A Study on Sex Estimation by Using Machine Learning Algorithms with Parameters Obtained from Computerized Tomography Images of the Cranium

Abstract

The aim of this study is to test whether sex prediction can be made by using machine learning algorithms (ML) with parameters taken from computerized tomography (CT) images of cranium and mandible skeleton which are known to be dimorphic. CT images of the cranium skeletons of 150 men and 150 women were included in the study. 25 parameters determined were tested with different ML algorithms. Accuracy (Acc), Specificity (Spe), Sensitivity (Sen), F1 score (F1), Matthews correlation coefficient (Mcc) values were included as performance criteria and Minitab 17 package program was used in descriptive statistical analyses. p ≤ 0.05 value was considered as statistically significant. In ML algorithms, the highest prediction was found with 0.90 Acc, 0.80 Mcc, 0.90 Spe, 0.90 Sen, 0.90 F1 values as a result of LR algorithms. As a result of confusion matrix, it was found that 27 of 30 males and 27 of 30 females were predicted correctly. Acc ratios of other MLs were found to be between 0.81 and 0.88. It has been concluded that the LR algorithm to be applied to the parameters obtained from CT images of the cranium skeleton will predict sex with high accuracy.

Dysregulated mammalian estrus cycle rescued by timed activation of VIP neurons in the circadian pacemaker and late afternoon light exposure

“…In addition, we identified a difference in SCN VIP activity at the peak time of ZT10 during receptive days. While ML has been used to identify sex differences in biological samples or behaviors ( 44-47 ) and estrous cycle stages from vaginal smears ( 48-50 ), this is, to our knowledge, the first demonstration of its use to identify estrous cycle states based on neuronal activity. Hypothesizing that SCN VIP activity at ZT10 acts as a permissive signal to convey the best timing to ensure fertility, we found that SCN VIP excitation either through the retinohypothalamic tract via light or via chemogenetically-specific activation at this unique feedback timepoint of ZT10 rescued estrous cycle regularity.…”

Determination of Gender By Machine Learning Algorithms, Through Using Craniocervical Junction Parameters and Dimensions of the Cervical Spinal Canal

“…There has been ample research in the literature on the gender determination, using very commonly the cranium, pelvis, and long bones with high accuracy ratio (2)(3)(4)(5)14). A gender study done on the cranial skeleton, using MLA from CT images, has determined the Acc ratios as 88% by LDA and 83% by QDA models (6). Likewise, another research has used pelvis CT images, finding the Acc ratios as 91-93% by DT model (15).…”

“…Even though DNA analyses seem to be the most reliable contemporary method -indeed it issince it requires professionals, advance laboratories, and high costs, new suitable methods are in need. Bones such as cranium, pelvis, and long bones have amply been used in gender determination since they display the highest dimorphism (2)(3)(4)(5)(6). Discriminant analysis has also been used previously for the gender determination (7).…”

Evaluation of Skull Morphometry in Computed Tomography Images and Calculation of the Cephalic Index

“…Europeans have macrocephalic skull volume (4). Accurate gender estimation can be achieved over 90% by logistic regression using skull metric differences (5). Skulls do not show sex discrimination in children until puberty.…”