Souza Jr., Luis Antonio

@informatica.ufes.br

Professor, Department of Informatics
Universidade Federal do Espírito Santo

Souza Jr., Luis Antonio


               Download Compact PDF  
https://researchid.co/la.souza

RESEARCH, TEACHING, or OTHER INTERESTS

Computer Science Applications, Artificial Intelligence, Computer Engineering, Management Information Systems
13

Scopus Publications

1421

Scholar Citations

17

Scholar h-index

20

Scholar i10-index

Scopus Publications

  • Error-Correcting Mean-Teacher: Corrections instead of consistency-targets applied to semi-supervised medical image segmentation
    Robert Mendel, David Rauber, Luis A. de Souza, João P. Papa, Christoph Palm
    Computers in Biology and Medicine, 2023
    Semantic segmentation is an essential task in medical imaging research. Many powerful deep-learning-based approaches can be employed for this problem, but they are dependent on the availability of an expansive labeled dataset. In this work, we augment such supervised segmentation models to be suitable for learning from unlabeled data. Our semi-supervised approach, termed Error-Correcting Mean-Teacher, uses an exponential moving average model like the original Mean Teacher but introduces our new paradigm of error correction. The original segmentation network is augmented to handle this secondary correction task. Both tasks build upon the core feature extraction layers of the model. For the correction task, features detected in the input image are fused with features detected in the predicted segmentation and further processed with task-specific decoder layers. The combination of image and segmentation features allows the model to correct present mistakes in the given input pair. The correction task is trained jointly on the labeled data. On unlabeled data, the exponential moving average of the original network corrects the student's prediction. The combined outputs of the students' prediction with the teachers' correction form the basis for the semi-supervised update. We evaluate our method with the 2017 and 2018 Robotic Scene Segmentation data, the ISIC 2017 and the BraTS 2020 Challenges, a proprietary Endoscopic Submucosal Dissection dataset, Cityscapes, and Pascal VOC 2012. Additionally, we analyze the impact of the individual components and examine the behavior when the amount of labeled data varies, with experiments performed on two distinct segmentation architectures. Our method shows improvements in terms of the mean Intersection over Union over the supervised baseline and competing methods. Code is available at https://github.com/CloneRob/ECMT.
  • Vessel and tissue recognition during third-space endoscopy using a deep learning algorithm
    Alanna Ebigbo, Robert Mendel, Markus W Scheppach, Andreas Probst, Neal Shahidi, Friederike Prinz, Carola Fleischmann, Christoph Römmele, Stefan Karl Goelder, Georg Braun, David Rauber, Tobias Rueckert, Luis A de Souza, Joao Papa, Michael Byrne, Christoph Palm, Helmut Messmann
    Gut, 2022
    In this study, we aimed to develop an artificial intelligence clinical decision support solution to mitigate operator-dependent limitations during complex endoscopic procedures such as endoscopic submucosal dissection and peroral endoscopic myotomy, for example, bleeding and perforation. A DeepLabv3-based model was trained to delineate vessels, tissue structures and instruments on endoscopic still images from such procedures. The mean cross-validated Intersection over Union and Dice Score were 63% and 76%, respectively. Applied to standardised video clips from third-space endoscopic procedures, the algorithm showed a mean vessel detection rate of 85% with a false-positive rate of 0.75/min. These performance statistics suggest a potential clinical benefit for procedure safety, time and also training.
  • Detecting atherosclerotic plaque calcifications of the carotid artery through optimum-path forest
    Danilo Samuel Jodas, Mateus Roder, Rafael Pires, Marcos Cleison Silva Santana, Luis A. de Souza, Leandro Aparecido Passos
