Transactions of Indian National Academy of Engineering, Journal Year: 2024, Volume and Issue: 9(4), P. 903 - 915
Published: Aug. 17, 2024
Language: Английский
Advances in medical technologies and clinical practice book series, Journal Year: 2025, Volume and Issue: unknown, P. 309 - 324
Published: Feb. 14, 2025
Smart medical devices are widely used to treat various disease, tracking patient's health status, and providing earlier treatments. The future healthcare depends upon the interrelations of humans intelligent through communication networks for monitoring status patient. gather data from makes smart decisions user measured data. In this chapter, significant research works carried out in wearable Internet Things classifies with categories as health, sports, athletics, daily activity, information, safety. adoption technologies such Artificial Intelligence, plays a crucial role system that brings transformation hospital patient examining physiological remotely recommendation. This chapter explores function devices, technologies, challenges healthcare.
Language: Английский
Citations
0
Applied Materials Today, Journal Year: 2024, Volume and Issue: 41, P. 102508 - 102508
Published: Nov. 14, 2024
Language: Английский
Citations
3
Nano-Structures & Nano-Objects, Journal Year: 2025, Volume and Issue: 42, P. 101473 - 101473
Published: April 3, 2025
Language: Английский
Citations
0
Healthcare Analytics, Journal Year: 2023, Volume and Issue: 4, P. 100265 - 100265
Published: Sept. 24, 2023
The first seven vertebrae of our spine are called the cervical spine. It supports weight head, encloses and safeguards spinal cord, permits a variety head motions. joined at rear bone by kind joint known as facet joint. These joints enable us to move necks forward, backward, twist. Fractures medical emergency that may lead lifelong paralysis or even death. If left untreated undetected, these fractures can worsen over time. Using computed tomography, fracture in individuals be accurately diagnosed. Given scarcity research on practical use deep learning methods detecting persons, it is imperative address this gap. This study uses dataset containing normal tomography images. proposed modified transfer-learning-based MobileNetV2, InceptionV3, Resnet50V2 models. An ablation was also conducted determine optimal custom layers for models data augmentation techniques. In addition, evaluation metrics have been used analyze compare model's performance. Among all approaches, MobileNetV2 with has achieved highest accuracy 99.75%. Furthermore, best-performing model deployed smartphone-based Android application.
Language: Английский
Citations
8
Preventive Medicine Reports, Journal Year: 2024, Volume and Issue: 41, P. 102710 - 102710
Published: March 29, 2024
To enhance the daily training quality of athletes without inducing significant physiological fatigue, aiming to achieve a balance between efficiency and load. Firstly, we developed an activity classification model using random forest algorithm introduced "effective rate" (the ratio effective time total time) as metric for assessing athlete efficiency. Secondly, method rating load was established, involving qualitative quantitative analyses fatigue through subjective scores heart rate data. Lastly, optimization system balance, utilizing multiple inertial sensors, created. Athlete states were categorized into nine types based on ratings, with corresponding management recommendations provided. Overall, this study, combining sports recognition assessment model, has excellent performance. The results indicate that prediction accuracy is high 94.70%. Additionally, average relative RPE score evaluation metrics, demonstrates good overall fit, validating feasibility model. This wearable devices monitor balanced It provides reference athletes' physical health levels, offering coaches relevant professionals.
Language: Английский
Citations
2
Actuators, Journal Year: 2024, Volume and Issue: 13(1), P. 38 - 38
Published: Jan. 17, 2024
The hydraulic pump plays a pivotal role in engineering machinery, and it is essential to continuously monitor its operating status. However, many vital signals for monitoring cannot be directly obtained practical applications. To address this, we propose soft sensor approach predicting the flow signal of based on graph convolutional network (GCN) long short-term memory (LSTM). Our innovative GCN-LSTM model intricately designed capture both spatial temporal interdependencies inherent complex such as pumps. We used GCN extract features LSTM process variables. evaluate performance pump, construct real-world experimental dataset with an actual shovel. further evaluated two public datasets, showing effectiveness pumps other operations.
Language: Английский
Citations
1
Deleted Journal, Journal Year: 2024, Volume and Issue: 20(6s), P. 2082 - 2091
Published: April 29, 2024
Biometrics recognition technology utilizes unique biological characteristics such as fingerprints, iris patterns, facial features, or voiceprints to identify and authenticate individuals. Through advanced algorithms pattern techniques, biometric systems capture analyze these physiological behavioral traits verify a person's identity. This offers high levels of security accuracy, making it valuable for various applications including access control, time attendance tracking, border security, digital payments. The construction an athletes' physical condition monitoring analysis system utilizing represents significant advancement in sports science. By integrating cutting-edge sensors algorithms, this provides real-time insights into parameters performance metrics. Athlete play pivotal roles optimizing training strategies, preventing injuries, enhancing overall athletic performance. In paper, we propose novel approach leveraging the Hybrid Multi-Instance Ensemble Classifier (HMIEC) combined with accurately classify athlete data assess their condition. Our study explores effectiveness HMIEC across multiple modalities, ECG-based, fingerprint-based, recognition, identifying athletes parameters. series experiments analyses, demonstrate HMIEC's superior classification compared other classifiers, sensitivity, specificity, area under curve (AUC). reductions heart rate from 75 65 beats per minute, increase oxygen saturation 98% 99%, decreases blood pressure readings 120/80 mmHg 110/70 mmHg, enhancements flexibility 30 35 centimeters, muscle strength 100 lbs 120 lbs, endurance capacity 20 25 minutes.
Language: Английский
Citations
1
Mobile Networks and Applications, Journal Year: 2024, Volume and Issue: unknown
Published: Sept. 3, 2024
Language: Английский
Citations
1
Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)
Published: May 30, 2024
Knowledge distillation is an effective approach for training robust multi-modal machine learning models when synchronous multimodal data are unavailable. However, traditional knowledge techniques have limitations in comprehensively transferring across modalities and models. This paper proposes a multiscale framework to address these limitations. Specifically, we introduce semantic graph mapping (SGM) loss function enable more comprehensive transfer between teacher student networks at multiple feature scales. We also design fusion tuning (FT) module fully utilize correlations within different types of the same modality networks. Furthermore, adopt transformer-based backbones improve compared convolutional neural apply proposed human activity recognition with baseline method, it improved by 2.31% 0.29% on MMAct UTD-MHAD datasets. Ablation studies validate necessity each component.
Language: Английский
Citations
1
Procedia Computer Science, Journal Year: 2024, Volume and Issue: 246, P. 4909 - 4915
Published: Jan. 1, 2024
Language: Английский
Citations
1