Breslow thickness: Geometric interpretation, potential pitfalls, and computer automated estimation

Nico Curti; Giulia Veronesi; Emi Dika; Cosimo Misciali; Emanuela Marcelli; Enrico Giampieri

doi:10.1016/j.prp.2022.154117

Breslow thickness: Geometric interpretation, potential pitfalls, and computer automated estimation

Pathol Res Pract. 2022 Oct:238:154117. doi: 10.1016/j.prp.2022.154117. Epub 2022 Sep 5.

Authors

Nico Curti¹, Giulia Veronesi², Emi Dika³, Cosimo Misciali⁴, Emanuela Marcelli¹, Enrico Giampieri¹

Affiliations

¹ eDIMESLab, Department of Experimental, Diagnostic and Specialty Medicine, University of Bologna, 40138 Bologna, Italy.
² Dermatology Unit, Department of Experimental, Diagnostic and Specialty Medicine, University of Bologna, 40138 Bologna, Italy.
³ Dermatology Unit, IRCCS Azienda OspedalieraUniversitaria di Bologna, Sant'Orsola Hospital, Bologna, Italy; Dermatology Unit, Department of Experimental, Diagnostic and Specialty Medicine, University of Bologna, 40138 Bologna, Italy. Electronic address: emi.dika3@unibo.it.
⁴ Dermatology Unit, IRCCS Azienda OspedalieraUniversitaria di Bologna, Sant'Orsola Hospital, Bologna, Italy.

PMID: 36126452
DOI: 10.1016/j.prp.2022.154117

Abstract

Breslow thickness is one of most important prognostic factor for cutaneous melanoma. To quantify the positions of the melanocytes, the Breslow thickness is defined on a distance metric that is reliable and easy to use in a clinical setting. In this letter, we want to highlight some pitfalls in this distance measurement arising from geometrical issues related to section bending and curling, and their consequences on computer automated estimation.

Keywords: Computational geometry; Computer aided diagnosis; Computer vision; Dermatology; Histology; Melanoma; Pathomics; Whole slide imaging.

Publication types

Letter