Magnetic Resonance Imaging Of Calvarial Diploe Thickness for Sex And Age Estimation of A Sample Of Egyptian Population

Document Type : Original Article

Authors

1 Department of Forensic Medicine and Clinical Toxicology, Faculty of Medicine, Suez Canal University, Ismailia, Egypt

2 Department of Radiodiagnosis, Faculty of Medicine, Suez Canal University, Ismailia, Egypt

Abstract

Background: Magnetic resonance imaging (MRI) is a mainstay radiological technique that is becoming more desirable in forensic identification for its non-invasive, non-ionizing and highly precise nature. Objectives: This study aimed to investigate whether sex and age could be estimated based on Calvarial Diploe Thickness (CDT) using MRI in a sample of the Egyptian population. Methods: The study was conducted on cranial MRI images of 306 adult Egyptians. For sex and age estimation, seven CDT measurements were assessed on 256 cranial MRI images of known sex and age. A binary logistic regression equation was developed for sex prediction. Multivariate linear regression analyses were performed to generate age prediction equations; equation (1), gender was included, while equation (2) was irrespective of gender. All regression equations were tested on 50 cranial MRI images. Results: The results revealed a significant binary logistical regression model for sex prediction of r2 0.442 and classification accuracy 71.6%. The Receiver Operating Characteristic (ROC) curve showed an area under the curve (AUC) 0.816, sensitivity 76.6 %, and specificity of 70.5%. Testing the equation showed accuracy for identifying males (78.2 %), accuracy for identifying females (70.4 %), and overall accuracy (74 %). Regarding age estimation, Pearson correlation showed no statistically significant correlation between age and all CDT measurements. Age prediction multivariate linear regression models (sex-stratified/sex-pooled) were statistically insignificant, with very low adjusted r2 and high standard error of estimate. Testing the multivariate linear regression equations showed Mean Absolute deviation (MAD) between the chronological age and predicted age of 9.09254 years and Mean Absolute Percentage Error (MAPE) of 28.458% for the multivariate linear regression equation (1) and MAD of 8.94469 years; and MAPE of 27.9829% for the multivariate linear regression equation (2). Conclusion: We concluded that the studied CDT measurements could be included as a new candidate in MRI-based sex estimation for Egyptians while are unfavorable for estimating age. Our results provide preliminary results for assessing a new approach for estimating sex and age for Egyptians. Applying such parameters on a greater sector of Egyptians for both sex and age estimation is highly recommended.

Keywords

Files

Share

How to cite

Statistics