Enhancing Multi–Class Prediction of Skin Lesions with Feature Importance Assessment

Numerous image processing techniques have been developed for the identification of various types of skin lesions. In real-world scenarios, the specific lesion type is often unknown in advance, leading to a multi-class prediction challenge. The available evidence underscores the importance of employi...

Full description

Saved in:
Bibliographic Details
Main Authors: Paulauskaite-Taraseviciene Agne, Sutiene Kristina, Dimsa Nojus, Valiukeviciene Skaidra
Format: Article
Language:English
Published: Sciendo 2024-09-01
Series:International Journal of Applied Mathematics and Computer Science
Subjects:
Online Access:https://doi.org/10.61822/amcs-2024-0041
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850283286403743744
author Paulauskaite-Taraseviciene Agne
Sutiene Kristina
Dimsa Nojus
Valiukeviciene Skaidra
author_facet Paulauskaite-Taraseviciene Agne
Sutiene Kristina
Dimsa Nojus
Valiukeviciene Skaidra
author_sort Paulauskaite-Taraseviciene Agne
collection DOAJ
description Numerous image processing techniques have been developed for the identification of various types of skin lesions. In real-world scenarios, the specific lesion type is often unknown in advance, leading to a multi-class prediction challenge. The available evidence underscores the importance of employing a comprehensive array of diverse features and subsequently identifying the most important ones as a crucial step in visual diagnostics. For this purpose, we addressed both binary and five-class classification tasks using a small dataset, with skin lesions prevalent in Lithuania. The model was trained using a rich set of 662 features, encompassing both conventional image features and graph-based ones, which were obtained from the superpixel graph generated using Delaunay triangulation. We explored the influence of feature importance determined by SHAP values, resulting in a weighted F1-score of 92.48% for the two-class classification and 71.21% for the five-class prediction.
format Article
id doaj-art-2447fc47dbb54481a31e620a4cc2f920
institution OA Journals
issn 2083-8492
language English
publishDate 2024-09-01
publisher Sciendo
record_format Article
series International Journal of Applied Mathematics and Computer Science
spelling doaj-art-2447fc47dbb54481a31e620a4cc2f9202025-08-20T01:47:48ZengSciendoInternational Journal of Applied Mathematics and Computer Science2083-84922024-09-0134461762910.61822/amcs-2024-0041Enhancing Multi–Class Prediction of Skin Lesions with Feature Importance AssessmentPaulauskaite-Taraseviciene Agne0Sutiene Kristina1Dimsa Nojus2Valiukeviciene Skaidra3Artificial Intelligence Centre Kaunas University of Technology K. Barsausko g. 59, 51423Kaunas, LithuaniaDepartment of Mathematical Modeling Kaunas University of Technology Studentu g. 50, 51368Kaunas, LithuaniaFaculty of Informatics Kaunas University of Technology Studentu g. 50, 51368Kaunas, Lithuania Department of Skin and Venereal Diseases Lithuanian University of Health Sciences A. Mickeviciaus g. 9, 44307Kaunas, LithuaniaNumerous image processing techniques have been developed for the identification of various types of skin lesions. In real-world scenarios, the specific lesion type is often unknown in advance, leading to a multi-class prediction challenge. The available evidence underscores the importance of employing a comprehensive array of diverse features and subsequently identifying the most important ones as a crucial step in visual diagnostics. For this purpose, we addressed both binary and five-class classification tasks using a small dataset, with skin lesions prevalent in Lithuania. The model was trained using a rich set of 662 features, encompassing both conventional image features and graph-based ones, which were obtained from the superpixel graph generated using Delaunay triangulation. We explored the influence of feature importance determined by SHAP values, resulting in a weighted F1-score of 92.48% for the two-class classification and 71.21% for the five-class prediction.https://doi.org/10.61822/amcs-2024-0041skin lesionfeature extractiongraph theorymulti-class predictionshap values.
spellingShingle Paulauskaite-Taraseviciene Agne
Sutiene Kristina
Dimsa Nojus
Valiukeviciene Skaidra
Enhancing Multi–Class Prediction of Skin Lesions with Feature Importance Assessment
International Journal of Applied Mathematics and Computer Science
skin lesion
feature extraction
graph theory
multi-class prediction
shap values.
title Enhancing Multi–Class Prediction of Skin Lesions with Feature Importance Assessment
title_full Enhancing Multi–Class Prediction of Skin Lesions with Feature Importance Assessment
title_fullStr Enhancing Multi–Class Prediction of Skin Lesions with Feature Importance Assessment
title_full_unstemmed Enhancing Multi–Class Prediction of Skin Lesions with Feature Importance Assessment
title_short Enhancing Multi–Class Prediction of Skin Lesions with Feature Importance Assessment
title_sort enhancing multi class prediction of skin lesions with feature importance assessment
topic skin lesion
feature extraction
graph theory
multi-class prediction
shap values.
url https://doi.org/10.61822/amcs-2024-0041
work_keys_str_mv AT paulauskaitetarasevicieneagne enhancingmulticlasspredictionofskinlesionswithfeatureimportanceassessment
AT sutienekristina enhancingmulticlasspredictionofskinlesionswithfeatureimportanceassessment
AT dimsanojus enhancingmulticlasspredictionofskinlesionswithfeatureimportanceassessment
AT valiukevicieneskaidra enhancingmulticlasspredictionofskinlesionswithfeatureimportanceassessment