Automated platform for histological race and sex comparison of human cortical bone

Research on histological bone variation in population is in its early stages in Malaysia and limited information is available about age graded race and sex comparison. This research performed race and sex comparison of histological cortical bone parameters in the Malaysian population and presented an automated system which could be used as assistance tool by forensic experts. Human bone specimen were collected from Hospital Universiti Kebangsaan Malaysia Medical Centre (UKMMC), Kuala Lumpur, Malaysia. Haversian canals were measured and five parameters were calculated for comparison. Comparison test (t-test/u-test) showed that the size of Haversian canals were significantly greater (p<0.05) in females (HCM fifth, sixth decade: 5955.8 μm2, 5788.0 μm2) than males (HCM fifth, sixth decade: 4117.6 μm2, 3965.1 μm2). In race comparison, total area covered by Haversian canals (bone porosity) was significantly greater (p<0.05) in Indian samples (HCA: 0.457mm2) compared to Chinese samples (HCA: 0.385mm2) in the second decade. However in fifth decade, total area covered by Chinese samples (HCA: 0.894mm2) was significantly greater (p<0.05) than Indian samples (HCA: 0.570mm2). Three main steps of histological comparison were focused for automation i.e. parameter calculation, data management and statistical comparisons. The system was designed with GUI which utilizes aforementioned automation step. Validation of the system was divided into two main parts. In first part, parameter measurement and calculation performed by the system were compared with existing tools in terms of percentage error in measurement (DinoCapture: 5.3%, L-measure: 5.1%, ImageJ: 4.7%, designed system: 4.0%) and consumed time for measurement (DinoCapture: 15-20min, L-measure: 15-20min, ImageJ: 20-25min, designed system: 1-2min). Similarly automated race and sex comparison performed by the system were compared with comparisons performed manually using SPSS software. Significance and t/z values showed no differences and did not change overall hypothesis of the comparison tests. Which implies that the automated system is efficient for histological race and sex comparisons.

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

• Abdullah, Hadi

• QA76 Computer software

Language dc:language

en