Search

Abstract Background WenXinWuYang, a novel portable Artificial Intelligence Electrocardiogram (AI-ECG) device, can detect many kinds of abnormal heart disease and perform a single-lead ECG, but its reliability and validity among pregnant women is unclear. The aim of this study was to assess the reliability and validity of heart rate, ECG measurements and diagnostic results by compared the portable device with a clinical 12-lead ECG among pregnant women. Methods We conducted a clinical study at a municipal-level maternal and child health care hospital. The pregnant women who visited the ECG room for ECG examination were invited to participate in this study. Each participant underwent three ECG recordings: one with conventional 12-lead ECG and two with WenXinWuYang ECG. The first WenXinWuYang ECG was recorded simultaneously with the 12-lead ECG. We collected heart rate, ECG measurements, and diagnostic results related to arrhythmias from both ECG devices. The data were then analyzed using Spearman rank correlation coefficients, consistency analyses, and Bland-Altman plots. Results The study included 287 ECG recordings from 99 pregnant women, with a balanced distribution across different stages of pregnancy. We observed strong to moderate correlations between the two WenXinWuYang measurements for heart rate (r = 0.847), PR interval (r = 0.728), QRS duration (r = 0.636), QT interval (r = 0.836), and QTc interval (r = 0.648), with a diagnostic consistency rate exceeding 90.0%. When compared with the 12-lead ECG, the mean differences for heart rate, PR interval, QRS duration, QT interval, and QTc interval were − 0.4 ± 3.1 bpm (r = 0.957), 14.6 ± 12.4 ms (r = 0.537), 7.0 ± 8.9 ms (r = 0.136), 2.1 ± 12.0 (r = 0.774), and 6.6 ± 16.5 (r = 0.663), respectively. Although the correlation coefficient was low in QRS duration, Bland-Altman results showed moderate to strong agreement between these intervals. Sinus rhythm recognition was fully consistent with the 12-lead ECG, with higher validity in detecting arrhythmias (sensitivity 84.2%, specificity 97.5%). Similar trends existed among different stages of pregnancy. Conclusion WenXinWuYang demonstrated acceptable reliability and validity in measuring heart rate, ECG measurements, and detecting arrhythmias among pregnant women.

Abstract Background Patients with severe coronary arterystenosis may present with apparently normal electrocardiograms (ECGs), making it difficult to detect adverse health conditions during routine screenings or physical examinations. Consequently, these patients might miss the optimal window for treatment. Methods We aimed to develop an effective model to distinguish severe coronary stenosis from no or mild coronary stenosis in patients with apparently normal ECGs. A total of 392 patients, including 138 with severe stenosis, were selected for the study. Deep learning (DL) models were trained from scratch and using pre-trained parameters via transfer learning. These models were evaluated based on ECG data alone and in combination with clinical information, including age, sex, hypertension, diabetes, dyslipidemia and smoking status. Results We found that DL models trained from scratch using ECG data alone achieved a specificity of 74.6% but exhibited low sensitivity (54.5%), comparable to the performance of logistic regression using clinical data. Adding clinical information to the ECG DL model trained from scratch improved sensitivity (90.9%) but reduced specificity (42.3%). The best performance was achieved by combining clinical information with the ECG transfer learning model, resulting in an area under the receiver operating characteristic curve (AUC) of 0.847, with 84.8% sensitivity and 70.4% specificity. Conclusions The findings demonstrate the effectiveness of DL models in identifying severe coronary stenosis in patients with apparently normal ECGs and validate an efficient approach utilizing existing ECG models. By employing transfer learning techniques, we can extract “deep features” that summarize the inherent information of ECGs with relatively low computational expense.

Background: Recent advancements in artificial intelligence (AI) have significantly improved atrial fibrillation (AF) detection using electrocardiography (ECG) data obtained during sinus rhythm (SR). However, the utility of printed ECG (pECG) records for AF detection, particularly in developing countries, remains unexplored. This study aims to assess the efficacy of an AI-based screening tool for paroxysmal AF (PAF) using pECGs during SR. Methods: We analyzed 5688 printed 12-lead SR-ECG records from 2192 patients admitted to Beijing Chaoyang Hospital between May 2011 to August 2022. All patients underwent catheter ablation for PAF (AF group) or other electrophysiological procedures (non-AF group). We developed a deep learning model to detect PAF from these printed SR-ECGs. The 2192 patients were randomly assigned to training (1972, 57.3% with PAF), validation (108, 57.4% with PAF), and test datasets (112, 57.1% with PAF). We developed an applet to digitize the printed ECG data and display the results within a few seconds. Our evaluation focused on sensitivity, specificity, accuracy, F1 score, the area under the receiver-operating characteristic curve (AUROC), and precision-recall curves (PRAUC). Results: The PAF detection algorithm demonstrated strong performance: sensitivity 87.5%, specificity 66.7%, accuracy 78.6%, F1 score 0.824, AUROC 0.871 and PRAUC 0.914. A gradient-weighted class activation map (Grad-CAM) revealed the model’s tailored focus on different ECG areas for personalized PAF detection. Conclusions: The deep-learning analysis of printed SR-ECG records shows high accuracy in PAF detection, suggesting its potential as a reliable screening tool in real-world clinical practice.

