A New Artificial Intelligence Approach for the Radiographic Classification of Sacroiliitis

• Esther Fernandez ^[3] ; Javier Garrigos ^[1] ; Jose Javier Martinez ^[1] ; Irene Cases ^[3] ; Manuel Jose Moreno ^[3] ; Luis Francisco Linares ^[3] ; Angel García ^[4] ; Jose Manuel Ferrandez ^[1] ; Eduardo Fernández ^[2]
  1. [1] Universidad Politécnica de Cartagena

     Universidad Politécnica de Cartagena

     Cartagena, España

     
  2. [2] Universidad Miguel Hernández de Elche

     Universidad Miguel Hernández de Elche

     Elche, España

     
  3. [3] Virgen de la Arrixaca, University Clinical Hospital, Murcia. Spain
  4. [4] University Hospital Torrevieja, Alicante, Spain

  Mostrar afiliaciones +
• Localización: Bio-inspired Systems and Applications: from Robotics to Ambient Intelligence: 9th International Work-Conference on the Interplay Between Natural and Artificial Computation, IWINAC 2022, Puerto de la Cruz, Tenerife, Spain, May 31 – June 3, 2022, Proceedings, Part II / José Manuel Ferrández Vicente (dir. congr.), José Ramón Álvarez Sánchez (dir. congr.), Félix de la Paz López (dir. congr.), Hojjat Adeli (aut.), 2022, ISBN 978-3-031-06527-9, págs. 383-390
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • Radiographs of the sacroiliac joints are commonly used as the first imaging methods for the diagnosis of Sacroileitis in patients with back pain. This study aims to develop and validate a new artificial intelligence approach for the automatic classification of the grade of sacroiliitis on conventional radiographs. We included a total of 267 patients with chronic back pain and clinical suggestion of Axial Spondyloarthritis who presented in a specialized center. Radiographs of sacroiliac joints were evaluated by 3 rheumatologists and 1 radiologist according to the modified New York criteria. For training the network, labeled sacroiliac joints images were resized and then several artificial neural networks were tested. Better results were achieved using the CNN-XGBoost architecture, which provided good generalizability and a high specificity with acceptable sensitivity in the detection of the grade of sacroiliitis, except for class 1. Although more studies are still needed, these artificial intelligence algorithms could potentially assist medical clinicians for the detection of radiographic sacroiliitis.