High dynamic range conversion of dermoscopy images in general dermatology conditions-A pilot study.

INTRODUCTION

Converting normal dermoscopy images to high dynamic range (HDR) images has been described to improve the quality of imaging in the context of pigmented skin tumors and vitiligo. We aimed to evaluate the effectiveness of HDR conversion in improving visualization of dermoscopy features in general non-malignant, dermatological conditions.

METHODS

Fifteen lesions were included in the study. Dermoscopy images (polarized 10×) were captured using a Dermlite® Foto ii Pro dermoscope attached to a Canon® 650D digital SLR camera. HDR images were created by using Topaz® software to convert the original dermoscopy image to HDR. All editing work was done using Adobe Photoshop CS6® (HDR software was used as a plug-in). Conversion parameters were kept uniform for all sets of image. Dermoscopy images were viewed by six dermatologists, who were blinded to the label, and for each image, the quality of visualization of dermoscopy features was graded on a global scale of 1 to 5 (1 lowest quality to 5 highest quality). For statistical analysis, Mann-Whitney U test was used to compare mean ratings across the four groups.

RESULTS

The mean rankings for the HDR converted images were higher than that of the original images. The difference was statistically significant (P < .001).

CONCLUSION

High dynamic range conversion of dermoscopy images is a technique, which can be effective in improving quality of visualization of dermoscopic features in general dermatology.