João M. M. Santana
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
João M. M. Santana has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
- Caio Bezerra Souto Maior
Center for Risk Analysis, Reliability Engineering and Environmental Modeling (CEERMA), Universidade Federal de Pernambuco, Recife, Brazil
2 shared publications David e A Martins
2 shared publications- Eduardo Jose Novaes Menezes2 shared publications
Isis Didier Lins
2 shared publications- Isis Didier Lins
Universidade Federal de Pernambuco, Brazil; Universidade Federal de Pernambuco, Brazil
2 shared publications João Mateus Marques De Santana
2 shared publications
Latest Publications
- 9 June 2022
- Published by Elsevier in Acta Tropica
- Vol. 233, 106551
Seroprevalence of SARS-CoV-2 on health professionals via Bayesian estimation: a Brazilian case study before and after vaccines
- 9 June 2022
- Published by Elsevier in Acta Tropica
- Vol. 233, 106551
Seroprevalence of SARS-CoV-2 on health professionals via Bayesian estimation: a Brazilian case study before and after vaccines
- Open Access
- 1 March 2021
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 16 (3) , e0247839
Convolutional neural network model based on radiological images to support COVID-19 diagnosis: Evaluating database biases
As SARS-CoV-2 has spread quickly throughout the world, the scientific community has spent major efforts on better understanding the characteristics of the virus and possible means to prevent, diagnose, and treat COVID-19. A valid approach presented in the literature is to develop an image-based method to support COVID-19 diagnosis using convolutional neural networks (CNN). Because the availability of radiological data is rather limited due to the novelty of COVID-19, several methodologies consider reduced datasets, which may be inadequate, biasing the model. Here, we performed an analysis combining six different databases using chest X-ray images from open datasets to distinguish images of infected patients while differentiating COVID-19 and pneumonia from ‘no-findings’ images. In addition, the performance of models created from fewer databases, which may imperceptibly overestimate their results, is discussed. Two CNN-based architectures were created to process images of different sizes (512 × 512, 768 × 768, 1024 × 1024, and 1536 × 1536). Our best model achieved a balanced accuracy (BA) of 87.7% in predicting one of the three classes (‘no-findings’, ‘COVID-19’, and ‘pneumonia’) and a specific balanced precision of 97.0% for ‘COVID-19’ class. We also provided binary classification with a precision of 91.0% for detection of sick patients (i.e., with COVID-19 or pneumonia) and 98.4% for COVID-19 detection (i.e., differentiating from ‘no-findings’ or ‘pneumonia’). Indeed, despite we achieved an unrealistic 97.2% BA performance for one specific case, the proposed methodology of using multiple databases achieved better and less inflated results than from models with specific image datasets for training. Thus, this framework is promising for a low-cost, fast, and noninvasive means to support the diagnosis of COVID-19.
- 1 March 2021
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 16 (3) , e0247839
Open AccessConvolutional neural network model based on radiological images to support COVID-19 diagnosis: Evaluating database biases
As SARS-CoV-2 has spread quickly throughout the world, the scientific community has spent major efforts on better understanding the characteristics of the virus and possible means to prevent, diagnose, and treat COVID-19. A valid approach presented in the literature is to develop an image-based method to support COVID-19 diagnosis using convolutional neural networks (CNN). Because the availability of radiological data is rather limited due to the novelty of COVID-19, several methodologies consider reduced datasets, which may be inadequate, biasing the model. Here, we performed an analysis combining six different databases using chest X-ray images from open datasets to distinguish images of infected patients while differentiating COVID-19 and pneumonia from ‘no-findings’ images. In addition, the performance of models created from fewer databases, which may imperceptibly overestimate their results, is discussed. Two CNN-based architectures were created to process images of different sizes (512 × 512, 768 × 768, 1024 × 1024, and 1536 × 1536). Our best model achieved a balanced accuracy (BA) of 87.7% in predicting one of the three classes (‘no-findings’, ‘COVID-19’, and ‘pneumonia’) and a specific balanced precision of 97.0% for ‘COVID-19’ class. We also provided binary classification with a precision of 91.0% for detection of sick patients (i.e., with COVID-19 or pneumonia) and 98.4% for COVID-19 detection (i.e., differentiating from ‘no-findings’ or ‘pneumonia’). Indeed, despite we achieved an unrealistic 97.2% BA performance for one specific case, the proposed methodology of using multiple databases achieved better and less inflated results than from models with specific image datasets for training. Thus, this framework is promising for a low-cost, fast, and noninvasive means to support the diagnosis of COVID-19.
