A Deep Learning Model for Preoperative Differentiation of Glioblastoma, Brain Metastasis and Primary Central Nervous System Lymphoma: A Pilot Study

Frontiers in Oncology

Marco Locatelli¹  Martina Giordano²  Paolo Palmisciano³  Giorgio Fiore⁴  Luigi Schisano⁴  Leonardo Tariciotti⁴  Valerio M. Caccavella⁵  Fabio Triulzi⁶  Giorgio Conte⁶  Giulia Remoli⁷  Luigi Gianmaria Remore⁹  Mauro Pluderi⁹  Manuela Caroli⁹  Stefano Borsa⁹  Giulio Bertani⁹  Giorgio Carrabba⁹ 

[1] Aldo Ravelli” Research Center for Neurotechnology and Experimental Brain Therapeutics, University of Milan, Milan, Italy;Department of Neurosurgery, Fondazione IRCCS Istituto Neurologico Carlo Besta, Milan, Italy;Department of Neurosurgery, Trauma Center, Gamma Knife Center, Cannizzaro Hospital, Catania, Italy;Department of Oncology and Hemato-Oncology, University of Milan, Milan, Italy;Department of Paediatric Orthopaedics and Traumatology, ASST Centro Specialistico Ortopedico Traumatologico Gaetano Pini-CTO, Milan, Italy;Department of Pathophysiology and Transplantation, University of Milan, Milan, Italy;National Center for Disease Prevention and Health Promotion, Italian National Institute of Health, Rome, Italy;Unit of Neuroradiology, Fondazione IRCCS Cà Granda Ospedale Maggiore Policlinico, Milan, Italy;Unit of Neurosurgery, Fondazione IRCCS Cà Granda Ospedale Maggiore Policlinico, Milan, Italy;

关键词: brain metastases;    deep learning;    glioblastoma;    machine learning;    primary central nervous system lymphoma (PCNSL);    artificial intelligence;   

DOI  :  10.3389/fonc.2022.816638

来源: DOAJ

【 摘 要 】

BackgroundNeuroimaging differentiation of glioblastoma, primary central nervous system lymphoma (PCNSL) and solitary brain metastasis (BM) remains challenging in specific cases showing similar appearances or atypical features. Overall, advanced MRI protocols have high diagnostic reliability, but their limited worldwide availability, coupled with the overlapping of specific neuroimaging features among tumor subgroups, represent significant drawbacks and entail disparities in the planning and management of these oncological patients.ObjectiveTo evaluate the classification performance metrics of a deep learning algorithm trained on T1-weighted gadolinium-enhanced (T1Gd) MRI scans of glioblastomas, atypical PCNSLs and BMs.Materials and MethodsWe enrolled 121 patients (glioblastoma: n=47; PCNSL: n=37; BM: n=37) who had undergone preoperative T1Gd-MRI and histopathological confirmation. Each lesion was segmented, and all ROIs were exported in a DICOM dataset. The patient cohort was then split in a training and hold-out test sets following a 70/30 ratio. A Resnet101 model, a deep neural network (DNN), was trained on the training set and validated on the hold-out test set to differentiate glioblastomas, PCNSLs and BMs on T1Gd-MRI scans.ResultsThe DNN achieved optimal classification performance in distinguishing PCNSLs (AUC: 0.98; 95%CI: 0.95 - 1.00) and glioblastomas (AUC: 0.90; 95%CI: 0.81 - 0.97) and moderate ability in differentiating BMs (AUC: 0.81; 95%CI: 0.70 - 0.95). This performance may allow clinicians to correctly identify patients eligible for lesion biopsy or surgical resection.ConclusionWe trained and internally validated a deep learning model able to reliably differentiate ambiguous cases of PCNSLs, glioblastoma and BMs by means of T1Gd-MRI. The proposed predictive model may provide a low-cost, easily-accessible and high-speed decision-making support for eligibility to diagnostic brain biopsy or maximal tumor resection in atypical cases.

【 授权许可】

Unknown