A two-stage workflow for vitiligo diagnosis: clinical characteristic classification and large language model (LLM)–based report generation - Takeaways - MDSpire

A two-stage workflow for vitiligo diagnosis: clinical characteristic classification and large language model (LLM)–based report generation

  • By

  • Kaiqiao He

  • Tianle Xu

  • Yining Feng

  • Yafei Lu

  • Xinju Wang

  • Linhan Jiang

  • Sen Guo

  • Yuanmin He

  • Wei Dai

  • Wei Zhang

  • Jianglin Zhang

  • Hongbing Lu

  • Dong Huang

  • Shuli Li

  • June 1, 2026

  • 0 min

Share

  • 1

    A dual-phase AI-assisted diagnostic system was developed to differentiate vitiligo from ten other hypopigmentary disorders.

  • 2

    The model achieved an AUC of 0.9906, with sensitivity of 98.29% and specificity of 93.73% in distinguishing vitiligo.

  • 3

    The AI model outperformed dermatologists in diagnostic sensitivity during tests with an independent set of 175 images.

  • 4

    The system generates structured clinical reports using a large language model, enhancing diagnostic transparency and decision-making.

  • 5

    This study addresses misdiagnosis in vitiligo, offering potential support in resource-limited settings through intelligent tools.

Original Source(s)

A two-stage workflow for vitiligo diagnosis: clinical characteristic classification and large language model (LLM)–based report generation

  • by Kaiqiao He, Tianle Xu, Yining Feng, Yafei Lu, Xinju Wang, Linhan Jiang, Sen Guo, Yuanmin He, Wei Dai, Wei Zhang, Jianglin Zhang, Hongbing Lu, Dong Huang, Shuli Li

  • June 1, 2026

Related Content

Climate, Humidity, and Population-Level Interest in Dry Skin: Infodemiology Analysis Using Google Trends Across the United States

Integrated systems toxicology identifies TCDD-responsive targets linked to immune dysregulation and treatment response in psoriasis

Single-cell landscape of immune remodeling in alopecia areata suggests MIF + fibroblasts and their potential ligand-receptor crosstalk with dendritic cells