Fitzpatrick Thresholding for Skin Image Segmentation

Accurate estimation of the body surface area (BSA) involved by a rash, such as psoriasis, is critical for assessing rash severity, selecting an initial treatment regimen, and following clinical treatment response. Attempts at segmentation of inflammatory skin disease such as psoriasis perform markedly worse on darker skin tones, potentially impeding equitable care. We assembled a psoriasis dataset sourced from six public atlases, annotated for Fitzpatrick skin type, and added detailed segmentation masks for every image. Reference models based on U-Net, ResU-Net, and SETR-small are trained without tone information. On the tuning split we sweep decision thresholds and select (i) global optima and (ii) per Fitzpatrick skin tone optima for Dice and binary IoU. Adapting Fitzpatrick specific thresholds lifted segmentation performance for the darkest subgroup (Fitz VI) by up to +31 % bIoU and +24 % Dice on UNet, with consistent, though smaller, gains in the same direction for ResU-Net (+25 % bIoU, +18 % Dice) and SETR-small (+17 % bIoU, +11 % Dice). Because Fitzpatrick skin tone classifiers trained on Fitzpatrick-17k now exceed 95 % accuracy, the cost of skin tone labeling required for this technique has fallen dramatically. Fitzpatrick thresholding is simple, model-agnostic, requires no architectural changes, no re-training, and is virtually cost free. We demonstrate the inclusion of Fitzpatrick thresholding as a potential future fairness baseline.

• Publication year

  2025

• Venue

  ISIC@MICCAI

• Publication date

  2025-10-08

• Fields of study

  Medicine, Computer Science, Engineering

• Identifiers
  DOI 10.1007/978-3-032-05825-6_2arXiv 2510.06655
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• A review of psoriasis image analysis based on machine learning

  2024cited by this paper
• CIRCLe: Color Invariant Representation Learning for Unbiased Classification of Skin Lesions

  2022cited by this paper
• FairPrune: Achieving Fairness Through Pruning for Dermatological Disease Diagnosis

  2022cited by this paper
• Estimation error of the body surface area in psoriasis: a comparative study of physician and computer‐assisted image analysis (ImageJ)

  2022cited by this paper
• Evaluating Deep Neural Networks Trained on Clinical Images in Dermatology with the Fitzpatrick 17k Dataset

  2021influential reference
• Disparities in dermatology AI performance on a diverse, curated clinical image set

  2021cited by this paper
• Representation and Misdiagnosis of Dark Skin in a Large-Scale Visual Diagnostic Challenge.

  2021cited by this paper
• Fairness of Classifiers Across Skin Tones in Dermatology

  2020cited by this paper
• Comparative analysis of the body surface area calculation method used in vitiligo extent score vs the hand unit method used in vitiligo area severity index

  2020cited by this paper
• Measurement properties of the product of investigator’s global assessment and body surface area in children and adults with atopic dermatitis

  2020cited by this paper
• Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers

  2020cited by this paper
• The VIA Annotation Software for Images, Audio and Video

  2019cited by this paper
• Road Extraction by Deep Residual U-Net

  2017cited by this paper
• On Fairness and Calibration

  2017cited by this paper
• The reliability of three psoriasis assessment tools: Psoriasis area and severity index, body surface area and physician global assessment.

  2017cited by this paper
• Automatic psoriasis lesion segmentation in two-dimensional skin images using multiscale superpixel clustering

  2017cited by this paper
• Equality of Opportunity in Supervised Learning

  2016cited by this paper
• Inherent Trade-Offs in the Fair Determination of Risk Scores

  2016cited by this paper
• U-Net: Convolutional Networks for Biomedical Image Segmentation

  2015cited by this paper

• No citing papers are available for this paper.