Development and Clinical Translation of Advanced PET Imaging for Real-Time Biology-Guided Radiotherapy

About This Grant

Abstract Radiation enable therapy (RT) i s one of the most common treatments for solid cancer patients. Modern RT techniques highly conformal dose delivery to target volumes.Positron Emission Tomography (PET) is emerging as a valuable tool in enhancing the precision and accuracy of radiation therapy. Unlike conventional imaging modalities such as CT or MRI, which rely on anatomical details, PET allows for the visualization of physiological function, which is particularly useful in oncology. PET-guided radiation therapy is a new technology that allows for more precise tumor delineation, ensuring that high-dose radiation is delivered to the areas of highest need while minimizing exposure to surrounding healthy tissues. By incorporating PET imaging into the planning procedures clinicians can include regions that might otherwise have been underdosed, reduce dose in areas unlikely to contain cancer cells, and intensify dose in regions showing high uptake. RefleXion's Biology-guided radiotherapy (BgRT), containing a 6 megavoltage (MV) photon radiotherapy device, a PET imaging system with two 90° C-shaped arcs of detectors, a kilovoltage (kV) X-ray CT imaging, and treatment planning subsystems, is a novel and the only radiotherapy delivery mechanism that is capable of aiming beamlets of external radiotherapy at PET emissions that originate from the target in real-time. Because the radiotherapy beamlets are guided to the tumor in real-time, this technology holds great promise for more effective treatment delivery, as well as reducing margins around the target and thereby reducing normal tissue toxicities. To allow real-time RT and PET imaging, the PET system of RefleXion BgRT is formed by two 90- degree arc detectors instead of a full ring system as in typical diagnostic PET scanners. In addition, the axial coverage of the PET system of RefleXion BgRT is 5.2 cm as compared to 20-35 cm in typical diagnostic PET. Therefore, the noise level of the RefleXion BgRT PET images is substantially higher due to lower sensitivity, subsequently affecting the image quality, quantification accuracy, motion estimation, and eventually treatment planning and delivery. In addition, the PET images of functional modeling are time-averaged, resulting in both motion blurring and mis-matched CT-based attenuation correction. In this Academic-Industrial Partnership (AIP) proposal, we formed a strong and ideal partnership between Yale University and RefleXion Medical, to develop, validate, and translate advanced PET denoising and motion management technologies to improve the quality of the BgRT delivered dose. In Aim 1, we will develop deep learning-based PET denoising algorithms for Functional Modeling (FM). In Aim 2, we will develop advanced algorithms for motion management and correction. In Aim 3, we will perform clinical evaluation of advanced PET imaging techniques on adaptive strategies for BgRT. In Aim 4, RefleXion will take the lead on clinical translation of the proposed technologies to end-users.