Artificial Intelligence in Medicine: AI Contribution to Wound Assessment and Care

Go online and do a search for “how to measure a wound.” You will find general principles, but you’ll also discover small variations in instructions on how to measure an open wound. How long is it? How wide? Are there irregular spaces below the skin surface or the floor of the wound where infection could set in? Assessing size, shape and irregularities is key for determining and monitoring treatment but challenging to standardize.

Currently, there are no universal criteria for quantitative and qualitative assessment. If two doctors evaluate a wound, they may not be in 100% agreement with each other, even though there are published, illustrated guidelines for wound measurement. Photographs of a wound show a disposable ruler positioned next to its edges an objective indicator of length and width, but precision can vary depending on who is doing the measuring. Some studies have shown that manually measuring by ruler can overestimate the wound size by as much as 40%.^[i]

Photographs also reveal what is called granulation tissue, and the amount and color of this substrate also decisions about wound care, and perceptions of how well it is healing. The percentage of granulation tissue (PGT) must be estimated at the time of the initial wound measurement, but this too can differ from one person to another.

Artificial intelligence as an assessment tool

A May, 2021 paper reports the development of an Artificial Intelligence (AI) algorithm as a wound assessment tool. The authors’ objective was to “evaluate the performance of AI-based software for wound assessment against manual wound assessments performed by wound care clinicians.” The team selected 199 photographs of various wounds with rulers positioned so as to show length and width measurements. Using this dataset, the team trained their AI algorithm to annotate quantitative wound dimensions and qualitative appearance of PGT. In addition, four wound care specialists completed their own annotations. To compute error (false positive and false negative areas), annotations were combined and compared by a) AI vs human and b) human vs human.

Findings

The paper established how crucial accurate wound assessment is, both quantitative and qualitative, for proper wound management and monitoring. Findings from the study conducted by the research team confirmed two key points:

a) Defining criterion-standard wound area and granulation tissue annotations for a broad range of wound types is challenging, and

b) AI technologies have the capacity to perform wound annotations with proficiency similar to human wound care specialists.^[ii]

Conclusion

This study provided an evaluation of an AI-based digital wound assessment tool. Wound specialists need “clear, quick, and precise wound analysis to guide best clinical practices,” yet human perceptions and biases pose a challenge to creating universal agreement on wound assessment, which in turn means a multitude of views on “key wound healing end points.”

The authors are optimistic that AI can not only improve the speed and accuracy of wound assessment, with improved patient outcomes, but it can also aid in standardizing wound assessment. This will be especially useful for managing chronic wounds, which is costly in terms of healthcare dollars and defeating for patients and physicians alike.

Wound care is yet another clinical area in which AI has enormous potential benefit.

NOTE: This content is solely for purposes of information and does not substitute for diagnostic or medical advice. Talk to your doctor if you are experiencing pelvic pain, or have any other health concerns or questions of a personal medical nature.

[i] Howell RS, Liu HH, Khan AA, et al. Development of a Method for Clinical Evaluation of Artificial Intelligence–Based Digital Wound Assessment Tools. JAMA Netw Open. 2021;4(5):e217234.
[ii] Ibid.