Automated human age at death estimation system from long bones histology

Human age estimation at death from bone histology is a frequent and important requirement in forensic anthropology. Usually human age at death is estimated manually from bone histology or morphology. Manual methods of age estimation from bone histology involve three main phases that includes, analysis of variations in microscopic characteristics of bone with age, developing age regression equation based on the variation analysis and estimation of age using regression equation. However manual age at death estimation is not only tedious and time consuming process but also prone to observation variability and produce subjective results. Furthermore, there exists no digital database that can store the information of bone samples of Malaysian population. Hence it is vital to develop a histological automated system for age at death estimation to eliminate the problems of manual methods. This study presents the development of automated system for human age at death estimation from bone histology. Six histological and two morphological parameters were analyzed in 44 samples of long bones (humerus, radius, ulna, tibia, fibula and femur). First, the measurements and analyses were carried out using manual methods and then an automated system was developed to eliminate the problems of the manual process. The system assists in automatic measurements and calculations of bone histological parameters, analysis of parameters with age, developing regression equation and estimation of age. The automatic system also provides a digital database capable of storing the information of all parameters. The results of the system shows that histological parameters specifically percentage area covered by Haversian canals and mean Haversian canal area possess the highest correlation with age. Morphological parameters do not show significant correlation with age in Malaysian population. Age regression equation is developed with SEE of 8.3 years. The automatic system estimates age within 10 years of the actual ages for 89% of the samples. The automatic system is evaluated by seven forensic anthropologists and is considered effortless and acceptable for automatic age at death estimation from bone histology.

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral

Grantor dc:publisher.institution

Universiti Tun Hussein Onn Malaysia

Year dc:date.issued

2019

Author dc:creator

• Khan, Ijaz

• QA Mathematics

Language dc:language

en