Mu-Yang Hsieh’s research while affiliated with National Taiwan University and other places

Accurate segmentation of the left ventricle (LV) using echocardiogram (Echo) images is essential for cardiovascular analysis. Conventional techniques are labor-intensive and exhibit inter-observer variability. Deep learning has emerged as a powerful tool for automated medical image segmentation, offering advantages in speed and potentially superior accuracy. This study explores the efficacy of employing a YOLO (You Only Look Once) segmentation model for automated LV segmentation in Echo images. YOLO, a cutting-edge object detection model, achieves exceptional speed–accuracy balance through its well-designed architecture. It utilizes efficient dilated convolutional layers and bottleneck blocks for feature extraction while incorporating innovations like path aggregation and spatial attention mechanisms. These attributes make YOLO a compelling candidate for adaptation to LV segmentation in Echo images. We posit that by fine-tuning a pre-trained YOLO-based model on a well-annotated Echo image dataset, we can leverage the model’s strengths in real-time processing and precise object localization to achieve robust LV segmentation. The proposed approach entails fine-tuning a pre-trained YOLO model on a rigorously labeled Echo image dataset. Model performance has been evaluated using established metrics such as mean Average Precision (mAP) at an Intersection over Union (IoU) threshold of 50% (mAP50) with 98.31% and across a range of IoU thresholds from 50% to 95% (mAP50:95) with 75.27%. Successful implementation of YOLO for LV segmentation has the potential to significantly expedite and standardize Echo image analysis. This advancement could translate to improved clinical decision-making and enhanced patient care.

Accurate segmentation of the Left Ventricle (LV) in Echocardiogram (Echo) images is essential for cardiovascular analysis. Conventional techniques are labor-intensive and exhibit inter-observer variability. Deep learning has emerged as a powerful tool for automated medical image segmentation, offering advantages in speed and potentially superior accuracy. This study explores the efficacy of employing the YOLO (You Only Look Once) segmentation model for automated LV segmentation in Echo images. YOLO, a cutting-edge object detection model, achieves exceptional speed-accuracy balance through its well-designed architecture. It utilizes efficient dilated convolutional layers and bottleneck blocks for feature extraction, while incorporating innovations like path aggregation and spatial attention mechanisms. These attributes make YOLO a compelling candidate for adaptation to LV segmentation in Echo images. We posit that by fine-tuning a pre-trained YOLO based model on a well-annotated Echo image dataset, we can leverage the model's strengths in real-time processing and precise object localization to achieve robust LV segmentation. The proposed approach entails fine-tuning a pre-trained YOLO model on a rigorously labeled Echo image dataset. Model performance has been evaluated using established metrics such as mean Average Precision (mAP) at an Intersection over Union (IoU) threshold of 50% (mAP50) with 98.31%and across a range of IoU thresholds from 50% to 95% (mAP50:95) with 75.27%. Successful implementation of YOLO for LV segmentation has the potential to significantly expedite and standardize Echo image analysis. This advancement could translate to improved clinical decision-making and enhanced patient care.

We report a 14-year-old adolescent who presented with acute limb ischemia caused by systemic lupus erythematosus-related antiphospholipid syndrome (APS). In the pediatric population, acute limb ischemia is rare. This case is unique in that after the initial medical treatment failed, interventional devices for acute stroke intervention were utilized to salvage the limb in our patient with a small tibial artery vessel to achieve procedural success. To provide limb salvage, operators may combine peripheral and neuro-intervention devices to maximize procedure success.

Objective: Previous studies on arteriovenous fistulas have demonstrated the potential benefit of drug-coated balloons (DCBs) in maintaining the patency of dialysis access. However, stenoses involving stent-grafts were excluded from these studies. Therefore, we aimed to evaluate the effectiveness of DCBs in treating stent-graft stenosis. Methods: This was a prospective, single-blinded, randomised controlled study. From March 2017 to April 2021, 40 patients with dysfunctional vascular access owing to stent-graft stenosis were randomised to treatment with a DCB or a conventional balloon. Clinical follow up was scheduled at 1, 3, and 6 months, and angiographic follow up was performed 6 months after the intervention. The primary outcome was angiographic late luminal loss at 6 months, and secondary outcomes included target lesion and access circuit primary patency at 6 months. Results: Thirty-six participants completed follow up angiography. The DCB group had a superior mean late luminal loss at 6 months compared with the control group (1.82 mm ± 1.83 mm vs. 3.63 mm ± 1.08 mm, respectively, p = .001). All 40 patients completed clinical follow up. The DCB group had a superior 6-month target lesion primary patency compared with the control group [hazard ratio (HR) 0.23, 95% confidence interval (CI) 0.07 - 0.71; p = .005). Additionally, the DCB group had a numerically higher 6-month access circuit primary patency rate than the control group, although the difference was not statistically significant (HR 0.54, 95% CI 0.26 - 1.11, p = .095). Conclusion: Conventional balloon angioplasty is not durable in stent-graft stenosis treatment. Treatment with DCBs provides less angiographic late luminal loss and potentially superior primary patency of the target lesion than treatment with conventional balloons. [ClinicalTrials ID: NCT03360279.].

Wet beriberi is a rare but fatal disease in modern society. The nonspecific clinical manifestations, including symptoms of heart failure and recalcitrant lactic acidosis, can prevent timely diagnosis. The use of a pulmonary artery catheter can promptly confirm a high cardiac output state and plays a crucial role in rapidly deteriorating cases. Appropriate treatment with intravenous administration of thiamine leads to dramatic recovery within hours. We present two cases of Shoshin beriberi, a fulminant variant of wet beriberi, diagnosed in 2016 and 2022 at our institute. The patients experienced haemodynamic collapse and refractory lactic acidosis, which were successfully diagnosed with the use of a pulmonary artery catheter and reversed by thiamine supplementation. We also reviewed 19 cases of wet beriberi reported between 2010 and 2022.