Search Results

Now showing 1 - 3 of 3

Some of the metrics are blocked by your
consent settings
Neural Architecture Search Using Trajectory Metaheuristics to Classify Coronary Stenosis
(IEEE, 2024)
Franco-Gaona, Erick
;
Avila-Garcia, Maria-Susana
;
Cruz-Aceves, Ivan
;
Orocio-Garcia, Hiram-Efrain
;
Escobedo-Gordillo, Andrés
Coronary stenosis is a disease that claims millions of lives each year. Early detection of this condition is crucial for patient survival. Currently, physicians perform detection by x-ray angiograms, however, the variability of diagnoses and the difficulty of access to expertise has led to the need for automated, computer-assisted diagnosis. In this work explores the use of deep learning to classify stenosis or non-stenosis in angiogram images using convolutional neural networks from scratch. A methodology to fine-tuning network architectures automatically using metaheuristic optimization techniques is proposed, demonstrating superior performance to fine-tuning empirically and proposing a new architecture in the literature to classify coronary stenosis. The proposed architectures achieved 86.02% and 95.67% F1-score with simulated annealing and iterated local search techniques, respectively. ©The authors ©IEEE
8
Some of the metrics are blocked by your
consent settings
Optimal Dataset Size for Fine-Tuning sEMG-Based Hand Gesture Recognition in Rehabilitation Prosthesis
(IEEE, 2024)
Escobedo-Gordillo, Andrés
;
Brieva, Jorge
;
Moya-Albor, Ernesto
;
Ponce, Hiram
;
Franco-Gaona, Erick
Surface electromyography (sEMG) has become a vital tool for controlling prostheses and rehabilitation using hand gesture recognition. However, the process of fine-tuning machine learning models to individual users often requires considerable amounts of data, which can be challenging to obtain due to user fatigue and discomfort. This work investigates the optimal dataset size needed for fine-tuning a pretrained Convolutional Neural Network (CNN) model for hand gesture recognition, using the NinaPro DB2 dataset. Our results show that training on just a third of the dataset achieves over 90% accuracy, highlighting a significant reduction in the data requirements compared to traditional methods. This approach can minimize the burden of data collection on users, making sEMG-based rehabilitation devices more practical and accessible. ©The authors ©IEEE
8
Some of the metrics are blocked by your
consent settings
Contactless Video-Based Vital-Sign Measurement Methods: A Data-Driven Review
(Springer, 2024-01-01)
Brieva, Jorge
;
Escobedo-Gordillo, Andrés
;
Moya-Albor, Ernesto
;
Ponce, Hiram
Nowadays, the healthcare is a priority for both governments and persons. Vital sign monitoring allows knowing the health status and is widely used for prevention, diagnosis, and treatment of determined illnesses. In particular, breathing and heart rate are traditionally considered the most relevant and accessible vital signs. However, oxygen saturation was essential in the COVID-19 pandemic. On the other hand, contact techniques to estimate these vital signs are a standard monitoring reference. However, non-contact estimation methods have gained relevance in the last few years in those cases where there is the possibility of suffering stress, pain, and skin irritation in specific situations, as in the case of vulnerable skin in burn patients and neonates. In this chapter, a review of contactless video-based vital-sign methods is presented. The selected methods have a data-driven approach as an alternative when there is not theoretical model of the physiological phenomenon. Finally, a new framework with a general data-driven approach to estimate the most used vital signs is proposed. This framework includes a region of interest extraction stage, a video magnification technique to reveals subtle changes, and a machine learning method to estimate the vital signs. In addition, each step describes some recommendations and best practices found ©Springer.
29

CRIS

Browse

Filters

Settings

Sort By

Results per page

Search Results