACKGROUND: Endocan is a novel human endothelial cell-specific molecule. The central role of leukocytes and endothelial dysfunction in the development of Behçet disease (BD) led us to hypothesize that endocan might be a marker of this disease. OBJECTIVE: We investigated the relationship between serum levels of endocan and disease activity in patients with BD. METHODS: In all, 33 patients (16 active, 17 inactive) with BD and 35 healthy persons were included in the study. Endocan and C-reactive protein were measured in all subjects. RESULTS: Patients with BD had significantly higher serum endocan levels. Mean serum levels of endocan were 1.29 ± 0.60 ng/mL (range: 0.58-2.99) in patients with BD and 0.75 ± 0.16 ng/mL (range: 0.48-1.21) in control subjects (P < .001). In patients with BD, serum endocan levels correlated moderately but significantly with C-reactive protein, erythrocyte sedimentation rate, and disease activity. Receiver operating characteristic curve analysis suggested that the optimum endocan level cut-off point for patients with BD was 0.87 ng/mL, with a sensitivity and specificity of 75.8% and 80%, respectively (area under curve 0.835, 95% confidence interval 0.738-0.932). LIMITATIONS: The main limitation of our study is the relatively small sample size. CONCLUSIONS: Circulating endocan may be a marker of BD activity.

Why aortic elasticity differs among classical and non-classical mitral valve prolapsed?

BACKGROUND: Mitral valve prolapse (MVP) is the most common valvular heart disease and characterized by the displacement of an abnormally thickened mitral valve leaflet into the left atrium during systole. There are two types of MVP, broadly classified as classic (thickness ≥5 mm) and non-classic (thickness <5 mm) according to the morphology of the leaflets. We aimed to investigate elastic properties of the aorta in young male patients with classical and non-classical MVP. MATERIAL/METHODS: In the present study, 63 young adult males (mean age: 22.7 ± 4.2) were included. Patients were divided into classic MVP (n = 27) and non-classic MVP (n = 36) groups. Aortic strain, aortic distensibility and aortic stiffness index were calculated by using aortic diameters obtained by echocardiography and blood pressures measured by sphygmomanometer. RESULTS: There was no significant difference between the groups in terms of age, body mass index, left ventricular mass and ejection fraction. When comparing the MVP group it was found that aortic strain and aortic distensibility were increased (p = 0.0027, p = 0.016, respectively) whereas the aortic stiffness index was decreased (p = 0.06) in the classical MVP group. CONCLUSION: We concluded that the elastic properties of the aorta is increased in patients with classic MVP. Further large scale studies should be performed to understand of morphological and physiological properties of the aorta in patients with MVP.

QT dispersion increases with low glomerular filtration rate in patients with coronary artery disease.

OBJECTIVE: We aimed to evaluate the relationship between estimated glomerular filtration rate (eGFR) and QT dispersion (QTd) in patients with coronary artery disease (CAD). METHODS: Sixty patients(mean age 62.72 ± 12.48 years) included 46 male, (mean age 60.89 ± 12.70 years)and 14 female (mean age 68.71± 9.86 years) were enrolled in this study. Patients were divided into 2 groups according to their eGFR using the 6 variable MDRD equation. Group 1 consisted of patients with estimated eGFR<60 ml/min/1.73m(2) and Group 2 consisted of patients witheGFR ≥ 60 ml/min/1.73m(2). RESULTS: Baseline patient characteristics were homogeneous in both groups except for age, gender and smoking.Also, the extent of CAD was similar in both groups (p > 0.05) QTd values were found higher in group 1 than those of group 2 (57.23 ± 40.65 ms vs. 31.23 ± 14.47 ms, p = 0.002). After adjustment for age, gender and smoking using one-way ANCOVA test, statistically significant difference in QTd still existedbetween the groups (p=0.038). CONCLUSION: QTd tends to be higher in patients with poor renal function independent of severity of angiographical CAD. QTd may be a potentially useful non-invasive test in the management of patients with poor renal function, especially those with CAD.

Evaluating radiofrequency electromagnetic field exposure in confined spaces: a systematic review of recent studies and future directions.

This study reviews recent research on Radiofrequency Electromagnetic Field (RF-EMF) exposure in confined environments, focusing on methodologies and parameters. Studies typically evaluate RF-EMF exposure using an electric field and specific absorption rate but fail to consider temperature rise in the tissues in confined environments. The study highlights the investigation of RF-EMF exposure in subterranean environments such as subways, tunnels and mines. Future research should evaluate the exposure of communication devices in such environments, considering the surrounding environment. Such studies will aid in understanding the risks and developing effective mitigation strategies to protect workers and the general public.

Deep learning for blood glucose level prediction: How well do models generalize across different data sets?

