Neural network classifier of hyperspectral images of skin pathologies

Vseslav O. Vinokurov; Irina A. Matveeva; Yulia A. Khristoforova; Oleg O. Myakinin; Ivan A. Bratchenko; Lyudmila A. Bratchenko; Alexander A. Moryatov; Sergey V. Kozlov; Alexander S. Machikhin; Ibrahim Abdulhalim; Valery P. Zakharov

doi:10.18287/2412-6179-CO-832

Neural network classifier of hyperspectral images of skin pathologies

Vseslav O. Vinokurov, Irina A. Matveeva, Yulia A. Khristoforova, Oleg O. Myakinin, Ivan A. Bratchenko, Lyudmila A. Bratchenko, Alexander A. Moryatov, Sergey V. Kozlov, Alexander S. Machikhin, Ibrahim Abdulhalim, Valery P. Zakharov

Research output: Contribution to journal › Article › peer-review

9 Scopus citations

Abstract

The paper presents results of using a neural network classifier to analyze images of malignant skin lesions obtained using a hyper-spectral camera. Using a three-block neural network of VGG architecture, we conducted the classification of a set of two-dimensional images of melanoma, papilloma and basal cell carcinoma, obtained in the range of 530 – 570 and 600 – 606 nm, characterized by the highest absorption of melanin and hemoglobin. The sufficiency of the inclusion in the training set of two-dimensional images of a limited spectral range is analyzed. The results obtained show significant prospects of using neural network algorithms for processing hyperspectral data for the classification of skin pathologies. With a relatively small set of training data used in the study, the classification accuracy for the three types of neoplasms was as high as 96 %.

Original language	English
Pages (from-to)	879-886
Number of pages	8
Journal	Computer Optics
Volume	45
Issue number	6
DOIs	https://doi.org/10.18287/2412-6179-CO-832
State	Published - 1 Nov 2021

Keywords

Basal cell carcinoma
Hemoglobin
Hyperspectral imaging
Melanin
Melanoma
Neural network classifier
Oncopathology
VGG

ASJC Scopus subject areas

Engineering (miscellaneous)
Computer Vision and Pattern Recognition

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.18287/2412-6179-CO-832

Cite this

@article{d420d5a4d37f41e6bdc33a3316eac6fb,

title = "Neural network classifier of hyperspectral images of skin pathologies",

abstract = "The paper presents results of using a neural network classifier to analyze images of malignant skin lesions obtained using a hyper-spectral camera. Using a three-block neural network of VGG architecture, we conducted the classification of a set of two-dimensional images of melanoma, papilloma and basal cell carcinoma, obtained in the range of 530 – 570 and 600 – 606 nm, characterized by the highest absorption of melanin and hemoglobin. The sufficiency of the inclusion in the training set of two-dimensional images of a limited spectral range is analyzed. The results obtained show significant prospects of using neural network algorithms for processing hyperspectral data for the classification of skin pathologies. With a relatively small set of training data used in the study, the classification accuracy for the three types of neoplasms was as high as 96 %.",

keywords = "Basal cell carcinoma, Hemoglobin, Hyperspectral imaging, Melanin, Melanoma, Neural network classifier, Oncopathology, VGG",

author = "Vinokurov, {Vseslav O.} and Matveeva, {Irina A.} and Khristoforova, {Yulia A.} and Myakinin, {Oleg O.} and Bratchenko, {Ivan A.} and Bratchenko, {Lyudmila A.} and Moryatov, {Alexander A.} and Kozlov, {Sergey V.} and Machikhin, {Alexander S.} and Ibrahim Abdulhalim and Zakharov, {Valery P.}",

year = "2021",

month = nov,

day = "1",

doi = "10.18287/2412-6179-CO-832",

language = "English",

volume = "45",

pages = "879--886",

journal = "Computer Optics",

issn = "0134-2452",

publisher = "Institution of Russian Academy of Sciences",

number = "6",

}

TY - JOUR

T1 - Neural network classifier of hyperspectral images of skin pathologies

AU - Vinokurov, Vseslav O.

AU - Matveeva, Irina A.

AU - Khristoforova, Yulia A.

AU - Myakinin, Oleg O.

AU - Bratchenko, Ivan A.

AU - Bratchenko, Lyudmila A.

AU - Moryatov, Alexander A.

AU - Kozlov, Sergey V.

AU - Machikhin, Alexander S.

AU - Abdulhalim, Ibrahim

AU - Zakharov, Valery P.

PY - 2021/11/1

Y1 - 2021/11/1

N2 - The paper presents results of using a neural network classifier to analyze images of malignant skin lesions obtained using a hyper-spectral camera. Using a three-block neural network of VGG architecture, we conducted the classification of a set of two-dimensional images of melanoma, papilloma and basal cell carcinoma, obtained in the range of 530 – 570 and 600 – 606 nm, characterized by the highest absorption of melanin and hemoglobin. The sufficiency of the inclusion in the training set of two-dimensional images of a limited spectral range is analyzed. The results obtained show significant prospects of using neural network algorithms for processing hyperspectral data for the classification of skin pathologies. With a relatively small set of training data used in the study, the classification accuracy for the three types of neoplasms was as high as 96 %.

AB - The paper presents results of using a neural network classifier to analyze images of malignant skin lesions obtained using a hyper-spectral camera. Using a three-block neural network of VGG architecture, we conducted the classification of a set of two-dimensional images of melanoma, papilloma and basal cell carcinoma, obtained in the range of 530 – 570 and 600 – 606 nm, characterized by the highest absorption of melanin and hemoglobin. The sufficiency of the inclusion in the training set of two-dimensional images of a limited spectral range is analyzed. The results obtained show significant prospects of using neural network algorithms for processing hyperspectral data for the classification of skin pathologies. With a relatively small set of training data used in the study, the classification accuracy for the three types of neoplasms was as high as 96 %.

KW - Basal cell carcinoma

KW - Hemoglobin

KW - Hyperspectral imaging

KW - Melanin

KW - Melanoma

KW - Neural network classifier

KW - Oncopathology

KW - VGG

UR - http://www.scopus.com/inward/record.url?scp=85120825142&partnerID=8YFLogxK

U2 - 10.18287/2412-6179-CO-832

DO - 10.18287/2412-6179-CO-832

M3 - Article

AN - SCOPUS:85120825142

SN - 0134-2452

VL - 45

SP - 879

EP - 886

JO - Computer Optics

JF - Computer Optics

IS - 6

ER -

Neural network classifier of hyperspectral images of skin pathologies

Abstract

Keywords

ASJC Scopus subject areas

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this