AI boosted molecular MRI for apoptosis detection in oncolytic virotherapy

Abstract

Oncolytic virotherapy is a promising treatment for high mortality cancers1. Non-invasive imaging of the underlying molecular processes is an essential tool for therapy optimization and assessment of viral spread, innate immunity, and therapeutic response2, 3. However, previous methods for imaging oncolytic viruses did not correlate with late viral activity4 or had poor sensitivity and specificity5. Similarly, methods developed to image treatment response, such as apoptosis, proved to be slow, nonspecific, or require the use of radioactive or metal-based contrast agents6-8. To date, no method has been widely adopted for clinical use. We describe here a new method for fast magnetic resonance molecular imaging with quantitative proton chemical-exchange specificity to monitor oncolytic virotherapy treatment response. A deep neural network enabled the computation of quantitative biomarker maps of protein and lipid/macromolecule concentrations as well as intracellular pH in a glioblastoma multiforme mouse brain tumor model. Early detection of apoptotic response to oncolytic virotherapy, characterized by decreased cytosolic pH and protein synthesis, was observed in agreement with histology. Clinical translation was demonstrated in a normal human subject, yielding molecular parameters in good agreement with literature values9. The developed method is directly applicable to a wide range of pathologies, including stroke10, cancer11-13, and neurological disorders14, 15.