Use of ImageJ as an image processing method for the assessment of post-surgical bruises

Purpose To develop, through ImageJ, an automated, non-invasive, objective method to analyze images of bruises that allowed quantifying variation in its size overtime for its later use in clinical trials. Methods The ImageJ software was used to create automated macros that were executed on 38 images of the untreated eyes of 19 patients that participated in a post-marketing, randomized, controlled clinical trial that assessed the efficacy of a cream to reduce post-blepharoplasty minor hematomas. Three image processing systems were used with the macros created with ImageJ: RGB, RGB2, and HSB. The area of the bruises and the percentage of reduction were estimated for each one. Ophthalmologists also reviewed photographs by direct visual examination. Results All three processing systems were useful for identifying the area of the bruise and studying its variation over time. RGB2 results were the most consistent with the direct visual examination conducted by ophthalmologists. Conclusion RGB2-automated image processing was considered the most appropriate for bruise analysis and represented an advantage over other manual techniques. However, it will be necessary to test it in clinical trials and other studies with a more significant number of samples and different locations.

Automated summary

An automated method for analyzing bruise images was developed using ImageJ software, with three processing systems utilized to estimate the area of bruises. The RGB2 system showed the most consistent results compared to direct visual examination by ophthalmologists.