AUTHOR=Lu Guang , Zhang Yuxin , Wang Wenjia , Miao Lixin , Mou Weiwei 

TITLE=Machine Learning and Deep Learning CT-Based Models for Predicting the Primary Central Nervous System Lymphoma and Glioma Types: A Multicenter Retrospective Study

JOURNAL=Frontiers in Neurology

VOLUME=Volume 13 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2022.905227

DOI=10.3389/fneur.2022.905227

ISSN=1664-2295

ABSTRACT=Purpose and Background. To distinguish primary central nervous system lymphoma (PCNSL) and glioma on Computed Tomography (CT) is an important task since treatment options differs vastly from the two diseases. This study aims to explore various machine learning and deep learning methods based on radiomics features extracted from CT scans to predict PCNSL and glioma types and compare the performance of different models.  
Methods A total of 101 patients from five Chinese medical centers with pathologically confirmed PCNSL and glioma were analyzed retrospectively, including 50 PCNSL and 51 glioma. After manual segmentation of the region of interest (ROI) on CT scans, 293 radiomics features of each patient were extracted. The radiomics features were used as input and then we established six machine learning, one deep learning model and three readers to identify the two types of tumors. The area under the receiver operating characteristic curve (AUC) and accuracy (ACC) were used to evaluate different models.
Results The cohort was split into a training (70, 70% patients) and validation cohort (31,30% patients) according to stratified sampling strategy. The deep neural network model performed best, with an accuracy of 0.886, 0.903, sensitivity of 0.914, 0.867, specificity of 0.857, 0.937, and AUC of 0.962, 0.913 in the training and validation cohort, respectively, which was significantly higher than the three primary physician’s diagnoses (ACCs ranged from 0.684 to 0.719, ρ<0.001 for all). Among all the machine learning models, the AUC ranged from 0.605 to 0.821 in the validation cohort.
Conclusion The established PCNSL and glioma prediction model based on deep neural networks method from CT radiomics features is feasible and provided high performance, which shows the potential to assist clinical decision-making.