    Optimum Path Forest Theory Algorithms and Applications, 2022
  • Endoscopic prediction of submucosal invasion in Barrett's cancer with the use of artificial intelligence: A pilot study
    Alanna Ebigbo, Robert Mendel, Tobias Rückert, Laurin Schuster, Andreas Probst, Johannes Manzeneder, Friederike Prinz, Matthias Mende, Ingo Steinbrück, Siegbert Faiss, David Rauber, Luis A. de Souza, João P. Papa, Pierre H. Deprez, Tsuneo Oyama, Akiko Takahashi, Stefan Seewald, Prateek Sharma, Michael F. Byrne, Christoph Palm, Helmut Messmann
    Endoscopy, 2021
    Background The accurate differentiation between T1a and T1b Barrett’s-related cancer has both therapeutic and prognostic implications but is challenging even for experienced physicians. We trained an artificial intelligence (AI) system on the basis of deep artificial neural networks (deep learning) to differentiate between T1a and T1b Barrett’s cancer on white-light images. Methods Endoscopic images from three tertiary care centers in Germany were collected retrospectively. A deep learning system was trained and tested using the principles of cross validation. A total of 230 white-light endoscopic images (108 T1a and 122 T1b) were evaluated using the AI system. For comparison, the images were also classified by experts specialized in endoscopic diagnosis and treatment of Barrett’s cancer. Results The sensitivity, specificity, F1 score, and accuracy of the AI system in the differentiation between T1a and T1b cancer lesions was 0.77, 0.64, 0.74, and 0.71, respectively. There was no statistically significant difference between the performance of the AI system and that of experts, who showed sensitivity, specificity, F1, and accuracy of 0.63, 0.78, 0.67, and 0.70, respectively. Conclusion This pilot study demonstrates the first multicenter application of an AI-based system in the prediction of submucosal invasion in endoscopic images of Barrett’s cancer. AI scored equally to international experts in the field, but more work is necessary to improve the system and apply it to video sequences and real-life settings. Nevertheless, the correct prediction of submucosal invasion in Barrett’s cancer remains challenging for both experts and AI.
  • Convolutional Neural Networks for the evaluation of cancer in Barrett's esophagus: Explainable AI to lighten up the black-box
    Luis A. de Souza, Robert Mendel, Sophia Strasser, Alanna Ebigbo, Andreas Probst, Helmut Messmann, João P. Papa, Christoph Palm
    Computers in Biology and Medicine, 2021
    Even though artificial intelligence and machine learning have demonstrated remarkable performances in medical image computing, their level of accountability and transparency must be provided in such evaluations. The reliability related to machine learning predictions must be explained and interpreted, especially if diagnosis support is addressed. For this task, the black-box nature of deep learning techniques must be lightened up to transfer its promising results into clinical practice. Hence, we aim to investigate the use of explainable artificial intelligence techniques to quantitatively highlight discriminative regions during the classification of early-cancerous tissues in Barrett's esophagus-diagnosed patients. Four Convolutional Neural Network models (AlexNet, SqueezeNet, ResNet50, and VGG16) were analyzed using five different interpretation techniques (saliency, guided backpropagation, integrated gradients, input × gradients, and DeepLIFT) to compare their agreement with experts' previous annotations of cancerous tissue. We could show that saliency attributes match best with the manual experts' delineations. Moreover, there is moderate to high correlation between the sensitivity of a model and the human-and-computer agreement. The results also lightened that the higher the model's sensitivity, the stronger the correlation of human and computational segmentation agreement. We observed a relevant relation between computational learning and experts' insights, demonstrating how human knowledge may influence the correct computational learning.
