Master of Science
Epilepsy is a degenerative brain disease characterized by abruption of neural activities that result in seizures. The onset of epileptic seizures are usually from a primary source - the epileptogenic foci (EF) which could be distributed to nearby neurons and tissues. Accurate localization of EF is critical in epilepsy cases where drug treatment has failed, and surgery is indicated to resect the EF to alleviate seizure. Typically, hybrid positron emission tomography (PET) and computed tomography (CT) imaging are performed to functionally localize the EF in drug-resistant epilepsy for surgical planning when anatomical abnormalities representing the EF cannot be identified on magnetic resonance imaging (MRI). The recent introduction of integrated PET/MRI scanning has significantly enhanced the localization of EF and eliminated the use of CT for PET attenuation correction (AC), minimizing radiation exposure particularly in radiosensitive patients such as pediatrics. The objective of this thesis is to develop an image analysis approach to further reduce PET radiation dose for optimal pediatric epilepsy PET/MRI. First, to eliminate CT for PET AC, a robust deep learning approach validated in pediatric populations for synthesizing CT from MRI was implemented after performing a rigorous systematic review and meta-analysis of state-of-the art machine learning (ML) AC methods. Next, a deep learning tool using the Self-SiMilARiTy-Aware Generative Adversarial framework (SMART) was developed and evaluated for denoising of PET images acquired with 90% reduction in PET dose, to generate high quality PET images. By combining ML-AC and SMART-PET, this work proposed an approach to drastically reduce the radiation exposure for high quality pediatric epilepsy PET imaging from ~6 mSV in PET/CT to ~0.5 mSV in low-dose PET/MRI.
Summary for Lay Audience
One in three epilepsy patients are considered to be drug-resistant when anti-epileptic medicines fail to alleviate seizure occurrence. Surgical resection is a standard of care treatment in drug-resistant epilepsy to remove the region of the brain triggering epileptic seizures. Accurate detection and characterization of the brain region that trigger seizures could increase the likelihood of seizure cessation and improve the quality of life after surgery. In approximately three out of five drug-resistant epileptic patients, anatomical localization of the brain lesion triggering the seizure cannot be detected on MRI. In these cases, PET/CT is used to measure changes in brain metabolism to detect the brain region with abnormal metabolism, representing the seizure trigger location. However, both PET and CT expose the patient to radiation risks with potential adverse health effects, which may be more severe in children and adolescents (pediatrics). In this thesis, we proposed an artificial intelligence-based image analysis approach for pediatric epilepsy PET/MRI to reduce up to 90% of radiation dose, by 1) eliminating CT, which is used solely for correcting PET signal attenuation and 2) denoising PET images to generate high quality images from very low injected PET dose. To eliminate CT, we implemented a machine-learning approach to generate a synthetic CT (sCT) from MRI input, this sCT were used for PET attenuation correction. To denoise the PET images, we developed a Self-SiMilARiTy-Aware deep learning-based tool (SMART) to obtain high-quality PET images by denoising PET images acquired with a 90% reduction in PET dosage. The proposed elimination of CT and PET image denoising will enable high quality pediatric epilepsy PET imaging with low radiation exposure.
Raymond, Confidence, "An Evaluation of a Deep learning approach for Radiation Dose Reduction in 18F-FDG PET/MRI Pediatric Epilepsy Imaging" (2023). Electronic Thesis and Dissertation Repository. 9111.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License