Use of Segmentation and Color Adjustment for Skin Tone Classification in Dermatological Images

Authors: F. Duarte

Abstract:

The work aims to evaluate the use of classical image processing methodologies towards skin tone classification in dermatological images. The skin tone is an important attribute when considering several factor for skin cancer diagnosis. Currently, there is a lack of clear methodologies to classify the skin tone based only on the dermatological image. In this work, a recent released dataset with the label for skin tone was used as reference for the evaluation of classical methodologies for segmentation and adjustment of color space for classification of skin tone in dermatological images. It was noticed that even though the classical methodologies can work fine for segmentation and color adjustment, classifying the skin tone without proper control of the acquisition of the sample images ended being very unreliable.

Keywords: Segmentation, classification, color space, skin tone, Fitzpatrick.

References:

[1] Kumar, A., Aelgani, V., Vohra, R. et al. Artificial intelligence bias in medical system designs: a systematic review. Multimed Tools Appl 83, 18005–18057 (2024).
[2] Gupta AK, Bharadwaj M, Mehrotra R. Skin Cancer Concerns in People of Color: Risk Factors and Prevention. Asian Pac J Cancer Prev. 2016 Dec 1;17(12):5257-5264.
[3] Kinyanjui, Newton M., et al. Estimating skin tone and effects on classification performance in dermatology datasets. arXiv preprint arXiv:1910.13268, 2019.
[4] Li X, Cui Z, Wu Y, Gu L, Harada T. Estimating and improving fairness with adversarial learning. arXiv:210304243
[cs]. Published online May 11, 2021.
[5] Barron, J. T. (2020). A generalization of Otsu’s method and minimum error thresholding. arXiv preprint arXiv:2007.07350.
[6] YKalb, T., Kushibar, K., Cintas, C., Lekadir, K., Diaz, O., Osuala, R. (2023, August 18). Revisiting Skin Tone Fairness in Dermatological Lesion Classification. arXiv.org. http://arxiv.org/abs/2308.09640
[7] Fitzpatrick, T.B.: The Validity and Practicality of Sun-Reactive Skin Types I Through VI. Archives of Dermatology 124(6), 869–871 (06 1988). https://doi.org/10.1001/archderm.1988.01670060015008
[8] Tschandl, P. (2018, January 1). The HAM10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions. Harvard Dataverse. https://doi.org/10.7910/dvn/dbw86t
[9] Lara, M. a. R., Kowalczuk, M. V. R., Eliceche, M. L., Ferraresso, M. G., Luna, D. R., Benitez, S. E., Mazzuoccolo, L. D. (2023). A dataset of skin lesion images collected in Argentina for the evaluation of AI tools in this population. Scientific Data, 10(1). https://doi.org/10.1038/s41597-023-02630-0