EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Smartphone-Based Dental Caries Classification Using Shape and Color Features with a Multilayer Perceptron

Yessi Jusman, Masayu Alya Nur ‘Aini, Muhammad Exsa Agniyanur, Sartika Puspita, Siti Nurul Aqmariah Mohd Kanafiah

BIO Web of Conferences · 2026 · 265 · 02008

DOI: 10.1051/bioconf/202626502008 ↗

Kurz gefasst

Fachartikel in der Zeitschrift BIO Web of Conferences, online erschienen am 17.11.2026. Verfasst von Yessi Jusman und 4 weiteren Autor:innen. Die Arbeit nennt 15 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Dental caries classification using smartphone-acquired images offers a potential approach to supporting accessible preliminary dental assessment. This study presents a dental caries classification system based on occlusal-view dental images captured using a smartphone. Image preprocessing was performed using resizing and Contrast Limited Adaptive Histogram Equalization (CLAHE) to standardize image dimensions and enhance local contrast. Subsequently, Color Moments and Fourier Shape Descriptors were combined to extract color and shape characteristics. Classification was conducted using a Multilayer Perceptron (MLP) trained with two algorithms: OSS and Conjugate Gradient Backpropagation with Powell–Beale Restarts (CGB). Both algorithms were evaluated using four hidden-node configurations (5, 10, 15, and 20). The results showed that MLP-OSS with 20 hidden nodes achieved the highest average testing accuracy of 74.17%, while MLP-CGB with 15 hidden nodes achieved the highest testing accuracy among the CGB configurations, at 70.18%. Although OSS with 20 hidden nodes provided the highest average testing accuracy, CGB with 15 hidden nodes exhibited lower variability in testing performance. These findings demonstrate the potential of combining handcrafted color and shape features with MLP classifiers for smartphone-based dental caries classification.

Zitieren

APA

Jusman, Y., ‘Aini, M. A. N., Agniyanur, M. E., Puspita, S., & Kanafiah, S. N. A. M. (2026). Smartphone-Based Dental Caries Classification Using Shape and Color Features with a Multilayer Perceptron. BIO Web of Conferences, 265, 02008. https://doi.org/10.1051/bioconf/202626502008

Vancouver

Jusman Y, ‘Aini MAN, Agniyanur ME, Puspita S, Kanafiah SNAM. Smartphone-Based Dental Caries Classification Using Shape and Color Features with a Multilayer Perceptron. BIO Web of Conferences. 2026;265:02008. doi:10.1051/bioconf/202626502008

BibTeX
@article{Jusman2026,
  author = {Jusman, Yessi and ‘Aini, Masayu Alya Nur and Agniyanur, Muhammad Exsa and Puspita, Sartika and Kanafiah, Siti Nurul Aqmariah Mohd},
  title = {Smartphone-Based Dental Caries Classification Using Shape and Color Features with a Multilayer Perceptron},
  journal = {BIO Web of Conferences},
  year = {2026},
  volume = {265},
  pages = {02008},
  doi = {10.1051/bioconf/202626502008}
}
Kostenlos starten →