Pelatihan Deteksi Lesi Periapikal Berbasis CNN–LSTM Retinex untuk Diagnosis Radiografis Dokter Gigi
DOI:
https://doi.org/10.58466/x5vqxr05Keywords:
Community Service, Periapical Lesion Detection, Deep Learning, CNN–LSTM, Retinex, Artificial Intelligence, Radiographic DiagnosisAbstract
Accurate periapical lesion diagnosis represents a significant clinical challenge for dentists, particularly in primary care clinics with limited access to Computer-Aided Diagnosis (CAD) systems. This Community Service (PKM) program aims to train dental medical personnel in using an AI-based periapical lesion detection system integrating CNN–LSTM architecture with Retinex image enhancement. The program was conducted at Primary Dental Clinics in the City Region over three months (October–December 2024), involving 32 dentists and 8 radiographers. Methods included socialization, intensive workshops, clinical case simulations, and technical mentoring. Evaluation results showed an average competency score increase of 34.7% (from 61.2 to 82.4 out of 100), with 87.5% of participants successfully operating the system independently. Participant satisfaction reached 89.1% (very satisfied category). The implemented AI system achieved periapical lesion detection accuracy of 98.4% with 97% recall, far exceeding manual diagnosis sensitivity averages of 70–85%. This PKM demonstrates that AI diagnostic technology transfer to primary dental clinical practitioners is feasible and significantly impacts dental healthcare quality improvement
Downloads
References
Chau, K., et al. (2025). A novel AI model for detecting periapical lesion on CBCT: CBCT-SAM. Journal of Dentistry, 153. https://doi.org/10.1016/j.jdent.2025.xx
Estrela, S., et al. (2008). Characterization of periapical lesions by cone-beam CT. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology & Endodontology, 106(6), 909–914.
Khan, A., et al. (2023). Hybrid CNN–RNN architectures for medical image classification: A systematic review. Computers in Biology and Medicine, 153, 106481.
Kurniawan, S. D. (2024). Hybrid Deep Learning Architecture Integrating CNN–LSTM and Retinex Enhancement for Automated Detection of Periapical Lesions on CBCT and Panoramic Radiographs. Smart Comp: Jurnalnya Orang Pintar Komputer, 13(1), 122–135. https://doi.org/10.30591/smartcomp.v13i1.4272
Latke, V., & Narawade, V. (2024). Detection of dental periapical lesions using Retinex based image enhancement and lightweight deep learning model. Image and Vision Computing, 146, 104998.
Patel, P., et al. (2019). Cone-beam computed tomography in endodontics. International Endodontic Journal, 52(10), 1360–1373.
Setzer, H., et al. (2018). Comparison of periapical diagnosis using CBCT and conventional radiography. Journal of Endodontics, 44(7), 1091–1098.
Ver Berne, J., et al. (2025). Automated classification of panoramic radiographs with inflammatory periapical lesions using a CNN-LSTM architecture. Journal of Dentistry, 156, 105387.
Wang, K., et al. (2024). Deep learning-based efficient diagnosis of periapical diseases with dental X-rays. Image and Vision Computing, 147, 105053.
Zhang, Y., et al. (2022). Illumination normalization with Retinex for dental panoramic image analysis. Medical Physics, 49(7), 4776–4788.
Diranna, K., Osmundson, E., Topps, J., Barakos, L., Gearhart, M., Cerwin, K., …, Strang, C. (2008). Asessment-centered teaching (A reflective practice). London: Sage.
Ermasari, G., Subagia, I. W., & Sudria, I. B. N. (2014). Kemampuan bertanya guru IPA dalam pengelolaan pembelajaran. Jurnal Pendidikan dan Pembelajaran IPA Indonesia, 4(1), 1-12. Retrieved from http://oldpasca.undiksha.ac.id/e-journal/index.php/jurnal_ipa/article/view/1111.
Feldt, L. S., & Brennan, R. (1989). Reliability. In R. L. Linn (Ed), Educational measurement (3rd ed.). New York, NY: Macmillan.
