Detecting Apical Lesions on Periapical Radiographs Using Transfer Learning

Detecting Apical Lesions on Periapical Radiographs Using Transfer Learning

Authors: Yueh Chuo; Wen-Ming Lin; Tsung-Yi Chen; Mei-Ling Chan; Yu-Sung Chang; Yan-Ru Lin; Yuan-Jin Lin; Yu-Han Shao; Chiung-An Chen; Shih-Lun Chen; Patricia Angela R. Abu

Summary:

This study proposes a CNN-based system using transfer learning to detect apical lesions on periapical radiographs. Using adaptive thresholding, advanced image enhancement, and multiple CNN architectures, the model achieved up to 96.21% accuracy with AlexNet. The approach improves clinical decision support and reduces diagnostic workload.

Check the original publication:

Key Words:

PA; CNN; tooth disease recognition; image segmentation; image preprocessing

Extracted Data:

Year:

2022

Modality:

Periapical Radiographs

Dataset:

490 images (415 normal, 75 lesion)

Dataset Split:

Training/Validation = 4:1

Network Architecture:

AlexNet; ResNet50; ResNet101; GoogLeNet

Metrics:

Accuracy: 96.21% (AlexNet)

AP - Professional Qty:

Dentists with more than 3 years of experience

AP - Supervisor Presence:

Yes

AP - Experience Level:

More than 3 years

AP - Expertise Area:

General Dentistry

AP - Tool or System:

MATLAB (Deep Network Designer)

ML Task:

Classification (Apical Lesion Detection)

Project Objective:

To develop a high-accuracy automatic apical lesion detection system using CNN transfer learning.

Clinical Relevance:

Automated high-accuracy detection of apical lesions may reduce dentist workload, improve diagnostic consistency, and support timely endodontic decisions.

Watch full video on YouTube