- 98 views
- 162 downloads
Development of a Detection Model for Curve Progression in Adolescents with Idiopathic Scoliosis
-
- Author / Creator
- Khodaei Jalalabadi, Mahdieh
-
Adolescent idiopathic scoliosis (AIS) is a three-dimensional (3D) structural spinal disorder recognized by lateral curvatures. The routine for monitoring of AIS is taking an X-ray every six months to check the curve severity and curve progression. Taking repetitive radiographs is not desirable for children since it can increase the risk of cancer. Based on the literature, demographic and radiographic parameters may help predict curve progression. However, there is no accurate prediction model available yet. The objectives of this thesis were to develop and validate a detection model by combining curve characteristics, reflection coefficient (RC) from ultrasound (US) signals, and clinical information to accurately detect the progressive cases thus avoiding unnecessary radiographs.
The specific objectives were
a) To identify potential clinical predictors based on prediction models from the literature
b) To determine if the US reflection coefficient correlates with curve severity and curve progression
c) To determine whether other US parameters obtained from a single US scan associated with curve progression
d) To develop a predictive model to detect the progressive curves of AIS
e) To validate the model based on clinical data
To identify prognostic factors of curve progression, a systematic review was conducted. The results showed limited or conflicting evidence for most parameters, requiring further investigation. Among the most frequently found potential parameters, age, body mass index (BMI), menarche status, frontal X-ray Cobb, number of curve (NOC), and Risser sign were selected for model development. As new parameters, I suggested the US Cobb change, the US Cobb angle in plane of maximum curvature (PMC), the kyphotic angle (KA), and axial vertebral rotation (AVR) measured on US images. Also, the US RC, which can provide the bone quality information, was investigated.
Experimental and clinical studies were conducted to ensure the validity and reliability of this index. The results showed the larger the RC index value, the stiffer the bone. Also, a clinical study demonstrated that this index could be measured reliably and 68% of children with AIS without progression had a larger RC value than the progression group.
The kyphotic angle was also considered as a potential predictor in the literature. A new method was developed and a pilot study to investigate the reliability of the KA measurement on US images was conducted. The results showed KA could be reliably measured. The maximum difference of the KA measurements between the US images and radiographs was 4°. When a larger clinical study was conducted, it further confirmed the reliability of the KA measurements. The factor which might affect the accuracy of the KA measurements was the posture of the participants during data acquisitions.
Similarly, AVR was also reported as one of the potential predictors and was studied. Before using the AVR as a parameter in model development, the reliability of the AVR measurements was studied. A clinical study was conducted, and the results showed a poor to moderate reliability, ICC≥0.49, with the maximum average difference between X-ray and US being 4.6°. The factors that showed a significant influence on the differences of the measurements were AVR measurements at the apical region and larger AVR severity. Prior to conduct a large clinical study, a pilot study on development of detection model was conducted. In this study age, menarche status, X-ray Cobb angle, RC index, and US Cobb change were used. Of these parameters, only RC and US Cobb change were retained as predictors of curve progression. To develop the final model, a large clinical study including 162 girls was conducted. Among those, 100 participants including 25 progression cases were used for model development. The selected parameters included demographic information: age, BMI, and menarche status; radiographic parameters: X-ray Cobb angle, NOC, and Risser sign; and the US parameters: US Cobb change, RC index, KA, maximum AVR, and PMC Cobb angle. The risk of progression equation was Log (p/1-p) =-1.40+0.28(US Cobb change)-39.45(RC)+1.36 (NOC). The model was then validated on 62 cases with 11 progressions. The sensitivity, specificity, and accuracy results were over 90%.
In conclusion, the developed model showed promising results and the accuracy was better than other studies in the literature. Further study on a larger sample size is needed to offer a stronger validation. -
- Subjects / Keywords
-
- Graduation date
- Spring 2022
-
- Type of Item
- Thesis
-
- Degree
- Doctor of Philosophy
-
- License
- This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.