Importance: Heart sound auscultation is a routinely used physical examination in clinical practice to identify potential cardiac abnormalities. However, accurate interpretation of heart sounds requires specialized training and experience, which limits its generalizability. Deep learning, a subset of machine learning, involves training artificial neural networks to learn from large datasets and perform complex tasks with intricate patterns. Over the past decade, deep learning has been successfully applied to heart sound analysis, achieving remarkable results and accumulating substantial heart sound data for model training. Although several reviews have summarized deep learning algorithms for heart sound analysis, there is a lack of comprehensive summaries regarding the available heart sound data and the clinical applications. Highlights: This review will compile the commonly used heart sound datasets, introduce the fundamentals and state-of-the-art techniques in heart sound analysis and deep learning, and summarize the current applications of deep learning for heart sound analysis, along with their limitations and areas for future improvement. Conclusions: The integration of deep learning into heart sound analysis represents a significant advancement in clinical practice. The growing availability of heart sound datasets and the continuous development of deep learning techniques contribute to the improvement and broader clinical adoption of these models. However, ongoing research is needed to address existing challenges and refine these technologies for broader clinical use.

Abstract Background Critical values are commonly used in clinical laboratory tests to define health-related conditions of varying degrees. Knowing the values, people can quickly become aware of health risks, and the health professionals can take immediate actions and save lives. Methods In this paper, we propose a method that extends the concept of critical value to one of the most commonly used physiological signals in the clinical environment—Electrocardiogram (ECG). We first construct a mapping from common ECG diagnostic conclusions to critical values. After that, we build a 61-layer deep convolutional neural network named CardioV, which is characterized by an ordinal classifier. Results We conduct experiments on a large public ECG dataset, and demonstrate that CardioV achieves a mean absolute error of 0.4984 and a ROC-AUC score of 0.8735. In addition, we find that the model performs better for extreme critical values and the younger age group, while gender does not affect the performance. The ablation study confirms that the ordinal classification mechanism suits for estimating the critical values which contain ranking information. Moreover, model interpretation techniques help us discover that CardioV focuses on the characteristic ECG locations during the critical value estimation process. Conclusions As an ordinal classifier, CardioV performs well in estimating ECG critical values that can help people quickly identify different heart conditions. We obtain ROC-AUC scores above 0.8 for all four critical value categories, and find that the extreme values (0 (no risk) and 3 (high risk)) have better model performance than the other two (1 (low risk) and 2 (medium risk)). Results also show that gender does not affect the performance, and the older age group has worse performance than the younger age group. In addition, visualization techniques reveal that the model pays more attention to characteristic ECG locations.

Bacterial infection is one of the most serious physiological conditions threatening human health. There is an increasing demand for more effective bacterial diagnosis and treatment through non-invasive approaches. Among current antibacterial strategies of non-invasive approaches, photothermal antibacterial therapy (PTAT) has pronounced advantages with properties of minor damage to normal tissue and little chance to trigger antimicrobial resistance. Therefore, we developed a fast and simple strategy that integrated the sensitive detection and photothermal therapy of bacteria by measuring adenosine triphosphate (ATP) bioluminescence following targeted photothermal lysis. First, 3-azido-d-alanine (d-AzAla) is selectively integrated into the cell walls of bacteria, photosensitizer dibenzocyclooctyne, and double sulfonic acid-modified indocyanine green (sulfo-DBCO-ICG) are subsequently designed to react with the modified bacteria through in vivo click chemistry. Next, the sulfo-DBCO-ICG modified bacteria under irradiation of 808 nm near-infrared laser was immediately detected by ATP bioluminescence following targeted photothermal lysis and even the number of bacteria on the infected tissue can be significantly reduced through PTAT. This method has demonstrated the ability to detect the presence of the bacteria for ATP value in 32 clinical samples. As a result, the ATP value over of 100 confirmed the presence of bacteria in clinical samples for 22 patients undergoing craniotomy and ten otitis media patients. Overall, this study paves a brand new avenue to facile diagnosis and a treatment platform for clinical bacterial infections.