Deep learning-based models for predicting blood glucose levels in diabetic patients can facilitate proactive measures to prevent critical events and are essential for closed-loop control therapy systems. However, selecting appropriate models from the literature may not always yield conclusive results, as the choice could be influenced by biases or misleading evaluations stemming from different methodologies, datasets, and preprocessing techniques. This study aims to compare and comprehensively analyze the performance of various deep learning models across diverse datasets to assess their applicability and generalizability across a broader spectrum of scenarios. Commonly used deep learning models for blood glucose level forecasting, such as feed-forward neural network, convolutional neural network, long short-term memory network (LSTM), temporal convolutional neural network, and self-attention network (SAN), are considered in this study. To evaluate the generalization capabilities of each model, four datasets of varying sizes, encompassing samples from different age groups and conditions, are utilized. Performance metrics include Root Mean Square Error (RMSE), Mean Absolute Difference (MAD), and Coefficient of Determination (CoD) for analytical asssessment, Clarke Error Grid (CEG) for clinical assessments, Kolmogorov-Smirnov (KS) test for statistical analysis, and generalization ability evaluations to obtain both coarse and granular insights. The experimental findings indicate that the LSTM model demonstrates superior performance with the lowest root mean square error and highest generalization capability among all other models, closely followed by SAN. The ability of LSTM and SAN to capture long-term dependencies in blood glucose data and their correlations with various influencing factors and events contribute to their enhanced performance. Despite the lower predictive performance, the FFN was able to capture patterns and trends in the data, suggesting its applicability in forecasting future direction. Moreover, this study helps in identifying the optimal model based on specific objectives, whether prioritizing generalization or accuracy.

Text mining of verbal autopsy narratives to extract mortality causes and most prevalent diseases using natural language processing.

Verbal autopsy (VA) narratives play a crucial role in understanding and documenting the causes of mortality, especially in regions lacking robust medical infrastructure. In this study, we propose a comprehensive approach to extract mortality causes and identify prevalent diseases from VA narratives utilizing advanced text mining techniques, so as to better understand the underlying health issues leading to mortality. Our methodology integrates n-gram-based language processing, Latent Dirichlet Allocation (LDA), and BERTopic, offering a multi-faceted analysis to enhance the accuracy and depth of information extraction. This is a retrospective study that uses secondary data analysis. We used data from the Agincourt Health and Demographic Surveillance Site (HDSS), which had 16338 observations collected between 1993 and 2015. Our text mining steps entailed data acquisition, pre-processing, feature extraction, topic segmentation, and discovered knowledge. The results suggest that the HDSS population may have died from mortality causes such as vomiting, chest/stomach pain, fever, coughing, loss of weight, low energy, headache. Additionally, we discovered that the most prevalent diseases entailed human immunodeficiency virus (HIV), tuberculosis (TB), diarrhoea, cancer, neurological disorders, malaria, diabetes, high blood pressure, chronic ailments (kidney, heart, lung, liver), maternal and accident related deaths. This study is relevant in that it avails valuable insights regarding mortality causes and most prevalent diseases using novel text mining approaches. These results can be integrated in the diagnosis pipeline for ease of human annotation and interpretation. As such, this will help with effective informed intervention programmes that can improve primary health care systems and chronic based delivery, thus increasing life expectancy.

Development and validation of an electrocardiographic artificial intelligence model for detection of peripartum cardiomyopathy.

BACKGROUND: This study used electrocardiogram data in conjunction with artificial intelligence methods as a noninvasive tool for detecting peripartum cardiomyopathy. OBJECTIVE: This study aimed to assess the efficacy of an artificial intelligence-based heart failure detection model for peripartum cardiomyopathy detection. STUDY DESIGN: We first built a deep-learning model for heart failure detection using retrospective data at the University of Tennessee Health Science Center. Cases were adult and nonpregnant female patients with a heart failure diagnosis; controls were adult nonpregnant female patients without heart failure. The model was then tested on an independent cohort of pregnant women at the University of Tennessee Health Science Center with or without peripartum cardiomyopathy. We also tested the model in an external cohort of pregnant women at Atrium Health Wake Forest Baptist. Key outcomes were assessed using the area under the receiver operating characteristic curve. We also repeated our analysis using only lead I electrocardiogram as an input to assess the feasibility of remote monitoring via wearables that can capture single-lead electrocardiogram data. RESULTS: The University of Tennessee Health Science Center heart failure cohort comprised 346,339 electrocardiograms from 142,601 patients. In this cohort, 60% of participants were Black and 37% were White, with an average age (standard deviation) of 53 (19) years. The heart failure detection model achieved an area under the curve of 0.92 on the holdout set. We then tested the ability of the heart failure model to detect peripartum cardiomyopathy in an independent University of Tennessee Health Science Center cohort of pregnant women and an external Atrium Health Wake Forest Baptist cohort of pregnant women. The independent University of Tennessee Health Science Center cohort included 158 electrocardiograms from 115 patients; our deep-learning model achieved an area under the curve of 0.83 (0.77-0.89) for this data set. The external Atrium Health Wake Forest Baptist cohort involved 80 electrocardiograms from 43 patients; our deep-learning model achieved an area under the curve of 0.94 (0.91-0.98) for this data set. For identifying peripartum cardiomyopathy diagnosed ≥10 days after delivery, the model achieved an area under the curve of 0.88 (0.81-0.94) for the University of Tennessee Health Science Center cohort and of 0.96 (0.93-0.99) for the Atrium Health Wake Forest Baptist cohort. When we repeated our analysis by building a heart failure detection model using only lead-I electrocardiograms, we obtained similarly high detection accuracies, with areas under the curve of 0.73 and 0.93 for the University of Tennessee Health Science Center and Atrium Health Wake Forest Baptist cohorts, respectively. CONCLUSION: Artificial intelligence can accurately detect peripartum cardiomyopathy from electrocardiograms alone. A simple electrocardiographic artificial intelligence-based peripartum screening could result in a timelier diagnosis. Given that results with 1-lead electrocardiogram data were similar to those obtained using all 12 leads, future studies will focus on remote screening for peripartum cardiomyopathy using smartwatches that can capture single-lead electrocardiogram data.

