Thesis Project Form

Pectus Excavatum (PE) is the most common congenital chest wall deformity in children, characterized by a depression of the anterior chest wall. This condition frequently causes deformation of the thoracic cavity, leading to cardiac compression and, in severe cases, significant cardiopulmonary impairments.
Current clinical assessments for PE rely on indices derived from CT or MRI imaging to evaluate thoracic deformation. However, CT imaging offers superior image quality, enhanced contrast, and higher spatial resolution compared to MRI, making it the preferred modality for detailed anatomical analysis, including both cardiac compression and thoracic depression. These metrics are critical for determining the severity of PE and guiding surgical decisions.
This thesis aims to optimize and extend an existing image processing framework to comprehensively analyze both cardiac compression and thoracic depression using preoperative CT scans. The goal is to improve the accuracy and automation of quantitative evaluations, addressing unresolved clinical challenges and increasing the robustness of the framework.

1. Adaptation of the Existing Framework:
o Develop a comprehensive image processing pipeline for preoperative CT scans capable of quantitatively evaluating both cardiac compression and thoracic depression.
o Optimize the pipeline to fully leverage the superior quality of CT images, ensuring precise segmentation and measurement of anatomical structures.
2. Automation and Efficiency:
o Introduce automation in the image processing workflow to minimize manual intervention, enabling faster and more reliable analysis.
o Extend the framework to address previously challenging cases, such as those involving female patients, whose anatomical variability posed difficulties in earlier methodologies.
3. Clinical Integration:
o Compare and validate the proposed framework against existing indices for thoracic deformation and cardiac compression.
o Assess the correlation between thoracic depression metrics and the degree of cardiac compression, providing a more holistic understanding of PE’s impact on patients.
4. Scalability and Generalization:
o Ensure that the framework is robust across diverse patient demographics, including variations in age, sex, and severity of PE.

1. Develop expertise in advanced CT image processing, including segmentation, feature extraction, and statistical analysis.
2. Design and implement automated workflows for the analysis of thoracic structures and cardiac compression, with a focus on clinical applicability.
3. Learn to validate the framework through clinical data and apply statistical methods to assess its reliability and relevance in surgical planning.

Place(s) where the thesis work will be carried out: DIBRIS (Department of Informatics, Bioengineering, Robotics, and Systems Engineering); Gaslini Pediatric Hospital