  • Abstract: Semi-supervised Segmentation Based on Error-correcting Supervision
    Robert Mendel, Luis Antonio de Souza Jr, David Rauber, João Paulo Papa, Christoph Palm
    Informatik Aktuell, 2021
  • Assisting Barrett's esophagus identification using endoscopic data augmentation based on Generative Adversarial Networks
    Luis A. de Souza, Leandro A. Passos, Robert Mendel, Alanna Ebigbo, Andreas Probst, Helmut Messmann, Christoph Palm, João P. Papa
    Computers in Biology and Medicine, 2020
  • Real-time use of artificial intelligence in the evaluation of cancer in Barrett's oesophagus
    Alanna Ebigbo, Robert Mendel, Andreas Probst, Johannes Manzeneder, Friederike Prinz, Luis A de Souza Jr., Joao Papa, Christoph Palm, Helmut Messmann
    Gut, 2020
    Based on previous work by our group with manual annotation of visible Barrett oesophagus (BE) cancer images, a real-time deep learning artificial intelligence (AI) system was developed. While an expert endoscopist conducts the endoscopic assessment of BE, our AI system captures random images from the real-time camera livestream and provides a global prediction (classification), as well as a dense prediction (segmentation) differentiating accurately between normal BE and early oesophageal adenocarcinoma (EAC). The AI system showed an accuracy of 89.9% on 14 cases with neoplastic BE. This paper follows up on our prior publication on the application of AI and deep learning in the evaluation of BE.1 2 In our initial publications, we developed a computer-aided diagnosis (CAD) model and demonstrated promising performance scores in the classification and segmentation domains during BE assessment.1 2 However, these results were achieved on optimal endoscopic images, which may not mirror the real-life situation sufficiently. To enable the seamless integration of AI-based image classification into the clinical workflow, our previous system was developed further to increase the speed of image analysis for classification and the resolution of the dense prediction, which shows the color-coded spatial distribution of cancer probabilities.1 2 Still based on deep convolutional neural nets (CNNs) and a residual net (ResNet) architecture with DeepLab V.3+, a state-of-the-art encoder–decoder network was adapted.3 To transfer the endoscopic livestream to our AI system, a capture card (Avermedia, Taiwan) was plugged to the endoscopic monitor. Online supplementary video 1 shows the setting of AI-based BE evaluation in the endoscopy room of the University Hospital Augsburg (figure 1). The AI prediction can be started at any time using either a button on the keyboard or a foot switch. The video clip shows examples of …
  • Learning visual representations with optimum-path forest and its applications to Barrett’s esophagus and adenocarcinoma diagnosis
    Luis A. de Souza, Luis C. S. Afonso, Alanna Ebigbo, Andreas Probst, Helmut Messmann, Robert Mendel, Christian Hook, Christoph Palm, João P. Papa
    Neural Computing and Applications, 2020
  • Semi-supervised Segmentation Based on Error-Correcting Supervision
    Robert Mendel, Luis Antonio de Souza, David Rauber, João Paulo Papa, Christoph Palm
    Lecture Notes in Computer Science Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics, 2020
  • Computer-aided diagnosis using deep learning in the evaluation of early oesophageal adenocarcinoma
    Alanna Ebigbo, Robert Mendel, Andreas Probst, Johannes Manzeneder, Luis Antonio de Souza Jr, João P Papa, Christoph Palm, Helmut Messmann
    Gut, 2019
  • Barrett's esophagus analysis using infinity Restricted Boltzmann Machines
    Leandro A. Passos, Luis A. de Souza Jr., Robert Mendel, Alanna Ebigbo, Andreas Probst, Helmut Messmann, Christoph Palm, João Paulo Papa
    Journal of Visual Communication and Image Representation, 2019
  • A survey on Barrett's esophagus analysis using machine learning
    Luis A. de Souza, Christoph Palm, Robert Mendel, Christian Hook, Alanna Ebigbo, Andreas Probst, Helmut Messmann, Silke Weber, João P. Papa
    Computers in Biology and Medicine, 2018

