Abnormalities Detection in Apert Syndrome using Hierarchical Clustering Algorithms

Abstract

Craniosynostosis syndrome is a congenital condition occurring due to the abnormal development of the skull, leading to abnormalities in skull morphology. Apert syndrome is one of common craniosynostosis syndrome in Malaysia and this syndrome can be categorized as the severe craniofacial disorders. The abnormalities of skull morphological in Apert syndrome patient can be identified as outliers which are investigated in this study. This study presents a skull morphological analysis based on a case study involving six paediatric patients diagnosed with Apert syndrome, alongside 22 control patients aged 0 to 12 years, all of whom underwent treatment at the University Malaya Medical Centre (UMMC). The computerized tomography scan (CTSCAN) data is provided by UMMC recorded from year 2012 until 2020 and the data is measured using MIMICS software by taking the measurement of cranial angles. The clustering-based procedures will be applied to identify the abnormalities in skull angle dataset. There are 12 skull angles and these angles are analysed using hierarchical clustering algorithms for identifying the outliers or abnormalities. The objective is to detect the abnormalities and determine the skull angles that associated with Apert syndrome (AS) in Malaysia population using clustering-based procedure. The abnormalities in Apert syndrome datasets are successfully detected by the algorithms and this study found that there are skull angles with specific location of angles are associated with Apert syndrome. This study also found that the location of skull angles for patients age 0-24 months old and >24 months old are different. The findings of this study can assist the surgical team in directing additional focus towards specific regions of the skull during the planning of interventions. This guidance has the potential to optimize surgical outcomes and reduce the risk of potential complications.