Multimodal Method for Differentiating Various Clinical Forms of Basal Cell Carcinoma and Benign Neoplasms In Vivo - Research

Sep 21, 2026

Upload by Logic

Correct classification of skin lesions is a key step in skin cancer screening, requiring high accuracy and interpretability. This study proposes a multimodal method for distinguishing various clinical forms of basal cell carcinoma and benign neoplasms using machine learning, evaluated on 37 lesions. The approach integrates diffuse reflectance spectroscopy, optical coherence tomography, and high-frequency ultrasound, extracting melanin, erythema, hemoglobin, texture, and tumor-shape features to train a classifier.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages20

Reading Time50 min

AccessFree to download

Abstract

Introduction

Background and risk factors

Clinical forms and treatment implications

diagnostics Article Multimodal Method for Differentiating Various Clinical Forms of Basal Cell Carcinoma and Benign Neoplasms In Vivo Yuriy I. Surkov 1,2,3,*, Isabella A. Serebryakova 1,2, Yana K. Kuzinova 4, Olga M. Konopatskova 3,4, Dmitriy V. Safronov 4, Sergey V. Kapralov 4, Elina A. Genina 1,2 and Valery V. Tuchin 1,2,3,5, * Citation: Surkov, Y.I.; Serebryakova, I.A.; Kuzinova, Y.K.; Konopatskova, O.M.; Safronov, D.V.; Kapralov, S.V.; Genina, E.A.; Tuchin, V.V. Multimodal Method for Differentiating Various Clinical Forms of Basal Cell Carcinoma and Benign Neoplasms In Vivo. Diagnostics 2024, 14, 202 . [https://doi.org/10.3390/](https://doi.org/10.3390/)[ ](https://doi.org/10.3390/)diagnostics14020202 Academic Editors: Fabiano Bini, Pavel A. Shilyagin, Elena B. Kiseleva and Ekaterina Vladimirovna Gubarkova Received: 2 November 2023 Revised: 15 January 2024 Accepted: 15 January 2024 Published: 17 January 2024 1 Institution of Physics, Saratov State University, 410012 Saratov, Russia; [s.izabell2014@gmail.com](s.izabell2014@gmail.com) (I.A.S.); [eagenina@yandex.ru](eagenina@yandex.ru) (E.A.G.) 2 Laboratory of Laser Molecular Imaging and Machine Learning, Tomsk State University, 634050 Tomsk, Russia 3 Laboratory of Biomedical Photoacoustic, Saratov State University, 410012 Saratov, Russia; [o.konopatskova@mail.ru](o.konopatskova@mail.ru) 4 Department of Faculty Surgery and Oncology, Saratov State Medical University, 410012 Saratov, Russia; [yan.prokhorova@gmail.com](yan.prokhorova@gmail.com) (Y.K.K.); [dr.safronof@yandex.ru](dr.safronof@yandex.ru) (D.V.S.); [sergejkapralov@yandex.ru](sergejkapralov@yandex.ru) (S.V.K.) 5 Institute of Precision Mechanics and Control, FRC “Saratov Scientific Centre of the Russian Academy of Sciences”, 410028 Saratov, Russia * Correspondence: [surkov9898@gmail.com](surkov9898@gmail.com) (Y.I.S.); [tuchinvv@mail.ru](tuchinvv@mail.ru) (V.V.T.) Abstract: Correct classification of skin lesions is a key step in skin cancer screening, which requires high accuracy and interpretability. This paper proposes a multimodal method for differentiating various clinical forms of basal cell carcinoma and benign neoplasms that includes machine learning. This study was conducted on 37 neoplasms, including benign neoplasms and five different clinical forms of basal cell carcinoma. The proposed multimodal screening method combines diffuse reflectance spectroscopy, optical coherence tomography and high-frequency ultrasound. Using diffuse reflectance spectroscopy, the coefficients of melanin pigmentation, erythema, hemoglobin content, and the slope coefficient of diffuse reflectance spectroscopy in the wavelength range 650–800 nm were determined. Statistical texture analysis of optical coherence tomography images was used to calculate first-and second-order statistical parameters. The analysis of ultrasound images assessed the shape of the tumor according to parameters such as area, perimeter, roundness and other characteristics. Based on the calculated parameters, a machine learning algorithm was developed to differentiate the various clinical forms of basal cell carcinoma. The proposed algorithm for classifying various forms of basal cell carcinoma and benign neoplasms provided a sensitivity of 70.6 ± 17.3%, specificity of 95.9 ± 2.5%, precision of 72.6 ± 14.2%, F1 score of 71.5 ± 15.6% and mean intersection over union of 57.6 ± 20.1% . Moreover, for differentiating basal cell carcinoma and benign neoplasms without taking into account the clinical form, the method achieved a sensitivity of 89.1 ± 8.0%, specificity of 95.1 ± 0.7%, F1 score of 89 .3 ± 3.4% and mean intersection over union of 82 .6 ± 10.8% . Keywords: basal cell carcinoma; benign neoplasm; diffuse reflectance spectroscopy; optical coherence tomography; high-frequency ultrasound; texture image analysis; multimodal diagnostic method; machine learning Copyright: © 2024 by the authors. Licensee MDPI, Basel, Switzerland. T

FAQs

What is the main goal of the proposed method?

To accurately differentiate various clinical forms of basal cell carcinoma from benign neoplasms while maintaining interpretability.

Which modalities are combined in the multimodal approach?

Diffuse reflectance spectroscopy, optical coherence tomography, and high-frequency ultrasound are used together with machine learning.

How are features extracted from the different modalities?

Diffuse reflectance spectroscopy estimates melanin pigmentation, erythema, hemoglobin content, and spectral slope, while OCT uses first- and second-order texture parameters and ultrasound evaluates tumor shape metrics such as area and perimeter.

Sep 21, 2026

Upload by Logic