Dermatology apps faced controversies when they were first launched but soon received respite when machine learning (ML) technologies were used for digital skin examinations.

Though advantages of clinically-tested automatic examinations are clearly evident, a recent paper published in JAMA Dermatology says that the guiding techniques for these algorithms may not give a correct diagnosis when examining patients with colored skin.

“Recently, ML has been used to create programs capable of distinguishing between images of benign and malignant moles with accuracy similar to that of board-certified dermatologists,” Dr. Adewole S. Adamson, of the department of dermatology at University of North Carolina at Chapel Hill, and Avery Smith, a software engineer at Fearless Solutions in Baltimore, noted in the journal. “This technology could greatly assist dermatologists in diagnosing and treating skin diseases, thereby improving patient care. However, if not developed with inclusivity in mind, ML could exacerbate healthcare disparities in dermatology.”

ML presents with varied applications in Dermatology owing to its visual capabilities. However, constructing an efficient algorithm still necessitates a large collection of high-quality data, the two wrote. Hitherto, however, most of the sample data is sourced from people with lighter skin color.

“In particular, this limitation could potentially have concerning consequences in the diagnosis of melanomas, which can look different on dark skin,” the authors wrote. “Although melanoma incidence is significantly more common among non-Hispanic white persons, this does not mean that patients of darker skin types should be excluded from potential benefits of early detection through ML.”

The authors emphasized the need for more photos of skin conditions occurring in patients with darker skin color in the database to adjust the inequality. Moreover, they suggested the development of ML software that can identify melanoma over varied skin types, or the creation of distinct algorithms exclusively intended to overcome this shortcoming.