Use of an Artificial Intelligence-enabled Wound Imaging System for Diabetic Foot Ulcers

In this study, the authors—Kai Siang Chan, Yam Meng Chan, Audrey Hui Min Tan, Shanying Liang, Yuan Teng Cho, Qiantai Hong, Enming Yong, Lester Rhan Chaen Chong, Li Zhang, Glenn Wei Leong Tan, Sadhana Chandrasekar, Zhiwen Joseph Lo—aimed to compare intrarater and interrater reliability of an artificial intelligence (AI) enabled wound imaging mobile application system with traditional wound measurement.¹ This study (“Clinical validation of an artificial intelligence-enabled wound imaging mobile application in diabetic foot ulcers”) was published on May 4, 2021, with the International Wound Journal.¹

The authors conducted a single-center, prospective, cross-sectional study on 28 patients with diabetic foot ulcers (DFUs) from June 2020 to January 2021 at a tertiary hospital. Patients who were 21 years or older were included in the study; those who were pregnant, breastfeeding, had a non-DFU ulcer, or unable to give consent to the treatment were excluded. From the 28 patients included in the study, there was a total of 75 wound episodes. There were 547 wound images analyzed.

The results indicated there was excellent intrarater reliability (>0.9) of C4W on 3 separate image captures of the same wound for length, width, and area. The interrater reliability also demonstrated as excellent (>0.9) for length, width, and area between the 3 separate image captures from the C4W device. For comparing the manual measurements taken by a trained wound nurse and their use of the C4W device, the authors reported good (0.75–0.9) interrater reliability for length, width, and area. Their study demonstrated high dependability with intrarater between devices and interrater dependability between the C4W imaging system and traditional wound measurements.

This study had limitations. The authors noted that they were unable to evaluate other characteristics (eg, depth and undermining of wounds). Other qualitative characteristics of the wounds, such as presence of exudates, eschar, and edema, were only reviewed by the wound nurse, not the C4W imaging system.  

The authors concluded that when the wound was assessed using the C4W system against traditional wound measurements by a trained wound specialist, there was good interrater and intrarater reliability in the case of DFUs. As a technical limitation, the system required manual adjustment to avoid capturing inaccurate detection of wound boundaries. However, the authors noted that the inaccuracies will improve as the technology develops and advances.

Read the full study, here.

—Cat Urbanski, Associate Digital Editor

Reference

1. Chan KS, Chan YM, Tan AHM, et al. Clinical validation of an artificial intelligence-enabled wound imaging mobile application in diabetic foot ulcers. Int Wound J. 2022;19(1):114-124. doi:10.1111/iwj.13603