Multiclass Kidney CT-Scan Classification Using Gabor Feature Extraction and Random Forest
JOIV : International Journal on Informatics Visualization · 31.05.2026 · 10 (3) · 1013
DOI: 10.62527/joiv.10.3.4211 ↗Kurz gefasst
Fachartikel in der Zeitschrift JOIV : International Journal on Informatics Visualization, online erschienen am 31.05.2026. Verfasst von Dede Kurniadi und 3 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).
Abstract (Originalsprache)
The kidneys are vital organs vulnerable to disorders such as stones, cysts, and tumors, which are often difficult to detect in early stages. CT scans play an essential role in diagnosis, yet manual interpretation is subjective and time-consuming. This study proposes an automatic multi-class kidney CT-scan classification model using the Machine Learning Life Cycle (MLLC) framework. The method combines Gabor-based texture extraction and the Random Forest algorithm. The dataset consists of 11,929 images categorized into four classes. Model training was performed using stratified 5-fold cross-validation, and generalization ability was evaluated through an inference phase on unseen data. The model using Gabor features achieved an accuracy of 99.74% ± 0.17%, precision of 99.78% ± 0.14%, recall of 99.67% ± 0.20%, F1-score of 99.72% ± 0.17%, and specificity of 99.91% ± 0.06%. In contrast, the model based on basic image features yielded 53.74% ± 0.75% accuracy, indicating its limited capability in capturing discriminative patterns. The Gabor-based approach demonstrated strong consistency, although a few misclassifications still occurred due to texture similarities among classes. The proposed model has potential for clinical decision support, particularly in resource-constrained healthcare settings that require lightweight and interpretable solutions. These findings also highlight opportunities for further research, such as combining Gabor-based texture representations with deeper CNN architectures and validating the model using multi-institution clinical datasets to enhance robustness and real-world applicability.
Zitieren
APA
Kurniadi, D., Fernando, E., Mulyani, A., & Aprinelia, M. M. (2026). Multiclass Kidney CT-Scan Classification Using Gabor Feature Extraction and Random Forest. JOIV : International Journal on Informatics Visualization, 10(3), 1013. https://doi.org/10.62527/joiv.10.3.4211
Vancouver
Kurniadi D, Fernando E, Mulyani A, Aprinelia MM. Multiclass Kidney CT-Scan Classification Using Gabor Feature Extraction and Random Forest. JOIV : International Journal on Informatics Visualization. 2026;10(3):1013. doi:10.62527/joiv.10.3.4211
BibTeX
@article{Kurniadi2026,
author = {Kurniadi, Dede and Fernando, Erick and Mulyani, Asri and Aprinelia, Muthia Mutmainah},
title = {Multiclass Kidney CT-Scan Classification Using Gabor Feature Extraction and Random Forest},
journal = {JOIV : International Journal on Informatics Visualization},
year = {2026},
volume = {10},
number = {3},
pages = {1013},
doi = {10.62527/joiv.10.3.4211}
}
EUVIMED