Convolutional neural network misclassification analysis in oral lesions: an error evaluation criterion by image characteristics.

Journal: Oral surgery, oral medicine, oral pathology and oral radiology
Published Date: Oct 11, 2023

Abstract

OBJECTIVE: This retrospective study analyzed the errors generated by a convolutional neural network (CNN) when performing automated classification of oral lesions according to their clinical characteristics, seeking to identify patterns in systemic errors in the intermediate layers of the CNN.

Authors

  • Rita Fabiane Teixeira Gomes
    Department of Oral Pathology, Faculdade de Odontologia, Federal University of Rio Grande do Sul (UFRGS), Porto Alegre 90035-003, Brazil.
  • Jean Schmith
    Polytechnic School, University of Vale do Rio dos Sinos-UNISINOS, São Leopoldo 93022-750, Brazil.
  • Rodrigo Marques de Figueiredo
    Polytechnic School, University of Vale do Rio dos Sinos-UNISINOS, São Leopoldo, Brazil; Technology in Automation and Electronics Laboratory-TECAE Lab, University of Vale do Rio dos Sinos-UNISINOS, São Leopoldo, Brazil.
  • Samuel Armbrust Freitas
    Department of Applied Computing, University of Vale do Rio dos Sinos-UNISINOS, São Leopoldo 93022-750, Brazil.
  • Giovanna Nunes Machado
    Polytechnic School, University of Vale do Rio dos Sinos-UNISINOS, São Leopoldo 93022-750, Brazil.
  • Juliana Romanini
    Oral Medicine, Otorhynolaringology Service, Hospital de Clínicas de Porto Alegre (HCPA), Porto Alegre 90035-003, Brazil.
  • Janete Dias Almeida
    Department of Biosciences and Oral Diagnostics, São Paulo State University, Campus São José dos Campos, São Paulo, Brazil.
  • Cassius Torres Pereira
    Department of Stomatology. Federal University of Paraná, Curitiba, Brazil.
  • Jonas de Almeida Rodrigues
    Department of Surgery and Orthopaedics, Faculdade de Odontologia-Federal University of Rio Grande do Sul-UFRGS, Porto Alegre, Brazil.
  • Vinicius Coelho Carrard
    Department of Oral Pathology, Faculdade de Odontologia, Federal University of Rio Grande do Sul (UFRGS), Porto Alegre 90035-003, Brazil.

Keywords

External Resources

Popular Topics
Recent Journals