A pilot study of an automated personal identification process: Applying machine learning to panoramic radiographs

Abstract

Purpose This study aimed to assess the usefulness of machine learning and automation techniques to match pairs of panoramic radiographs for personal identification. Materials and Methods Two hundred panoramic radiographs from 100 patients (50 males and 50 females) were randomly selected from a private radiological service database. Initially, 14 linear and angular measurements of the radiographs were made by an expert. Eight ratio indices derived from the original measurements were applied to a statistical algorithm to match radiographs from the same patients, simulating a semi-automated personal identification process. Subsequently, measurements were automatically generated using a deep neural network for image recognition, simulating a fully automated personal identification process. Results Approximately 85% of the radiographs were correctly matched by the automated personal identification process. In a limited number of cases, the image recognition algorithm identified 2 potential matches for the same individual. No statistically significant differences were found between measurements performed by the expert on panoramic radiographs from the same patients. Conclusion Personal identification might be performed with the aid of image recognition algorithms and machine learning techniques. This approach will likely facilitate the complex task of personal identification by performing an initial screening of radiographs and matching ante-mortem and post-mortem images from the same individuals.

Automatic Detection of Teeth and Dental Treatment Patterns on Dental Panoramic Radiographs Using Deep Neural Networks

“…The mainstream themes of existing studies had been to find possible candidates based on the similarity of the image itself or to measure similarities based on various identifiers by extracting certain features or patterns from DPRs [ 13 , 27–30 ]. The latest previous studies using deep learning demonstrated an average accuracy of about 60%–80% [ 12 , 14 , 27 , 30 ]; however, this study showed a higher accuracy. Extracting features or performing comparisons on the images could be intuitive and fast, but it might also be affected by several endogenous variables such as the distortion, contrast, and resolution of DPRs.…”

Forensic Dentistry as an Analysis Tool for Sex Estimation: A Review of Current Techniques

“…Ortiz et al explored the usefulness of machine learning and automation by matching pairs of panoramic radiographs for personal identification and found an approximate accuracy rate of 85% by the automated process. 47 Esmaeilyfard et al used measurements made from cone beam computed tomography (CBCT) images of the first mandibular molar and evaluated them for sexual differences and found that the predictive model was able to predict sex from linear measurements of the first mandibular molar from CBCT images to over 90% accuracy. 48 The increasing use of imaging and radiography around the world could shift how we practice forensic dentistry, by incorporating more elements of technology and automated machine learning to assist us with disaster victim identification or age and sex estimation.…”

3D superimposition of human dentition contours in personal identification: A preliminary study

“…A variety of algorithms have been created for artificial intelligence (AI) recognition based on panoramic dental radiographs, which have been widely utilized in dental practice and could offer a metrical assessment of dental anatomy [16]. Heinrich [17] and Ortiz [18] simulated AM and PM panoramic radiographs of victims, and 85% of the cases could be reliably TA B L E 1 The RMSE value of genuine pair and the minimum RMSE value of imposter pair, N representing the number of registrations. identified.…”

“…A variety of algorithms have been created for artificial intelligence (AI) recognition based on panoramic dental radiographs, which have been widely utilized in dental practice and could offer a metrical assessment of dental anatomy [16]. Heinrich [17] and Ortiz [18] simulated AM and PM panoramic radiographs of victims, and 85% of the cases could be reliably identified. Matsuda [19] validated the accuracy of personal identification with paired orthopantomographs using six well‐known convolutional neural network (CNN) architectures.…”