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

Fuente: Zenodo
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Paulauskaite-Taraseviciene, Agne, Sutiene, Kristina, Dimsa, Nojus, Valiukeviciene, Skaidra
Format: Recurso digital
Sprache:Englisch
Veröffentlicht: Zenodo 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866901433160302592
author Paulauskaite-Taraseviciene, Agne
Sutiene, Kristina
Dimsa, Nojus
Valiukeviciene, Skaidra
author_facet Paulauskaite-Taraseviciene, Agne
Sutiene, Kristina
Dimsa, Nojus
Valiukeviciene, Skaidra
contents <p>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.</p>
format Recurso digital
id zenodo_https___doi_org_10_61822_amcs-2024-0041
institution Zenodo
language eng
publishDate 2024
publisher Zenodo
record_format zenodo
spellingShingle Enhancing Multi–Class Prediction of Skin Lesions with Feature Importance Assessment
Paulauskaite-Taraseviciene, Agne
Sutiene, Kristina
Dimsa, Nojus
Valiukeviciene, Skaidra
skin lesion
feature extraction
Graph theory
multi-class prediction
SHAP values
<p>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.</p>
title 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