TY - JOUR
T1 - Hyperspectral Imaging for Non-invasive Diagnostics of Melanocytic Lesions
AU - Paoli, John
AU - Pölönen, Ilkka
AU - Salmivuori, Mari
AU - Räsänen, Janne
AU - Zaar, Oscar
AU - Polesie, Sam
AU - Koskenmies, Sari
AU - Pitkänen, Sari
AU - Övermark, Meri
AU - Isoherranen, Kirsi
AU - Juteau, Susanna
AU - Ranki, Annamari
AU - Grönroos, Mari
AU - Neittaanmäki, Noora
N1 - Publisher Copyright:
© 2022, Medical Journals/Acta D-V. All rights reserved.
PY - 2022
Y1 - 2022
N2 - Malignant melanoma poses a clinical diagnostic pro-blem, since a large number of benign lesions are excised to find a single melanoma. This study assessed the accuracy of a novel non-invasive diagnostic tech-nology, hyperspectral imaging, for melanoma detection. Lesions were imaged prior to excision and histo-pathological analysis. A deep neural network algorithm was trained twice to distinguish between histopatho-logically verified malignant and benign melanocytic lesions and to classify the separate subgroups. Further-more, 2 different approaches were used: a majority vote classification and a pixel-wise classification. The study included 325 lesions from 285 patients. Of these, 74 were invasive melanoma, 88 melanoma in situ, 115 dysplastic naevi, and 48 non-dysplastic naevi. The study included a training set of 358,800 pixels and a vali-dation set of 7,313 pixels, which was then tested with a training set of 24,375 pixels. The majority vote classification achieved high overall sensitivity of 95% and a specificity of 92% (95% confidence interval (95% CI) 0.024–0.029) in differentiating malignant from benign lesions. In the pixel-wise classification, the overall sensitivity and specificity were both 82% (95% CI 0.005–0.005). When divided into 4 subgroups, the diagnostic accuracy was lower. Hyperspectral imaging provides high sensitivity and specificity in distinguis-hing between naevi and melanoma. This novel method still needs further validation.
AB - Malignant melanoma poses a clinical diagnostic pro-blem, since a large number of benign lesions are excised to find a single melanoma. This study assessed the accuracy of a novel non-invasive diagnostic tech-nology, hyperspectral imaging, for melanoma detection. Lesions were imaged prior to excision and histo-pathological analysis. A deep neural network algorithm was trained twice to distinguish between histopatho-logically verified malignant and benign melanocytic lesions and to classify the separate subgroups. Further-more, 2 different approaches were used: a majority vote classification and a pixel-wise classification. The study included 325 lesions from 285 patients. Of these, 74 were invasive melanoma, 88 melanoma in situ, 115 dysplastic naevi, and 48 non-dysplastic naevi. The study included a training set of 358,800 pixels and a vali-dation set of 7,313 pixels, which was then tested with a training set of 24,375 pixels. The majority vote classification achieved high overall sensitivity of 95% and a specificity of 92% (95% confidence interval (95% CI) 0.024–0.029) in differentiating malignant from benign lesions. In the pixel-wise classification, the overall sensitivity and specificity were both 82% (95% CI 0.005–0.005). When divided into 4 subgroups, the diagnostic accuracy was lower. Hyperspectral imaging provides high sensitivity and specificity in distinguis-hing between naevi and melanoma. This novel method still needs further validation.
KW - hyperspectral imaging
KW - machine learning
KW - malignant melanoma
KW - non-invasive diagnostic
KW - 3121 General medicine, internal medicine and other clinical medicine
U2 - 10.2340/actadv.v102.2045
DO - 10.2340/actadv.v102.2045
M3 - Article
C2 - 36281811
AN - SCOPUS:85141937803
SN - 0001-5555
VL - 102
JO - Acta Dermato-Venereologica
JF - Acta Dermato-Venereologica
M1 - adv00815
ER -