Optimizing ensemble U-Net architectures for robust coronary vessel segmentation in angiographic images.

Automated coronary angiography assessment requires precise vessel segmentation, a task complicated by uneven contrast filling and background noise. Our research introduces an ensemble U-Net model, SE-RegUNet, designed to accurately segment coronary vessels using 100 labeled angiographies from angiographic images. SE-RegUNet incorporates RegNet encoders and squeeze-and-excitation blocks to enhance feature extraction. A dual-phase image preprocessing strategy further improves the model's performance, employing unsharp masking and contrast-limited adaptive histogram equalization. Following fivefold cross-validation and Ranger21 optimization, the SE-RegUNet 4GF model emerged as the most effective, evidenced by performance metrics such as a Dice score of 0.72 and an accuracy of 0.97. Its potential for real-world application is highlighted by its ability to process images at 41.6 frames per second. External validation on the DCA1 dataset demonstrated the model's consistent robustness, achieving a Dice score of 0.76 and an accuracy of 0.97. The SE-RegUNet 4GF model's precision in segmenting blood vessels in coronary angiographies showcases its remarkable efficiency and accuracy. However, further development and clinical testing are necessary before it can be routinely implemented in medical practice.

Machine learning approaches for biomarker discovery to predict large-artery atherosclerosis.

Large-artery atherosclerosis (LAA) is a leading cause of cerebrovascular disease. However, LAA diagnosis is costly and needs professional identification. Many metabolites have been identified as biomarkers of specific traits. However, there are inconsistent findings regarding suitable biomarkers for the prediction of LAA. In this study, we propose a new method integrates multiple machine learning algorithms and feature selection method to handle multidimensional data. Among the six machine learning models, logistic regression (LR) model exhibited the best prediction performance. The value of area under the receiver operating characteristic curve (AUC) was 0.92 when 62 features were incorporated in the external validation set for the LR model. In this model, LAA could be well predicted by clinical risk factors including body mass index, smoking, and medications for controlling diabetes, hypertension, and hyperlipidemia as well as metabolites involved in aminoacyl-tRNA biosynthesis and lipid metabolism. In addition, we found that 27 features were present among the five adopted models that could provide good results. If these 27 features were used in the LR model, an AUC value of 0.93 could be achieved. Our study has demonstrated the effectiveness of combining machine learning algorithms with recursive feature elimination and cross-validation methods for biomarker identification. Moreover, we have shown that using shared features can yield more reliable correlations than either model, which can be valuable for future identification of LAA.

Corrigendum: Artificial intelligence-assisted remote detection of ST-elevation myocardial infarction using a mini-12-lead electrocardiogram device in prehospital ambulance care.

[This corrects the article DOI: 10.3389/fcvm.2022.1001982.].

Code-Switching Automatic Speech Recognition for Nursing Record Documentation: System Development and Evaluation.

BACKGROUND: Taiwan has insufficient nursing resources due to the high turnover rate of health care providers. Therefore, reducing the heavy workload of these employees is essential. Herein, speech transcription, which has various potential clinical applications, was employed for the documentation of nursing records. The requirement of including only one speaker per transcription facilitated data collection and system development. Moreover, authorization from patients was unnecessary. OBJECTIVE: The aim of this study was to construct a speech recognition system for nursing records such that health care providers can complete nursing records without typing or with only a few edits. METHODS: Nursing records in Taiwan are mainly written in Mandarin, with technical terms and abbreviations presented in both Mandarin and English. Therefore, the training set consisted of English code-switching information. Next, transfer learning (TL) and meta-TL (MTL) methods, which perform favorably in code-switching scenarios, were applied. RESULTS: As of September 2021, the China Medical University Hospital Artificial Intelligence Speech (CMaiSpeech) data set was established by manually annotating approximately 100 hours of recordings from 525 speakers. The word error rate (WER) of the benchmark model of syllable-based TL was 29.54% in code-switching. The WER of the proposed model of syllable-based MTL was 22.20% in code-switching. The test set comprised 17,247 words. Moreover, in a clinical case, the proposed model of syllable-based MTL yielded a WER of 31.06% in code-switching. The clinical test set contained 1159 words. CONCLUSIONS: This paper has two main contributions. First, the CMaiSpeech data set-a Mandarin-English corpus-has been established. Health care providers in Taiwan are often compelled to use a mixture of Mandarin and English in nursing records. Second, an automatic speech recognition system for nursing record document conversion was proposed. The proposed system can shorten the work handover time and further reduce the workload of health care providers.

