CBI + R: A Fusion Approach to Assist Dermatological Diagnoses.

Autor: Jiji, G. Wiselin, Rajesh, A., Raj, P. Johnson Durai
Předmět:
Zdroj: International Journal of Image & Graphics; Jan2021, Vol. 21 Issue 1, pN.PAG-N.PAG, 19p
Abstrakt: With the emerge of advanced technologies such as high-resolution cameras and computational power, it seems to ease to built a better dermatological diagnostic system. However, the identification of skin disease is still a challenging problem with the origination of various skin diseases. In this paper, we proposed a new fusion architecture — CBI + R to support the diagnosis in multiple skin diseases. The architecture combines Content-Based Image Retrieval (CBIR) and Case-Based Reasoning (CBR) technology together to facilitate medical diagnosis. CBIR is to retrieve digital dermoscopy images from a data repository using the shape, texture and color features. Along with these features, CBR is incorporated which contains symptoms, case history and treatment plan of the disease. Experiments on a set of 1210 images yielded an accuracy of 98.2%. This was a superior retrieval and diagnosis performance in comparison with the state-of-the-art works. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index