To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.

 
 

Description

Convolutional Neural Network based Machine Learning for Ameloglyphics: A Forensic Analysis

Sanjana Shetty, Sowmya SV , Dominic Augustine, Saiprasad Alva, Mukul Saini Author Affiliations: Department of Oral & Maxillofacial Pathology and Oral Microbiology, Faculty of Dental Sciences, MS Ramaiah University of Applied Sciences, MSR Nagar, Bengaluru-560054, Karnataka, India.

Purpose:

Tooth prints, considered to be the hard tissue analogues of finger prints have been studied extensively over the years by manual and in some cases, digital methods. While Artificial intelligence and Machine Learning have witnessed a steady rise in their applications in various fields with promising results, its utility in ameloglyphics has not been tried and tested. This study employed Machine learning through Convolutional Neural Network (CNN) to analyse enamel prints. The aim of this study was to analyse tooth prints through Convolutional Neural Network based technology and correlate the patterns with gender and age.

Methods:

The study was done on a sample size of 39 extracted deciduous teeth and 51 extracted permanent teeth. The surface of the teeth were acid etched and the enamel prints were taken by means of cellulose acetate strips. The obtained prints were photographed, subjected to manual analysis and classified into three patterns. CNN was then used for training and testing the data sets.

Results:

CNN was successfully trained and tested for its ability to identify and differentiate ameloglyphic patterns. Significant differences were observed between the enamel prints of the two genders and between the analysed age groups. Ameloglyphic patterns, being unique to individuals, act as aids in personal identification and hold immense value in mass disaster situations where soft tissues being friable are seldom preserved. Enamel being highly resilient and resistant to various degrading actions such as heat and acid, can be a crucial tool for human identification in such circumstances. Artificial Intelligence and Machine Learning based CNN for the analysis of these enamel prints can simplify and potentially replace the conventional methods.

Conclusions:

Ameloglyphics for personal identification is a significant forensic tool. Ameloglyphic analysis via CNN based machine learning was found to be accurate, cost effective and time efficient. The analysis of tooth prints by manual means can be a cumbersome process and the incorporation of AI and ML for the same, as observed in this study, can overcome this drawback. Hence, CNN for ameloglyphic analysis is a reliable tool.

Disciplines

Investigative Techniques | Oral Biology and Oral Pathology | Other Dentistry

Document Type

Event

Share

COinS
 

Convolutional Neural Network based Machine Learning for Ameloglyphics: A Forensic Analysis

Convolutional Neural Network based Machine Learning for Ameloglyphics: A Forensic Analysis

Sanjana Shetty, Sowmya SV , Dominic Augustine, Saiprasad Alva, Mukul Saini Author Affiliations: Department of Oral & Maxillofacial Pathology and Oral Microbiology, Faculty of Dental Sciences, MS Ramaiah University of Applied Sciences, MSR Nagar, Bengaluru-560054, Karnataka, India.

Purpose:

Tooth prints, considered to be the hard tissue analogues of finger prints have been studied extensively over the years by manual and in some cases, digital methods. While Artificial intelligence and Machine Learning have witnessed a steady rise in their applications in various fields with promising results, its utility in ameloglyphics has not been tried and tested. This study employed Machine learning through Convolutional Neural Network (CNN) to analyse enamel prints. The aim of this study was to analyse tooth prints through Convolutional Neural Network based technology and correlate the patterns with gender and age.

Methods:

The study was done on a sample size of 39 extracted deciduous teeth and 51 extracted permanent teeth. The surface of the teeth were acid etched and the enamel prints were taken by means of cellulose acetate strips. The obtained prints were photographed, subjected to manual analysis and classified into three patterns. CNN was then used for training and testing the data sets.

Results:

CNN was successfully trained and tested for its ability to identify and differentiate ameloglyphic patterns. Significant differences were observed between the enamel prints of the two genders and between the analysed age groups. Ameloglyphic patterns, being unique to individuals, act as aids in personal identification and hold immense value in mass disaster situations where soft tissues being friable are seldom preserved. Enamel being highly resilient and resistant to various degrading actions such as heat and acid, can be a crucial tool for human identification in such circumstances. Artificial Intelligence and Machine Learning based CNN for the analysis of these enamel prints can simplify and potentially replace the conventional methods.

Conclusions:

Ameloglyphics for personal identification is a significant forensic tool. Ameloglyphic analysis via CNN based machine learning was found to be accurate, cost effective and time efficient. The analysis of tooth prints by manual means can be a cumbersome process and the incorporation of AI and ML for the same, as observed in this study, can overcome this drawback. Hence, CNN for ameloglyphic analysis is a reliable tool.