Diagnostic accuracy of photographic imaging in the granulation of diabetic ulcers by segmentation

Abstract

Introduction: The evaluation of the state of the ulcers by photographic image is carried out when the red and pink colors are found, which correspond to granulation.

Objective: Determine the sensitivity, specificity and accuracy of the photographic image with respect to the histological study in the granulation of diabetic ulcers.

Methods: The design was a diagnostic test performed on 29 diabetic patients with 45 uninfected diabetic ulcers, in which the direct observation of an area of granulation by photographic image was compared as a diagnostic test in the evaluation referring to the standard by pathological anatomy, through a punch biopsy. The photographic image was obtained using a CATS61 Smartphone and analyzed through segmentation in red and black colors with the ImageJ software. The study was authorized by an ethics committee. Statistics were performed with SPSS 22 and EPIDAT 4.4 softwares.

Results: Diabetic ulcers presented an average of 3.03 ± 2.39 cm long and 2.26 ± 1.62 cm wide; of type 2 ulcer according to Wagner in 73.3 %, and type A ulcer, according to the University of Texas at 60%. Photographic imaging tests showed sensitivity, specificity and accuracy in 90%, 33.3% and 61.6%, respectively.

Conclusions: The identification of the photographic image and the histological study of diabetic ulcers with granulation were feasible. The sensitivity, specificity and accuracy of the photographic image were high, low and moderate.