A biopsy-validated study of combined DL-upscaled MRSI and (18F)-FACBC PET for presurgical glioma diagnosis.
Authors
Affiliations (8)
Affiliations (8)
- Department of Diagnostic Imaging, Akershus University Hospital, Lørenskog, Norway.
- Institute of Clinical Medicine, University of Oslo, Oslo, Norway.
- Department of Circulation and Medical Imaging, Norwegian University of Science and Technology (NTNU), Trondheim, Norway.
- Department of Radiology and Nuclear Medicine, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway.
- Department of Neurosurgery, Ophthalmology and Otorhinolaryngology, University Hospital of North Norway, Tromsø, Norway.
- Department of Neurosurgery, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway.
- Department of Neuroscience, Norwegian University of Science and Technology, Trondheim, Norway.
- Department of Electrical Engineering and Computer Science, University of Stavanger, Stavanger, Norway.
Abstract
Medical imaging is crucial for glioma management. Combined with MRI, amino acid PET may improve glioma diagnosis, biopsy targeting, and tumor delineation compared to structural MRI alone. Magnetic resonance spectroscopic imaging (MRSI) complements both structural MRI and PET by detecting metabolites such as N-acetylaspartate (NAA), creatine (Cr), and choline (Cho), which are markers for brain health and tumor malignancy, but is challenged by low spatial resolution. This study evaluates whether upscaled MRSI enhanced by deep learning can improve diagnostic accuracy and serve as a complement or alternative to PET for glioma classification. Ten glioma patients (ages 24-80) were included. Presurgical [18F]-FACBC PET/MRI, including proton 2D MRSI, was acquired for all patients. Thirty-two image-guided biopsies were sampled during surgery from the patients and classified as glioma tissue or non-tumor tissue, and according to <i>IDH1</i> status. For each biopsy location, tumor-to-background ratio (TBR) and standardized uptake value (SUV) from PET, and tCho/NAA and tCho/tCr ratios from MRSI were calculated. ROC analysis was used to assess the accuracy of [18F]-FACBC PET and upscaled MRSI, and the combinations of these in classifying glioma vs non-tumor tissue and <i>IDH1</i> status. Among the individual biomarkers evaluated, the tCho/NAA ratio from the deep learning-based model demonstrated the highest apparent AUC for classifying glioma vs. non-tumor tissue (AUC = 0.87, 95% CI: 0.66-1.0), compared with tCho/tCr (AUC = 0.81, 95% CI: 0.54-1:00), SUV (AUC = 0.71, 95% CI: 0.49-0.90), and TBR (AUC = 0.68, 95% CI: 0.48-0.86). Combining TBR with tCho/NAA and/or tCho/tCr improved tissue classification compared to either modality alone, where TBR+tCho/NAA+tCho/tCr showed the best results (AUC = 0.91, 95% CI: 0.71-1.0). MRSI was a poor predictor for <i>IDH1</i>-status (tCho/NAA: AUC = 0.67, 95% CI: 0.44-0.88 and tCho/tCr: AUC = 0.38, 95% CI: 0.17-0.60), while PET was an excellent predictor (SUV: AUC = 0.83, 95% CI: 0.66-0.95 and TBR: AUC = 0.82, 95% CI: 0.65-0.94) and the combination of SUV, tCho/NAA, and tCho/tCr was a best predictor (AUC = 0.96, 95% CI: 0.88-1.0). Combining DL-upscaled MRSI with [18F]-FACBC PET appeared to enhance discrimination between glioma and non-tumor tissue, a finding that warrants validation in larger, prospective studies.