Aprimorando a classificação automática de tumores cerebrais com técnicas de aprendizado profundo

Autores

  • Willian de Vargas UFCSPA
  • Dieine Estela Bernieri Schiavon UFCSPA
  • Viviane Rodrigues Botelho UFCSPA
  • Thatiane Alves Pianoski UFCSPA
  • Carla Diniz Lopes Becker UFCSPA

DOI:

https://doi.org/10.59681/2175-4411.v16.iEspecial.2024.1253

Palavras-chave:

Aprendizado Profundo, Classificação Automática, Tumores cerebrais

Resumo

A precisão na classificação automática de tumores cerebrais desempenha um papel determinante para a confiabilidade do método para aplicações na saúde. Erros de


classificação podem resultar em diagnósticos imprecisos, levando a abordagens inadequadas e potencialmente prejudiciais. Objetivo: Propor uma abordagem visando minimizar erros de classificação. Método: Desenvolvemos um modelo de rede neural convolucional em duas etapas: primeiro, quatro modelos binários para tumores que apresentam maiores desafios de diferenciação; depois, um modelo Ensemble para classificação multiclasse. Adicionalmente, empregamos uma técnica para interpretar as previsões dos modelos e identificar as regiões de interesse nas imagens médicas. Resultados: Os resultados demonstram que a abordagem proposta alcança uma acurácia de 98%. Conclusão: Este trabalho trouxe contribuições para a aplicação de aprendizado profundo na classificação de tumores cerebrais, destacando a importância de abordagens transparentes e robustas para garantir precisão e segurança nas previsões.

Biografia do Autor

Willian de Vargas, UFCSPA

Master’s Student, Federal University of Health Sciences of Porto Alegre – UFCSPA, Porto Alegre (RS), Brazil.

Dieine Estela Bernieri Schiavon, UFCSPA

Master’s Student, Federal University of Health Sciences of Porto Alegre – UFCSPA, Porto Alegre (RS), Brazil.

Viviane Rodrigues Botelho, UFCSPA

Ph.D., Federal University of Health Sciences of Porto Alegre – UFCSPA, DECESA, Porto Alegre (RS), Brazil.

Thatiane Alves Pianoski, UFCSPA

Ph.D., Federal University of Health Sciences of Porto Alegre – UFCSPA, DECESA, Porto Alegre (RS), Brazil.

Carla Diniz Lopes Becker, UFCSPA

Ph.D., Federal University of Health Sciences of Porto Alegre – UFCSPA, DECESA, Porto Alegre (RS), Brazil.

Referências

Câncer do sistema nervoso central [Internet]. Instituto Nacional de Câncer - INCA. [cited 2024 Jan 26]. Available from: https://www.gov.br/inca/pt-br/assuntos/cancer/tipos/sistema- nervoso-central

Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA: a Cancer Journal for Clinicians. 2021 Feb 4;71(3):209–49.

Nowosielski M, Galldiks N, Iglseder S, Kickingereder P, von Deimling A, Bendszus M, et al. Diagnostic challenges in meningioma. Neuro-Oncology [Internet]. 2017 Nov 1 [cited 2024 Jan 26];19(12):1588–98. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5716093/

Wilms G, Lammens M, Marchal G, Demaerel P, Verplancke J, Van Calenbergh F, et al. Prominent dural enhancement adjacent to nonmeningiomatous malignant lesions on contrast-enhanced MR images. AJNR American journal of neuroradiology [Internet]. 1991 [cited 2024 Jan 26];12(4):761–4. Available from: https://pubmed.ncbi.nlm.nih.gov/1882761/

Bhuvaji S. SartajBhuvaji/Brain-Tumor-Classification-DataSet [Internet]. GitHub. 2023 [cited 2024 Jan 26]. Available from: https://github.com/SartajBhuvaji/Brain-Tumor-Classification- DataSet

Cheng, J. (2017). Brain tumor dataset [Internet]. [cited 2024 Jan 26]. Available from: https://figshare.com/articles/dataset/brain_tumor_dataset/1512427. DOI: 10.6084/m9.figshare.1512427.v5

Ahmedhamada0. Br35H: Brain Tumor Detection 2020 [Internet]. [cited 2024 Jan 26]. www.kaggle.com. Available from: https://www.kaggle.com/datasets/ahmedhamada0/brain- tumor-detection

Ribeiro MT, Singh S, Guestrin C. “Why Should I Trust You?” Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD ’16 [Internet]. 2016; [cited 2024 Jan 26]. Available from: https://arxiv.org/pdf/1602.04938.pdf.

