Detection and Classification of Knee Osteoarthritis
Joseph Humberto Cueva,
Darwin Castillo,
Héctor Espinós-Morató,
David Durán,
Patricia Díaz,
Vasudevan Lakshminarayanan
Affiliations
Joseph Humberto Cueva
Departamento de Química, Facultad de Ciencias Exactas y Naturales, Universidad Técnica Particular de Loja, San Cayetano Alto s/n, Loja 11-01-608, Ecuador
Darwin Castillo
Departamento de Química, Facultad de Ciencias Exactas y Naturales, Universidad Técnica Particular de Loja, San Cayetano Alto s/n, Loja 11-01-608, Ecuador
Héctor Espinós-Morató
Escuela de Ciencia, Ingeniería y Diseño, Universidad Europea de Valencia, Paseo de la Alameda 7, 46010 Valencia, Spain
David Durán
Applied Data Science Lab (ADaS Lab), Facultat Informàtica, Multimedia i Telecomunicacions, Universitat Oberta de Catalunya, Avenida Tibidabo 39-43, 08035 Barcelona, Spain
Patricia Díaz
Facultad de Ciencias Médicas, Universidad Técnica Particular de Loja, San Cayetano Alto s/n, Loja 11-01-608, Ecuador
Vasudevan Lakshminarayanan
Theoretical and Experimental Epistemology Lab, School of Optometry and Vision Science, University of Waterloo, Waterloo, ON N2L3G1, Canada
Osteoarthritis (OA) affects nearly 240 million people worldwide. Knee OA is the most common type of arthritis, especially in older adults. Physicians measure the severity of knee OA according to the Kellgren and Lawrence (KL) scale through visual inspection of X-ray or MR images. We propose a semi-automatic CADx model based on Deep Siamese convolutional neural networks and a fine-tuned ResNet-34 to simultaneously detect OA lesions in the two knees according to the KL scale. The training was done using a public dataset, whereas the validations were performed with a private dataset. Some problems of the imbalanced dataset were solved using transfer learning. The model results average of the multi-class accuracy is 61%, presenting better performance results for classifying classes KL-0, KL-3, and KL-4 than KL-1 and KL-2. The classification results were compared and validated using the classification of experienced radiologists.
