A Hybrid Transformer-Convolutional Neural Network for Segmentation of Intracerebral Hemorrhage and Perihematomal Edema on Non-Contrast Head Computed Tomography (CT) with Uncertainty Quantification to Improve Confidence DOI Creative Commons
Tran Anh Tuan, Dmitriy Desser, Tal Zeevi

и другие.

Bioengineering, Год журнала: 2024, Номер 11(12), С. 1274 - 1274

Опубликована: Дек. 15, 2024

Intracerebral hemorrhage (ICH) and perihematomal edema (PHE) are key imaging markers of primary secondary brain injury in hemorrhagic stroke. Accurate segmentation quantification ICH PHE can help with prognostication guide treatment planning. In this study, we combined Swin-Unet Transformers nnU-NETv2 convolutional network for on non-contrast head CTs. We also applied test-time data augmentations to assess individual-level prediction uncertainty, ensuring high confidence prediction. The model was trained 1782 CT scans from a multicentric trial tested two independent datasets Yale (n = 396) University Berlin Charité Hospital Medical Center Hamburg-Eppendorf 943). Model performance evaluated the Dice coefficient Volume Similarity (VS). Our dual Swin-nnUNET achieved median (95% interval) 0.93 (0.90–0.95) VS 0.97 (0.95–0.98) ICH, 0.70 (0.64–0.75) 0.87 (0.80–0.93) cohort. 0.86 (0.80–0.90) 0.91 (0.85–0.95) 0.65 (0.56–0.70) (0.77–0.93) Berlin/Hamburg-Eppendorf Prediction uncertainty associated lower accuracy, smaller ICH/PHE volumes, infratentorial location. results highlight benefits transformer-convolutional neural architecture augmentation quantification.

Язык: Английский

Strain sensing behavior of conductive polymer composites under corrosion fatigue DOI

Zhi Wu,

Enrico Zappino, Jianying Hu

и другие.

International Journal of Mechanical Sciences, Год журнала: 2025, Номер 288, С. 110034 - 110034

Опубликована: Янв. 31, 2025

Язык: Английский

Процитировано

0
Recent progress of biosensors based on thermoelectric effects for monitoring physical activity and environment monitoring DOI Open Access

Xiao-Jie Tang,

Cai Qi,

Qiang Sun

и другие.

Soft Science, Год журнала: 2025, Номер 5(1)

Опубликована: Фев. 13, 2025

Thermoelectric (TE) materials and sensors have emerged as a frontier in health environmental monitoring, offering silent, simple, reliable alternative to traditional power generation methods by harnessing waste heat into usable electrical energy. They also offer superior stability longevity, making them ideal for long-term monitoring applications. Furthermore, when compared other self-powered biosensors, TE excel their ability operate wide range of temperatures conditions, providing more consistent source sensor operation. This review delves the recent advancements TE-based sensors, highlighting multifunctional capabilities real-time sensing. We explore fundamental principles conversion, including Seebeck effect, assess performance metric, specifically figure-of-merit (ZT ). The integration with flexible wearable electronics is discussed, emphasizing high efficiency mechanical robustness. Applications devices internet things (IoT)-integrated systems are underscored, particularly fire detection personal monitoring. Challenges material limitations, miniaturization, scalability addressed, focus on future research directions enhance sustainability longevity sensors. provides comprehensive overview development technology its trajectory, importance ongoing address current challenges realize these innovative devices.

Язык: Английский

Процитировано

0
Elastomer with Microchannel Nanofiber Array Inspired by Rabbit Cornea Achieves Rapid Liquid Spreading and Reduction of Frictional Vibration Noise DOI Creative Commons
Bowen Zhang, Lei Jiang, Ruochen Fang

и другие.

Biomimetics, Год журнала: 2025, Номер 10(3), С. 164 - 164

Опубликована: Март 7, 2025

Reducing friction-induced squeal noise is a common issue in daily life and industrial production, particularly for elastomers. However, adjusting structure wettability wet environments to solve the remains challenge. Here, inspired by rabbit corneas, microchannel nanofiber array composite superhydrophilic elastomer material was prepared achieve rapid liquid spreading optimize distribution. Researchers have found that when depth of groove 400 μm length 5000 nm, water rapidly spreads on surface only 430 ms. This reduces self-excited vibration caused friction, thereby reducing squealing 20 decibels (dB). article proposes novel direct biomimetic reduction strategy, which has great potential solving friction problems industry life, such as automotive wiper blades, engines, oil lubricated bearings, etc.

Язык: Английский

Процитировано

0
A Hybrid Transformer-Convolutional Neural Network for Segmentation of Intracerebral Hemorrhage and Perihematomal Edema on Non-Contrast Head Computed Tomography (CT) with Uncertainty Quantification to Improve Confidence DOI Creative Commons
Tran Anh Tuan, Dmitriy Desser, Tal Zeevi

и другие.

Bioengineering, Год журнала: 2024, Номер 11(12), С. 1274 - 1274

Опубликована: Дек. 15, 2024

Intracerebral hemorrhage (ICH) and perihematomal edema (PHE) are key imaging markers of primary secondary brain injury in hemorrhagic stroke. Accurate segmentation quantification ICH PHE can help with prognostication guide treatment planning. In this study, we combined Swin-Unet Transformers nnU-NETv2 convolutional network for on non-contrast head CTs. We also applied test-time data augmentations to assess individual-level prediction uncertainty, ensuring high confidence prediction. The model was trained 1782 CT scans from a multicentric trial tested two independent datasets Yale (n = 396) University Berlin Charité Hospital Medical Center Hamburg-Eppendorf 943). Model performance evaluated the Dice coefficient Volume Similarity (VS). Our dual Swin-nnUNET achieved median (95% interval) 0.93 (0.90–0.95) VS 0.97 (0.95–0.98) ICH, 0.70 (0.64–0.75) 0.87 (0.80–0.93) cohort. 0.86 (0.80–0.90) 0.91 (0.85–0.95) 0.65 (0.56–0.70) (0.77–0.93) Berlin/Hamburg-Eppendorf Prediction uncertainty associated lower accuracy, smaller ICH/PHE volumes, infratentorial location. results highlight benefits transformer-convolutional neural architecture augmentation quantification.

Язык: Английский

Процитировано

0