PhD Scientific Days 2022

Budapest, 6-7 July 2022

Clinical Medicine IV.

Quantitative Multispectral Imaging for the Estimation of the Breslow Thickness of Melanoma

Noémi Nóra Varga1, Szabolcs Bozsányi1, Klára Farkas1, András Bánvölgyi1, Kende Lőrincz1, Ilze Lihacova2, Alexey Lihachev2, Emilija Vija Plorina2, Áron Bartha3, Antal Jobbágy1, Enikő Kuroli1,4, György Paragh5, Péter Holló1, Márta Medvecz1, Norbert Kiss1 and Norbert M. Wikonkál1

1 Department of Dermatology, Venereology and Dermatooncology, Semmelweis University, Budapest, Hungary
2 Biophotonics Laboratory, Institute of Atomic Physics and Spectroscopy, University of Latvia, Riga, Latvia
3 2nd Department of Pediatrics, Semmelweis University, Budapest, Hungary
4 1st Department of Pathology and Experimental Cancer Research, Semmelweis University, Budapest, Hungary
5 Department of Dermatology, Roswell Park Comprehensive Cancer Center, Buffalo, NY, United States

Text of the abstract

Introduction: Melanoma is a malignant melanocytic tumor that is responsible for the most skin cancer-related deaths. Histopathological evaluation after surgical excision is the gold standard method for diagnosing melanoma. Based on the Breslow thickness, the histologically confirmed tumor depth, the clinicians define the required surgical safety margin. However, it is not available at the time of the initial melanoma diagnosis and decision making.
Aims: In this work, we aimed to develop a novel algorithm using multispectral imaging (MSI) for the prediction of Breslow thickness of melanoma. Furthermore, we aimed compare the performance of this algorithm to dermoscopy.
Method: In this study, we have examined 100 patients with histologically verified primary melanomas. Based on the shape descriptors and intensity values of the MSI images, we created an algorithm to classify melanomas into three subgroups according to tumor thickness: ≤1 mm, 1-2 mm and >2 mm. Clinical photographs and dermoscopic images of the 100 melanomas were shown to dermatologists and dermatology residents to classify the lesion based on tumor thickness.
Results: We developed a novel MSI algorithm that was capable of the classification of melanomas into the above-mentioned three subgroups with the sensitivity of 78.00% and the specificity of 89.00% and a substantial agreement (κ = 0.67). We compared our results to the performance of dermatologists and dermatology residents who reached a sensitivity of 60.38% and specificity of 80.86% with a moderate agreement (κ = 0.41).
Conclusion: To the best of our knowledge, we were the first to analyze melanoma tumor thickness using MSI to classify melanomas into 3 subgroups of great clinical relevance. When we compared our findings to that of dermatologist and dermatology residents, we found superior sensitivity and specificity. Based on our findings, this novel imaging method may be used predict the appropriate safety margins for curative melanoma excision based on the estimated Breslow thickness.
Funding: This work was supported by grants from the EFOP-3.6.3-VEKOP-16-2017-00009 (S.B., N.N.V.), the ÚNKP-21-4-II-SE-10 (N.K.) and ÚNKP-21-2-I-SE-41 (N.N.V.).