Utilizing Multi-layer Perceptron for Esophageal Cancer Classification Through Machine Learning Methods DOI Open Access
Sandeep Kumar, Jagendra Singh, Vinayakumar Ravi

et al.

The Open Public Health Journal, Journal Year: 2024, Volume and Issue: 17(1)

Published: Oct. 7, 2024

Aims This research paper aims to check the effectiveness of a variety machine learning models in classifying esophageal cancer through MRI scans. The current study encompasses Convolutional Neural Network (CNN), K-Nearest Neighbor (KNN), Recurrent (RNN), and Visual Geometry Group 16 (VGG16), among others which are elaborated this paper. identify most accurate model facilitate increased, improved diagnostic accuracy revolutionize early detection methods for dreadful disease. ultimate goal is, therefore, improve clinical practice performance its results with advanced techniques medical diagnosis. Background Esophageal poses critical problem oncologists since pathology is quite complex, death rate exceptionally high. Proper essential effective treatment survival. positive, but conventional not sensitive have low specificity. Recent progress brings new possibility high sensitivity specificity explores potentiality different machine-learning scans complement constraints traditional diagnostics approach. Objective aimed at verifying whether CNN, KNN, RNN, VGG16, amongst other models, correctly from review establishing all these best all. It plays role developing mechanisms that increase patient outcome confidence setting. Methods applies approach comparative analysis by using four unique classify was made possible intensive training validation standardized set data. model’s assessed evaluation metrics, included accuracy, precision, recall, F1 score. Results In cancers scans, found VGG16 be an adequate model, 96.66%. CNN took second position, 94.5%, showing efficient spatial pattern recognition. KNN RNN also showed commendable performance, accuracies 91.44% 88.97%, respectively, portraying their strengths proximity-based handling sequential These findings underline potential add significant value processes diagnosis models. Conclusion concluded techniques, mainly had escalated precision imaging. great while displayed detection, followed RNN. Thus, opportunities introducing computational clinics, might transform strategies patient-centered outcomes oncology.

Language: Английский

Deep learning-based automated method for enhancing excavator activity recognition in far-field construction site surveillance videos DOI
Yejin Shin, Seungwon Seo, Choongwan Koo

et al.

Automation in Construction, Journal Year: 2025, Volume and Issue: 173, P. 106099 - 106099

Published: March 1, 2025

Language: Английский

Citations

0
Analysis of the Climate Impact on Occupational Health and Safety Using Heat Stress Indexes DOI Open Access
Guilherme Neto Ferrari,

Guilherme Custódio dos Santos,

Paulo César Ossani

et al.

International Journal of Environmental Research and Public Health, Journal Year: 2025, Volume and Issue: 22(1), P. 130 - 130

Published: Jan. 20, 2025

Workers may be exposed to conditions that put their physical and mental integrity at risk, from workplace settings climate characteristics. Heat stress is a harmful health condition caused by exceeding the human body’s tolerance limits, leading illness increasing chance of work accidents. indexes, such as Humidex Index (HI), are used measure these impacts. These indexes significant they provide quantitative heat experienced workers, taking into account both environmental individual factors. Objective: This study aims compare multiple relating them historical Brazilian occupational accident data. Methods: We selected eight applied correspondence analysis each one, statistical method generates graphs visualize association between variables in database. Results: The comparison indicated seven presented similar behavior. It was also possible relate ranges index values with specific characteristics Conclusions: technique allowed us analyze relationship accidents showed choice does not significantly alter results for most studied.

Language: Английский

Citations

0
A Segmented Classification and Regression Machine Learning Approach for Correcting Precipitation Forecast at 4–6 h Leadtimes DOI
Y.M. Xie, Linye Song, Mingxuan Chen

et al.

Journal of Meteorological Research, Journal Year: 2025, Volume and Issue: 39(1), P. 79 - 99

Published: Feb. 1, 2025

Language: Английский

Citations

0
Utilizing Multi-layer Perceptron for Esophageal Cancer Classification Through Machine Learning Methods DOI Open Access
Sandeep Kumar, Jagendra Singh, Vinayakumar Ravi

et al.

The Open Public Health Journal, Journal Year: 2024, Volume and Issue: 17(1)

Published: Oct. 7, 2024

Aims This research paper aims to check the effectiveness of a variety machine learning models in classifying esophageal cancer through MRI scans. The current study encompasses Convolutional Neural Network (CNN), K-Nearest Neighbor (KNN), Recurrent (RNN), and Visual Geometry Group 16 (VGG16), among others which are elaborated this paper. identify most accurate model facilitate increased, improved diagnostic accuracy revolutionize early detection methods for dreadful disease. ultimate goal is, therefore, improve clinical practice performance its results with advanced techniques medical diagnosis. Background Esophageal poses critical problem oncologists since pathology is quite complex, death rate exceptionally high. Proper essential effective treatment survival. positive, but conventional not sensitive have low specificity. Recent progress brings new possibility high sensitivity specificity explores potentiality different machine-learning scans complement constraints traditional diagnostics approach. Objective aimed at verifying whether CNN, KNN, RNN, VGG16, amongst other models, correctly from review establishing all these best all. It plays role developing mechanisms that increase patient outcome confidence setting. Methods applies approach comparative analysis by using four unique classify was made possible intensive training validation standardized set data. model’s assessed evaluation metrics, included accuracy, precision, recall, F1 score. Results In cancers scans, found VGG16 be an adequate model, 96.66%. CNN took second position, 94.5%, showing efficient spatial pattern recognition. KNN RNN also showed commendable performance, accuracies 91.44% 88.97%, respectively, portraying their strengths proximity-based handling sequential These findings underline potential add significant value processes diagnosis models. Conclusion concluded techniques, mainly had escalated precision imaging. great while displayed detection, followed RNN. Thus, opportunities introducing computational clinics, might transform strategies patient-centered outcomes oncology.

Language: Английский

Citations

1