AI-based preeclampsia detection and prediction with electrocardiogram data.

Introduction: More than 76,000 women die yearly from preeclampsia and hypertensive disorders of pregnancy. Early diagnosis and management of preeclampsia can improve outcomes for both mother and baby. In this study, we developed artificial intelligence models to detect and predict preeclampsia from electrocardiograms (ECGs) in point-of-care settings. Methods: Ten-second 12-lead ECG data was obtained from two large health care settings: University of Tennessee Health Science Center (UTHSC) and Atrium Health Wake Forest Baptist (AHWFB). UTHSC data was split into 80% training and 20% holdout data. The model used a modified ResNet convolutional neural network, taking one-dimensional raw ECG signals comprising 12 channels as an input, to predict risk of preeclampsia. Sub-analyses were performed to assess the predictive accuracy for preeclampsia prediction within 30, 60, or 90 days before diagnosis. Results: The UTHSC cohort included 904 ECGs from 759 females (78.8% African American) with a mean ± sd age of 27.3 ± 5.0 years. The AHWFB cohort included 817 ECGs from 141 females (45.4 African American) with a mean ± sd age of 27.4 ± 5.9 years. The cross-validated ECG-AI model yielded an AUC (95% CI) of 0.85 (0.77-0.93) on UTHSC holdout data, and an AUC (95% CI) of 0.81 (0.77-0.84) on AHWFB data. The sub-analysis of different time windows before preeclampsia prediction resulted in AUCs (95% CI) of 0.92 (0.84-1.00), 0.89 (0.81-0.98) and 0.90 (0.81-0.98) when tested on ECGs 30 days, 60 days and 90 days, respectively, before diagnosis. When assessed on early onset preeclampsia (preeclampsia diagnosed at <34 weeks of pregnancy), the model's AUC (95% CI) was 0.98 (0.89-1.00). Discussion: We conclude that preeclampsia can be identified with high accuracy via application of AI models to ECG data.

Feasibility of remote monitoring for fatal coronary heart disease using Apple Watch ECGs.

Background: Fatal coronary heart disease (FCHD) is often described as sudden cardiac death (affects >4 million people/year), where coronary artery disease is the only identified condition. Electrocardiographic artificial intelligence (ECG-AI) models for FCHD risk prediction using ECG data from wearable devices could enable wider screening/monitoring efforts. Objectives: To develop a single-lead ECG-based deep learning model for FCHD risk prediction and assess concordance between clinical and Apple Watch ECGs. Methods: An FCHD single-lead ("lead I" from 12-lead ECGs) ECG-AI model was developed using 167,662 ECGs (50,132 patients) from the University of Tennessee Health Sciences Center. Eighty percent of the data (5-fold cross-validation) was used for training and 20% as a holdout. Cox proportional hazards (CPH) models incorporating ECG-AI predictions with age, sex, and race were also developed. The models were tested on paired clinical single-lead and Apple Watch ECGs from 243 St. Jude Lifetime Cohort Study participants. The correlation and concordance of the predictions were assessed using Pearson correlation (R), Spearman correlation (ρ), and Cohen's kappa. Results: The ECG-AI and CPH models resulted in AUC = 0.76 and 0.79, respectively, on the 20% holdout and AUC = 0.85 and 0.87 on the Atrium Health Wake Forest Baptist external validation data. There was moderate-strong positive correlation between predictions (R = 0.74, ρ = 0.67, and κ = 0.58) when tested on the 243 paired ECGs. The clinical (lead I) and Apple Watch predictions led to the same low/high-risk FCHD classification for 99% of the participants. CPH prediction correlation resulted in an R = 0.81, ρ = 0.76, and κ = 0.78. Conclusion: Risk of FCHD can be predicted from single-lead ECGs obtained from wearable devices and are statistically concordant with lead I of a 12-lead ECG.