[Computer-assisted image processing for quantifying histopathologic variables in the healing of colonic anastomosis in dogs]

Abstract

Background: The authors present the experimental results of the computerized quantifying of tissular structures involved in the reparative process of colonic anastomosis performed by manual suture and biofragmentable ring. The quantified variables in this study were: oedema fluid, myofiber tissue, blood vessel and cellular nuclei.

Material and methods: An image processing software developed at Laboratório de Informática Dedicado à Odontologia (LIDO) was utilized to quantifying the pathognomonic alterations in the inflammatory process in colonic anastomosis performed in 14 dogs. The results were compared to those obtained through traditional way diagnosis by two pathologists in view of counterproof measures. The criteria for these diagnoses were defined in levels represented by absent, light, moderate and intensive which were compared to analysis performed by the computer.

Results: There was significant statistical difference between two techniques: the biofragmentable ring technique exhibited low oedema fluid, organized myofiber tissue and higher number of alongated cellular nuclei in relation to manual suture technique.

Conclusion: The analysis of histometric variables through computational image processing was considered efficient and powerful to quantify the main tissular inflammatory and reparative changing.