Abstract:
The most common thoracic cage abnormality is pectus excavatum (PE), which is assessed using three CT indices: Haller, correction, and asymmetry. Manually performing this analysis is time-consuming and variable. This paper proposes a fully automatic framework for PE severity quantification from CT images, consisting of three steps: identification of the sternum’s greatest depression point, detection of 8 anatomical keypoints relevant for severity assessment, and geometric regularization and extraction of measurements. The first two steps use Unet++ heatmap regression networks, including a novel variant to predict 1D confidence maps. The framework was tested on a 269-CT database. In a subset of patients, intra-observer, inter-observer, and intra-patient variability of the estimated indices were compared. The developed system had a mean relative absolute error of 4.41%, 5.22%, and 1.86% for the Haller, correction, and asymmetry indices, respectively. The proposed framework outperformed the expert in intra-patient analysis, showing higher reproducibility between indices extracted from CTs of the same patient. These findings support the clinical framework for automatic, accurate, and reproducible PE severity quantification.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