RECENT SCHOLAR PUBLICATIONS

  • PRISM: a clinically interpretable stepwise framework for multimodal skin cancer diagnosis
    PHG Bouzon, WF da Rocha, LA de Souza Jr, AGC Pacheco
    Scientific Reports , 2026
    2026
  • LiwTERM-r: A Revised Lightweight Transformer-based Model for Multimodal Skin Lesion Detection Robust to Incomplete Input
    LA de Souza Júnior, AGC Pacheco, TO dos Santos, WF da Rocha, ...
    Journal of the Brazilian Computer Society 32 (1), 305-315 , 2026
    2026
  • Dermalyze: uma aplicação para auxílio à triagem de lesões de pele baseado em aprendizado profundo
    EP Magesk, LA de Souza Júnior, PHL Frasson, AGC Pacheco
    Brazilian Symposium on Multimedia and the Web (WebMedia), 185-193 , 2025
    2025
  • Layer-selective deep representation to improve esophageal cancer classification
    LA Souza Jr, LA Passos, MCS Santana, R Mendel, D Rauber, A Ebigbo, ...
    Medical & Biological Engineering & Computing 62 (11), 3355-3372 , 2024
    2024
    Citations: 2
  • Retraction Note: Learning visual representations with optimum-path forest and its applications to Barrett’s esophagus and adenocarcinoma diagnosis
    LA de Souza Jr, LCS Afonso, A Ebigbo, A Probst, H Messmann, R Mendel, ...
    Neural Computing and Applications 36 (28), 17753-17753 , 2024
    2024
  • LiwTERM: A lightweight transformer-based model for dermatological multimodal lesion detection
    LA Souza, AGC Pacheco, GG De Angelo, T Oliveira-Santos, C Palm, ...
    2024 37th SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), 1-6 , 2024
    2024
    Citations: 10
  • A review of deep learning‐based approaches for deepfake content detection
    LA Passos, D Jodas, KAP Costa, LA Souza Júnior, D Rodrigues, ...
    Expert Systems 41 (8), e13570 , 2024
    2024
    Citations: 150
  • DeepCraftFuse: visual and deeply-learnable features work better together for esophageal cancer detection in patients with Barrett’s esophagus
    LA Souza Jr, AGC Pacheco, LA Passos, MCS Santana, R Mendel, ...
    Neural Computing and Applications 36 (18), 10445-10459 , 2024
    2024
    Citations: 3
  • Error-correcting mean-teacher: Corrections instead of consistency-targets applied to semi-supervised medical image segmentation
    R Mendel, D Rauber, LA de Souza Jr, JP Papa, C Palm
    Computers in Biology and Medicine 154, 106585 , 2023
    2023
    Citations: 27
  • Vessel and tissue recognition during third-space endoscopy using a deep learning algorithm
    A Ebigbo, R Mendel, MW Scheppach, A Probst, N Shahidi, F Prinz, ...
    Gut 71 (12), 2388-2390 , 2022
    2022
    Citations: 38
  • A fuzzy distance-based ensemble of deep models for cervical cancer detection
    R Pramanik, M Biswas, S Sen, LA de Souza Júnior, JP Papa, R Sarkar
    Computer Methods and Programs in Biomedicine 219, 106776 , 2022
    2022
    Citations: 138
  • Auxílio ao diagnóstico automático do esôfago de Barrett utilizando aprendizado de máquina
    LA Souza Júnior
    Universidade Federal de São Carlos , 2022
    2022
  • Auxílio ao diagnóstico automático do esôfago de Barrett utilizando aprendizado de máquina
    LA de Souza Júnior
    Federal University of São Carlos, Brazil , 2022
    2022
  • Detecting atherosclerotic plaque calcifications of the carotid artery through optimum-path forest
    DS Jodas, M Roder, R Pires, MCS Santana, LA de Souza Jr, LA Passos
    Optimum-Path Forest, 137-154 , 2022
    2022
    Citations: 3
  • Endoscopic prediction of submucosal invasion in Barrett’s cancer with the use of artificial intelligence: a pilot study
    A Ebigbo, R Mendel, T Rückert, L Schuster, A Probst, J Manzeneder, ...
    Endoscopy 53 (09), 878-883 , 2021
    2021
    Citations: 84
  • Convolutional Neural Networks for the evaluation of cancer in Barrett's esophagus: Explainable AI to lighten up the black-box
    LA de Souza Jr, R Mendel, S Strasser, A Ebigbo, A Probst, H Messmann, ...
    Computers in Biology and Medicine 135, 104578 , 2021
    2021
    Citations: 72
  • Fine-tuning Generative Adversarial Networks using Metaheuristics: A Case Study on Barrett's Esophagus Identification
    LA Souza, LA Passos, R Mendel, A Ebigbo, A Probst, H Messmann, ...
    Bildverarbeitung für die Medizin 2021: Proceedings, German Workshop on … , 2021
    2021
    Citations: 18
  • Assisting Barrett's esophagus identification using endoscopic data augmentation based on Generative Adversarial Networks
    LA de Souza Jr, LA Passos, R Mendel, A Ebigbo, A Probst, H Messmann, ...
    Computers in Biology and Medicine 126, 104029 , 2020
    2020
    Citations: 57
  • Semi-supervised segmentation based on error-correcting supervision
    R Mendel, LA de Souza Jr, D Rauber, JP Papa, C Palm
    European conference on computer vision, 141-157 , 2020
    2020
    Citations: 147
  • Real-time use of artificial intelligence in the evaluation of cancer in Barrett’s oesophagus
    A Ebigbo, R Mendel, A Probst, J Manzeneder, F Prinz, LA de Souza Jr, ...
    Gut 69 (4), 615-616 , 2020
    2020
    Citations: 214

