Correlation Analysis Between Mandibular And Craniofacial Bone And Cervical Vertebrae In Skeletal Class ? Patients Based On Genetic Algorithms

Objective:To investigate the correlation between the length of mandible and the measurement items of cervical bone and craniofacial bone in the Skeletal Class ? patients in northeast China,and use Genetic Algorithms(GAS)to establish the relevant equations to express the relationship between mandible and cervical bone and craniofacial bone intuitively and quantitatively,and to provide basis for clinical diagnosis and treatment of orthodontic and orthognathic.Method:201 children and adolescents aged 8-20 years were selected,75 males group,mean age(13.54±2.84)years old,and 126 females group,mean age(13.44±2.65)years old.An independent sample t test was performed on the two groups to analyze whether the measured values of the mandible were statistically different.Five samples were randomly selected from the two groups as test samples,and the rest were experimental samples.Dolphin imaging 11.9 software(Partterson,USA)was used to cephalometric analysis to measure mandibular,cranial and facial bone and cervical vertebrae related data.The male group were:male class ? divisionl group(M1 group);male class ? division 2 group(M2 group),female group were:female class ? division 1 group(W1 group);female class ? division 2 group(W2 group),the male and female groups were independently tested by t-test to analyze whether the mandibular measurements of the same sex class ? division 1 groups and class ? division 2 groups were statistically different.Combine the combinations without statistical differences into one group.Using the sensitivity analysis method based on genetic algorithm,the parameters of craniofacial bone and cervical vertebrae with strong sensitivity to mandibular parameters were selected from the measurement results,that is,predictive factor recognition;and the correlation equation was established by genetic algorithm.The predicted values obtained from the correlation equations are compared with the measured values.Results:SPSS 17.0 analysis showed a statistically significant difference in multiple measurements of mandible of different sexes.(p<0.05)The Independent sample T test was carried out in the two groups of M1 and M2,and there was no statistical difference between the two groups of mandibular measurements.(p>0.05)The Independent sample T test was carried out in the two groups of W1 and W2,and there was no statistical difference between the two groups of mandibular measurements.(P>0.05)Through sensitivity analysis,the parameters with the strongest correlation between the measured values of mandible are H4 and SN,and the parameters with strong correlation are SN-Ar,the facial high ratio SGo/NGn,Y axis angle and mandibular angle Ar-Go-Gn.Genetic algorithms have been used to establish the relationship equations between mandible and craniofacial bone and cervical vertebrae of different sexes.The equation is:Male:Ar-Pg=28.415+1.818×H4+0.746×SN(r2=0.0568 p=0.000);Co-Gn=35.523+2.18×H4+0.787×SN(r2=0.556 p=0.000);Co-Go=-71.694+1.456×H4+0.564×SN+0.355×SN-Ar+0.529×SGo/NGn(r2=0.518 p=0.000)?Female:Ar-Pg=15.168+1.706×H4+0.675×SN+0.31×SN-Ar-0.29×Y axis(r2=0.611 p=0.000);Go-Pg=0.852×H4+0.446×SN-0.842×SGo/NGn-0.647×Y axis-0.609x Ar-Go-Gn+203.292(r2=0.734 p=0.000);Go-Me=244.294+1.106×H4-1.029×SGo/NGn-0.725×Y axis-0.638x Ar-Go-Gn(r2=0.700 p=0.000);Go-Gn=246.371+1.241×H4-1.05×SGo/NGn-0.776xY axis-0.570×Ar-Go-Gn(r2=0.664 p=0.000);Ar-Go=-91.453+1.207×H4+0.953×SGo/NGn+0.539×Y axis +0.170×Ar-Go-Gn(r2=0.647 p=0.000);Co-Go=-90.261+1.216×H4+0.412×SN+0.935×SGo/NGn+0.65×Y axis(r2=0.708 p=0.000)There is no statistically significant difference between the predicted values and the measured values of the equations established by GAS(P>0.05).Result:1.There are statistical differences in the lengths of mandible of different genders in skeletal class ? patients.Gender factors should be taken into account in mandibular growth,development and prediction analysis.2.There was no statistical difference in the length of the mandible between the class ? division 1 and the class ? division 2 in skeletal class ?patients of the same sex3.The mandibular length equation based on sensitivity analysis and genetic algorithm is statistically significant.It can directly and quantitatively express the relationship between mandible,craniofacial bone and cervical vertebra.It can provide reference basis for clinical diagnosis and treatment of Orthodontics and Orthognathic,and predict to a certain extent.