Date of Award
5-1-2022
Thesis Type
phd
Document Type
Thesis (Restricted Access)
Divisions
science
Department
Institute of Biological Sciences
Institution
Universiti Malaya
Abstract
Machine learning methods have been used in this study to analyze and predict the required healing time among paediatric orthopaedic patients. To our best knowledge, there is no study reported using machine learning methods to predict paediatric orthopaedic fracture healing time. In this study, we examined the fracture healing time in children using Random forest (RF), Self-Organizing Feature map (SOM) and support vector regression (SVR) The study sample was obtained from the paediatric orthopaedic unit at University Malaya Medical Centre, radiographs of the upper limb and lower limb fractures from children under twelve years, with ages recorded from the date and time of initial injury. Inputs assessment extracted from radiographic images included the following features: type of fracture, angulation of the fracture, the contact area percentage of the fracture, age, gender, bone type, type of fracture, and the number of bones involved. all of which were determined from the radiographic images. RF and SVR were used to select variables affecting bone healing time. Then, SOM was applied for analysis of the relationship between the selected variables with fracture healing time. Findings from this study identified fracture angulation and distance, age and bone part as important variables in explaining the fracture healing pattern. Root mean square error (RMSE) was used as a performance measure and SOM was used in this study for visualization and ordination of factors associated with healing time. Based on the outcomes obtained from the models it is concluded that SVR and SOM techniques can be used to assist in the analysis of the healing time efficiently especially in paediatric cases as it can additionally signal a non-unintentional injury or abnormal restoration, that affect the time required for bone fracture healing. Predicting healing time can be used as a tool in the treatment process for general practitioners and medical officers and in the follow-up period. We also have developed decision support using the AO trauma guide to determine the type of fracture and its management. The system prototype is available at kidsfractureexpert.com/.
Note
Thesis (PhD) - Faculty of Science, Universiti Malaya, 2022.
Recommended Citation
Lau, Chia Fong, "Paediatric orthopaedic fracture healing prediction system / Lau Chia Fong" (2022). Student Works (2020-2029). 1175.
https://knova.um.edu.my/student_works_2020s/1175