Thesis Project Form

In pediatric patients, diffuse intrinsic pontine glioma (DIPG) represents the most difficult brain tumor to treat and the leading cause of brain tumor-related death.

Despite recent remarkable genomic discoveries, radiotherapy (RT) remains the current standard of care, providing transient clinical improvement and a limited survival benefit.
Positron Emission Tomography (PET) imaging with amino-acid tracers, such as 18F-dihydroxyphenylalanine (DOPA), is a promising diagnostic tool able to provide non-invasive information of brain tumors, including prediction of disease evolution.
A pioneering study conducted in collaboration between Giannina Gaslini Children’s Institute and Ospedale Galliera has evaluated diagnostic and prognostic information obtained by 18F-DOPA PET in a group of children with newly diagnosed DIPGs, revealing the potential of this imaging modality in providing crucial information for evaluating the metabolism of DIPGs.

Aforementioned work showed good correlation with 18F-DOPA PET uptake intensity and extent (PET tumor volume and uniformity) with clinical scores for treatment response and survival of pediatric patients with DIPGs.
However, in the present study and, more generally, in clinical routine, quantitative markers extracted from PET scans are measured manually by clinicians with limited computer support. This process is time consuming and potentially subjected to inaccuracy and individual variability in measurements.
The proposed work has the aim to develop an image processing pipeline to automatize computation of clinical indexes and thus to provide radiologists with a user-friendly computer-aid tool specifically conceived for pediatric PET images.
Final designed framework will be tested on the same cohort previously analyzed with manual, semi-quantitative methods in order to compare performance of proposed tool and of manual measurements.
Future developments may include extension of this pipeline to other kinds of pediatric brain tumors.
Long-term relapses of this tool may contribute to the establishment of well-timed, personalized and more effective therapies, aiming to improve overall survival and quality of life of DIPG patients.

Image processing and statistical analysis of clinical data of pediatric patients affected by different degrees of DIPGs. Extraction of quantitative markers from PET scans implies development of ad-hoc image processing pipelines (e.g. in Matlab or Python) specifically adapted to the age range under analysis.