الفهرس 
Logistic regressionOnly 14 pages are availabe for public view
 |  | from | 161 |  |  |
| | | from | 161 | | |
Abstract
Statistical Model To Identify Factors Leading To Autism In Children With Application On Pediatric Hospital Ain Shams University”
1- Introduction:
Autism spectrum disorders (ASDs) represent a group of neurodevelopmental disorders characterized by impairments in verbal and non-verbal communication, social withdrawal and stereotypical behaviors, which may or may not be associated with cognitive deficits, self-injurious behaviors and other neurological comorbidities.
Autism has been said to be the fastest growing developmental disability in the world. A dramatic increase in incidences of ASDs has occurred in the past 25 years. In many countries the prevalence of the current ASD is about 1 in 100, whereas in the 1970s and early 1980s it was about 1 in 2500-3000. This dramatic increase has raised many concerns regarding possible causes, with much attention given to toxic exposure in the perinatal period and broadening of the diagnostic criteria, however the underlying causes of autism are still not fully understood and there is a growing agreement among clinicians and scientists that autism ensue from an interaction between biological vulnerability factors and environmental or iatrogenic insults. One environmental factor that has received significant attention is the body burden of mercury, lead, and other toxic metals.
Studies from the Middle East on this issue have been particularly rare. In a survey on mental health research in the Arab world over 25 years period, publications on child psychiatry on topics such as autism were found to be under presented and most of the presented researches have come from western countries. In fact, autism has become the subject of interest in the developing countries specifically in the Arab world in the late 1990s.
Different factors especially environmental risk factors as exposure to mercury, lead and aluminum, have been studied in this thesis on a sample of Egyptian autistic children compared with their matched controls to reach to the possible risk factors that may lead to autism.
Measurement of heavy metal content in hair is often used as a marker of exposure, because it correlates with the past blood levels, provides a continuous records, and the collection of the samples is essentially non- invasive procedure which is especially useful for testing children.
The parametric statistical techniques represented in Logistic regression and discriminant analysis were applied in this study by using the Statistical Package For Social Sciences (SPSS v.22.0) while the non- parametric statistical techniques represented in Classification and regression tree (CART), Artificial neural networks and Random forest models were applied by using STATISTICA v. 10 software. It was also intended in this study to allow some methodologies used in the modeling process to determine how each method have resulted in simple, parsimonious and meaningful explanation of variables and how they work to get better classification of autism and healthy control status alike.