Stripe rust and powdery mildew are both devastating diseases for durum and common wheat. Pyramiding of genes conferring resistance to one or more diseases in a single cultivar is an important breeding approach to provide broader spectra of resistances in wheat improvement. A new powdery mildew resistance gene originating from wild emmer (Triticum turgidum var. dicoccoides) backcrossed into common wheat (T. aestivum) line WE35 was identified. It conferred an intermediate level of resistance to Blumeria graminis f. sp. tritici isolate E09 at the seedling stage and a high level of resistance at the adult plant stage. Genetic analysis showed that the powdery mildew resistance in WE35 was controlled by a dominant gene designated Pm64. Bulked segregant analysis (BSA) and molecular mapping indicated that Pm64 was located in chromosome bin 2BL4-0.50–0.89. Polymorphic markers were developed from the corresponding genomic regions of Chinese Spring wheat and wild emmer accession Zavitan to delimit Pm64 to a 0.55 cM genetic interval between markers WGGBH1364 and WGGBH612, corresponding to a 15 Mb genomic region on Chinese Spring and Zavitan 2BL, respectively. The genetic linkage map of Pm64 is critical for fine mapping and cloning. Pm64 was completely linked in repulsion with stripe rust resistance gene Yr5. Analysis of a larger segregating population might identify a recombinant line with both genes as a valuable resource in breeding for resistance to powdery mildew and stripe rust. Keywords: Blumeria graminis, Genetic linkage map, Yellow rust, Triticum aestivum, Triticum dicoccoides

Aiming at the limitation of the convolution kernel with a fixed receptive field and unknown prior to optimal network width in U-Net, multi-scale U-Net (MSU-Net) is proposed by us for medical image segmentation. First, multiple convolution sequence is used to extract more semantic features from the images. Second, the convolution kernel with different receptive fields is used to make features more diverse. The problem of unknown network width is alleviated by efficient integration of convolution kernel with different receptive fields. In addition, the multi-scale block is extended to other variants of the original U-Net to verify its universality. Five different medical image segmentation datasets are used to evaluate MSU-Net. A variety of imaging modalities are included in these datasets, such as electron microscopy, dermoscope, ultrasound, etc. Intersection over Union (IoU) of MSU-Net on each dataset are 0.771, 0.867, 0.708, 0.900, and 0.702, respectively. Experimental results show that MSU-Net achieves the best performance on different datasets. Our implementation is available at https://github.com/CN-zdy/MSU_Net.

Multimodal medical images provide significant amounts of complementary semantic information. Therefore, multimodal medical imaging has been widely used in the segmentation of gliomas through computational neural networks. However, inputting images from different sources directly to the network does not achieve the best segmentation effect. This paper describes a convolutional neural network called F-S-Net that fuses the information from multimodal medical images and uses the semantic information contained within these images for glioma segmentation. The architecture of F-S-Net is formed by cascading two sub-networks. The first sub-network projects the multimodal medical images into the same semantic space, which ensures they have the same semantic metric. The second sub-network uses a dual encoder structure (DES) and a channel spatial attention block (CSAB) to extract more detailed information and focus on the lesion area. DES and CSAB are integrated into U-Net architectures. A multimodal glioma dataset collected by Yijishan Hospital of Wannan Medical College is used to train and evaluate the network. F-S-Net is found to achieve a dice coefficient of 0.9052 and Jaccard similarity of 0.8280, outperforming several previous segmentation methods.

Leaf senescence is normally the last stage of plant development. Early senescence of functional leaves significantly reduces the photosynthetic time and efficiency, seriously affecting grain yield and quality in wheat. Discovering genes responsible for early leaf senescence (els) are necessary for developing novel germplasms and cultivars with delayed leaf-senescence through molecular manipulation and marker assisted selection. In this study, we identified an early leaf senescence line M114 in a derivative of a wheat breeding population. Genetic analysis indicated that early leaf senescence in M114 is controlled by a single recessive gene, provisionally designated els1. By applying bulked segregant analysis and RNA-Seq (BSR-Seq), seven polymorphic markers linked to els1 were developed and the gene was located on chromosome arm 2BS in a 1.5 cM genetic interval between markers WGGB303 and WGGB305. A co-segregating marker, WGGB302, provide a starting point for fine mapping and map-based cloning of els1. Keywords: Bulk segregant analysis, Genetic mapping, Leaf senescence, Triticum aestivum

Wireless sensor network is drawing intensive research attention and being applied in various practical scenarios. The integration of WSN in industrial manufacturing process provides low cost industrial automation systems. The benefits of WSN include rapid deployment, self-organization, and being maintenance free. However, wireless sensor nodes are battery powered; replacing or recharging battery is usually impossible after deployment, which makes energy a scarce resource for WSN. Furthermore, in industrial applications, environment information is gathered and delivered in a timely manner, which makes traditional energy saving methods such as listen-sleeping duty approach less efficient. In this work, a novel dynamic address encoding scheme based on packet arrival rate is proposed; the scheme can reduce the protocol overhead by reducing MAC and network address size and thus saves energy needed for data transmission. The implementation of the scheme is fully distributed and needs little modification of the existing applications. Simulation results show the proposed scheme is efficient and robust on both the MAC layer and the network layer.