Predictive value of 0.35 T magnetic resonance imaging radiomic features in stereotactic ablative body radiotherapy of pancreatic cancer: A pilot study

• DOI: 10.1002/mp.14200
• Publication Date: 2020-04-24T13:51:56Z
• Authors: Garrett Simpson; Benjamin Spieler; Nesrin Dogan; Lorraine Portelance; Eric A. Mellon; Deukwoo Kwon; John C. Ford; Fei Yang
• Keywords: Pancreatic Neoplasms; 03 medical and health sciences; 0302 clinical medicine; Humans; Pilot Projects; Radiosurgery; Magnetic Resonance Imaging; Neoadjuvant Therapy; 3. Good health

Purpose The aim of this study was to evaluate the potential and feasibility radiomic features extracted from low field strength (0.35 T) magnetic resonance images (MRIs) in predicting treatment response for patients with pancreatic cancer undergoing stereotactic body radiotherapy (SBRT). Methods Twenty unresected, non‐metastatic ductal adenocarcinoma (PDAC) were enrolled, all whom received neoadjuvant chemotherapy followed by five‐fraction MR‐guided SBRT a radiation dose range 33−50 Gy. For each patient, five daily setup scans acquired hybrid 0.35 T MRI/radiotherapy unit. Tumor heterogeneity quantified gross tumor volume (GTV) averaged over course treatment. Random forest (RF) adaptive least absolute shrinkage selection operator (LASSO) classification models constructed identify radiomics predictive response. Predictive capability top‐performing then evaluated using receiver operating characteristic area under curve (AUC) obtained leave‐one‐out cross‐validation. Results Half 20 showed treatment, defined regression on histopathology or follow‐up dynamic contrast‐enhanced computed tomography (CT). most selected RF method GLCM energy GLSZM gray‐level variance. RF‐based model achieved an AUC = 0.81 95% confidence interval [0.594 1] LASSO algorithm as only feature, achieving [0.596 1]. Conclusion findings suggest that during may contain information about PDAC Using supports continued expansion analysis based MR hold providing timely indications