Artificial intelligence-assisted remote detection of ST-elevation myocardial infarction using a mini-12-lead electrocardiogram device in prehospital ambulance care.

Objective: To implement an all-day online artificial intelligence (AI)-assisted detection of ST-elevation myocardial infarction (STEMI) by prehospital 12-lead electrocardiograms (ECGs) to facilitate patient triage for timely reperfusion therapy. Methods: The proposed AI model combines a convolutional neural network and long short-term memory (CNN-LSTM) to predict STEMI on prehospital 12-lead ECGs obtained from mini-12-lead ECG devices equipped in ambulance vehicles in Central Taiwan. Emergency medical technicians (EMTs) from the 14 AI-implemented fire stations performed the on-site 12-lead ECG examinations using the mini portable device. The 12-lead ECG signals were transmitted to the AI center of China Medical University Hospital to classify the recordings as "STEMI" or "Not STEMI". In 11 non-AI fire stations, the ECG data were transmitted to a secure network and read by available on-line emergency physicians. The response time was defined as the time interval between the ECG transmission and ECG interpretation feedback. Results: Between July 17, 2021, and March 26, 2022, the AI model classified 362 prehospital 12-lead ECGs obtained from 275 consecutive patients who had called the 119 dispatch centers of fire stations in Central Taiwan for symptoms of chest pain or shortness of breath. The AI's response time to the EMTs in ambulance vehicles was 37.2 ± 11.3 s, which was shorter than the online physicians' response time from 11 other fire stations with no AI implementation (113.2 ± 369.4 s, P < 0.001) after analyzing another set of 335 prehospital 12-lead ECGs. The evaluation metrics including accuracy, precision, specificity, recall, area under the receiver operating characteristic curve, and F1 score to assess the overall AI performance in the remote detection of STEMI were 0.992, 0.889, 0.994, 0.941, 0.997, and 0.914, respectively. During the study period, the AI model promptly identified 10 STEMI patients who underwent primary percutaneous coronary intervention (PPCI) with a median contact-to-door time of 18.5 (IQR: 16-20.8) minutes. Conclusion: Implementation of an all-day real-time AI-assisted remote detection of STEMI on prehospital 12-lead ECGs in the field is feasible with a high diagnostic accuracy rate. This approach may help minimize preventable delays in contact-to-treatment times for STEMI patients who require PPCI.

Comparison of multiple imputation algorithms and verification using whole-genome sequencing in the CMUH genetic biobank.

A genome-wide association study (GWAS) can be conducted to systematically analyze the contributions of genetic factors to a wide variety of complex diseases. Nevertheless, existing GWASs have provided highly ethnic specific data. Accordingly, to provide data specific to Taiwan, we established a large-scale genetic database in a single medical institution at the China Medical University Hospital. With current technological limitations, microarray analysis can detect only a limited number of single-nucleotide polymorphisms (SNPs) with a minor allele frequency of >1%. Nevertheless, imputation represents a useful alternative means of expanding data. In this study, we compared four imputation algorithms in terms of various metrics. We observed that among the compared algorithms, Beagle5.2 achieved the fastest calculation speed, smallest storage space, highest specificity, and highest number of high-quality variants. We obtained 15,277,414 high-quality variants in 175,871 people by using Beagle5.2. In our internal verification process, Beagle5.2 exhibited an accuracy rate of up to 98.75%. We also conducted external verification. Our imputed variants had a 79.91% mapping rate and 90.41% accuracy. These results will be combined with clinical data in future research. We have made the results available for researchers to use in formulating imputation algorithms, in addition to establishing a complete SNP database for GWAS and PRS researchers. We believe that these data can help improve overall medical capabilities, particularly precision medicine, in Taiwan.