MOST CITED SCHOLAR PUBLICATIONS

  • Real-time use of artificial intelligence in the evaluation of cancer in Barrett’s oesophagus
    A Ebigbo, R Mendel, A Probst, J Manzeneder, F Prinz, LA de Souza Jr, ...
    Gut 69 (4), 615-616 , 2020
    2020
    Citations: 214
  • Computer-aided diagnosis using deep learning in the evaluation of early oesophageal adenocarcinoma
    A Ebigbo, R Mendel, A Probst, J Manzeneder, LA Souza Jr, JP Papa, ...
    Gut 68 (7), 1143-1145 , 2019
    2019
    Citations: 195
  • A review of deep learning‐based approaches for deepfake content detection
    LA Passos, D Jodas, KAP Costa, LA Souza Júnior, D Rodrigues, ...
    Expert Systems 41 (8), e13570 , 2024
    2024
    Citations: 150
  • Semi-supervised segmentation based on error-correcting supervision
    R Mendel, LA de Souza Jr, D Rauber, JP Papa, C Palm
    European conference on computer vision, 141-157 , 2020
    2020
    Citations: 147
  • A fuzzy distance-based ensemble of deep models for cervical cancer detection
    R Pramanik, M Biswas, S Sen, LA de Souza Júnior, JP Papa, R Sarkar
    Computer Methods and Programs in Biomedicine 219, 106776 , 2022
    2022
    Citations: 138
  • Endoscopic prediction of submucosal invasion in Barrett’s cancer with the use of artificial intelligence: a pilot study
    A Ebigbo, R Mendel, T Rückert, L Schuster, A Probst, J Manzeneder, ...
    Endoscopy 53 (09), 878-883 , 2021
    2021
    Citations: 84
  • Convolutional Neural Networks for the evaluation of cancer in Barrett's esophagus: Explainable AI to lighten up the black-box
    LA de Souza Jr, R Mendel, S Strasser, A Ebigbo, A Probst, H Messmann, ...
    Computers in Biology and Medicine 135, 104578 , 2021
    2021
    Citations: 72
  • A survey on Barrett's esophagus analysis using machine learning
    LA de Souza Jr, C Palm, R Mendel, C Hook, A Ebigbo, A Probst, ...
    Computers in biology and medicine 96, 203-213 , 2018
    2018
    Citations: 62
  • A technical review of artificial intelligence as applied to gastrointestinal endoscopy: clarifying the terminology
    A Ebigbo, C Palm, A Probst, R Mendel, J Manzeneder, F Prinz, ...
    Endoscopy international open 7 (12), E1616-E1623 , 2019
    2019
    Citations: 58
  • Assisting Barrett's esophagus identification using endoscopic data augmentation based on Generative Adversarial Networks
    LA de Souza Jr, LA Passos, R Mendel, A Ebigbo, A Probst, H Messmann, ...
    Computers in Biology and Medicine 126, 104029 , 2020
    2020
    Citations: 57
  • Automatic frontal sinus recognition in computed tomography images for person identification
    LA de Souza Jr, AN Marana, SAT Weber
    Forensic science international 286, 252-264 , 2018
    2018
    Citations: 43
  • Vessel and tissue recognition during third-space endoscopy using a deep learning algorithm
    A Ebigbo, R Mendel, MW Scheppach, A Probst, N Shahidi, F Prinz, ...
    Gut 71 (12), 2388-2390 , 2022
    2022
    Citations: 38
  • Error-correcting mean-teacher: Corrections instead of consistency-targets applied to semi-supervised medical image segmentation
    R Mendel, D Rauber, LA de Souza Jr, JP Papa, C Palm
    Computers in Biology and Medicine 154, 106585 , 2023
    2023
    Citations: 27
  • Barrett’s esophagus analysis using infinity restricted Boltzmann machines
    LA Passos, LA de Souza Jr, R Mendel, A Ebigbo, A Probst, H Messmann, ...
    Journal of Visual Communication and Image Representation 59, 475-485 , 2019
    2019
    Citations: 26
  • Barrett's esophagus identification using optimum-path forest
    LA De Souza, LCS Afonso, C Palm, JP Papa
    2017 30th SIBGRAPI conference on graphics, patterns and images (SIBGRAPI … , 2017
    2017
    Citations: 25
  • Barrett’s esophagus analysis using SURF features
    L Souza, C Hook, JP Papa, C Palm
    Bildverarbeitung für die Medizin 2017: Algorithmen-Systeme-Anwendungen … , 2017
    2017
    Citations: 22
  • Fine-tuning Generative Adversarial Networks using Metaheuristics: A Case Study on Barrett's Esophagus Identification
    LA Souza, LA Passos, R Mendel, A Ebigbo, A Probst, H Messmann, ...
    Bildverarbeitung für die Medizin 2021: Proceedings, German Workshop on … , 2021
    2021
    Citations: 18
  • RETRACTED ARTICLE: Learning visual representations with optimum-path forest and its applications to Barrett’s esophagus and adenocarcinoma diagnosis
    LA de Souza Jr, LCS Afonso, A Ebigbo, A Probst, H Messmann, R Mendel, ...
    Neural Computing and Applications 32 (3), 759-775 , 2020
    2020
    Citations: 13
  • Barrett's esophagus identification using color Co-occurrence matrices
    L Souza, A Ebigbo, A Probst, H Messmann, JP Papa, R Mendel, C Palm
    2018 31st SIBGRAPI conference on graphics, patterns and images (SIBGRAPI … , 2018
    2018
    Citations: 13
  • LiwTERM: A lightweight transformer-based model for dermatological multimodal lesion detection
    LA Souza, AGC Pacheco, GG De Angelo, T Oliveira-Santos, C Palm, ...
    2024 37th SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), 1-6 , 2024
    2024
    Citations: 10