Rasool M, Ismail NA, Boulila W, Ammar A, Samma H, Yafooz WMS, et al. A Hybrid Deep Learning Model for Brain Tumour Classification. Entropy [Internet]. 2022 Jun 8;24(6):799. [cited 2024 Jan 26]. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9222774/

Rasool M, Ismail NA, Al-Dhaqm A, Yafooz WMS, Alsaeedi A. A Novel Approach for Classifying Brain Tumours Combining a SqueezeNet Model with SVM and Fine-Tuning. Electronics. 2022 Dec 29;12(1):149.

Mahajan, N., & Chavan, H. (2023). A Robust Approach for Brain Tumor Detection using Transfer Learning. Available from: https://ieeexplore.ieee.org/abstract/document/10220906?casa_token=ZmdSd_mFuJ0AAAA A:l5j9izaUrS-LA26hcIujQgSDBscVkYwPc9MsmLEmXBrS3XRWZ2- dOfnBend_8uD_RDg4scQ7gg

Gómez-Guzmán MA, Jiménez-Beristaín L, García-Guerrero EE, López-Bonilla OR, Tamayo-Perez UJ, Esqueda-Elizondo JJ, et al. Classifying Brain Tumors on Magnetic Resonance Imaging by Using Convolutional Neural Networks. Electronics [Internet]. 2023 Jan 1;12(4):955. [cited 2024 Jan 26]. Available from: https://www.mdpi.com/2079- 9292/12/4/955

Özkaraca O, Bağrıaçık Oİ, Gürüler H, Khan F, Hussain J, Khan J, et al. Multiple Brain Tumor Classification with Dense CNN Architecture Using Brain MRI Images. Life. 2023 Jan 28;13(2):349.

Nickparvar, M. (2023). Brain Tumor MRI Dataset [Internet]. [cited 2024 Jan 26]. www.kaggle.com. Available from: https://www.kaggle.com/datasets/masoudnickparvar/brain- tumor-mri-dataset/

Filatov D, Yar GNAH. Brain Tumor Diagnosis and Classification via Pre-Trained Convolutional Neural Networks [Internet]. arXiv.org. 2022. [cited 2024 Jan 26]. Available from: https://arxiv.org/abs/2208.00768

Glioma Object Detection Dataset and Pre-Trained Model by test [Internet]. Roboflow. [cited 2024 Jan 26]. Available from: https://universe.roboflow.com/test-786lz/glioma-4mibx

Meningioma Detector Object Detection Dataset and Pre-Trained Model by Kelompok BRIN [Internet]. Roboflow. [cited 2024 Jan 26]. Available from: https://universe.roboflow.com/kelompok-brin-dxxif/meningioma-detector

brain-tumor-classification Classification Dataset by opendataacademia [Internet]. Roboflow. [cited 2024 Jan 26]. Available from: https://universe.roboflow.com/opendataacademia/brain- tumor-classification-qkeoa

Desai PK. Multiple meningiomas | Radiology Case | Radiopaedia.org [Internet].

Radiopaedia. [cited 2024 Jan 26]. Available from: https://radiopaedia.org/cases/multiple- meningiomas-1

Gomes L. Glioma de Baixo Grau - Tumores Cerebrais [Internet]. Dr. Gustavo Noleto. 2022 [cited 2024 Jan 26]. Available from: https://www.neurodrgustavonoleto.com/glioma-de-baixo- grau/

Downloads

Publicado

19-11-2024

Como Citar

de Vargas, W., Schiavon, D. E. B., Botelho, V. R., Pianoski, T. A., & Becker, C. D. L. (2024). Aprimorando a classificação automática de tumores cerebrais com técnicas de aprendizado profundo. Journal of Health Informatics, 16(Especial). https://doi.org/10.59681/2175-4411.v16.iEspecial.2024.1253

Artigos mais lidos pelo mesmo(s) autor(es)