- 1 December 2020
- Published by Instituto Brasileiro de Petroleo e Gas in Rio Oil and Gas Expo and Conference
- Vol. 20 (2020) , 134-135
Reliability estimation for interval control valve installed in subsea wells of brazilian pre-salt
- 1 December 2020
- Published by Instituto Brasileiro de Petroleo e Gas in Rio Oil and Gas Expo and Conference
- Vol. 20 (2020) , 134-135
Reliability estimation for interval control valve installed in subsea wells of brazilian pre-salt
- 1 December 2020
- Published by Instituto Brasileiro de Petroleo e Gas in Rio Oil and Gas Expo and Conference
- Vol. 20 (2020) , 135-136
Technical assurance in new technology development projects: a reliability-based approach
- 1 December 2020
- Published by Instituto Brasileiro de Petroleo e Gas in Rio Oil and Gas Expo and Conference
- Vol. 20 (2020) , 135-136
Technical assurance in new technology development projects: a reliability-based approach
- Open Access
- 15 June 2018
- Published by Taylor & Francis
- p. 2863-2868
Personal protective equipment detection in industrial facilities using camera video streaming
Organizations and industries must ensure safe operation of their facilities, employing rigorous risk management techniques for planning and executing their activities. The use of Personal Protective Equipment (PPE) represents the closest layer of protection to workers, and can considerably reduce the risk of exposure to hazards, being critical for safety in industrial environments. The hazards addressed by protective equipment include physical, acoustic, electrical, heat and chemicals. Despite the substantial efforts in increasing awareness about the benefits of PPE to strive towards zero accident philosophy, operators often neglect its use when not being supervised. However, organizations commonly have surveillance cameras installed which might provide useful visual information on correct usage of PPE. In this context, computer vision is an interdisciplinary field that seeks to automate tasks that the human visual system can do and includes domains of signal and image processing, pattern recognition and artificial intelligence. Moreover, object recognition is a prominent technology from computer vision for finding and identifying objects in an image or video sequence. Then, this work aims to create an automatic PPE detection from surveillance cameras and other video streams using computer vision and machine learning. Equipment such as helmets, safety glasses, earplugs and other garments are checked for whether they are being used by operators in a real-time monitoring, alerting supervisors to prevent accidents and ensure a safer environment.
- 15 June 2018
- Published by Taylor & Francis
- p. 2863-2868
Open AccessPersonal protective equipment detection in industrial facilities using camera video streaming
Organizations and industries must ensure safe operation of their facilities, employing rigorous risk management techniques for planning and executing their activities. The use of Personal Protective Equipment (PPE) represents the closest layer of protection to workers, and can considerably reduce the risk of exposure to hazards, being critical for safety in industrial environments. The hazards addressed by protective equipment include physical, acoustic, electrical, heat and chemicals. Despite the substantial efforts in increasing awareness about the benefits of PPE to strive towards zero accident philosophy, operators often neglect its use when not being supervised. However, organizations commonly have surveillance cameras installed which might provide useful visual information on correct usage of PPE. In this context, computer vision is an interdisciplinary field that seeks to automate tasks that the human visual system can do and includes domains of signal and image processing, pattern recognition and artificial intelligence. Moreover, object recognition is a prominent technology from computer vision for finding and identifying objects in an image or video sequence. Then, this work aims to create an automatic PPE detection from surveillance cameras and other video streams using computer vision and machine learning. Equipment such as helmets, safety glasses, earplugs and other garments are checked for whether they are being used by operators in a real-time monitoring, alerting supervisors to prevent accidents and ensure a safer environment.