Development of Prognostic and Predictive Models Using Multidimensional Signatures for Glioblastoma

Date: 2022/04/22 - 2022/04/22

Academic Seminar: Development of Prognostic and Predictive Models Using Multidimensional Signatures for Glioblastoma

Speaker: Prof. Ning Wen, Medical Imaging Technology Institute at Ruijin Hospital, Shanghai Jiaotong University School of Medicine

Time: 9:00 a.m.-10:00 a.m., April 22nd, 2022 ( Beijing Time)

Location: via feishu

Abstract

Glioblastoma Multiforme (GBM), a World Health Organization (WHO) grade IV tumor, is the most common primary malignant brain cancer with median overall survival (OS) of 15 months despite best available treatments. Although one- and three-year OS rates are 39.7% and 10.1%, respectively, it is of note that some patients do significantly better than others. The classification of Glioma based on genomic features is integral to the WHO classification and has shown strong prognostic value independent of age and histology. Despite molecular advancements, the recurrent and molecularly heterogeneous nature of GBM garners its resilience. The tumor cells and tumor microenvironment rapidly adapt to treatment and there is not an available method for continuous monitoring of such adaptive changes at a molecular level. Clinical care for patients with GBM is primed for advancement through the use of neural networks that augment imaging and genomic analysis and predict progression-free (PFS) and OS. Such applications can also be applied to therapeutic navigation and discovery.

Integrating molecular biomarkers with multiparametric MRI (mpMRI) features in a multiplex setting has an enormous potential to identify GBM biomarkers crucial for: 1) improving patient risk stratification, 2) developing precision treatment and 3) predicting outcome of salvage treatment. We aim to improve GBM treatment effectiveness by 1) developing machine learning models to segment intratumor sub-regions on mpMRI and identifying prognostic imaging features prior to surgical intervention, 2) developing pathway associated neural network models for prognosis by integrating gene expression patterns, oncogenic pathways and imaging features.

Biography

Dr. Ning Wen received his Ph.D. degree in Medical Physics from Wayne State University and obtained his MBA degree from the University of Michigan. Before joining Medical Imaging Technology Institute at Ruijin Hospital, Shanghai Jiaotong University School of Medicine, Dr. Wen worked at Henry Ford Health System and served multiple roles, including Director of Clinical Physics and Director of Translational Research. He also had an adjunct professor position at the Department of Oncology, Wayne State University.

Dr. Wen has extensive experience and interests in stereotactic radiosurgery and is Co-director of a stereotactic radiosurgery course, which has trained over 300 doctors, physicists, and therapists from across the globe. Dr. Wen’s current research focuses on developing supervised and unsupervised machine learning algorithms to analyze osmics data derived from medical images and cancer genome for patient risk stratification and treatment response evaluation. Dr. Wen is a recipient of multiple grant awards from the American Cancer Society, National Cancer Institute, and other agencies. He has published over 70 peer-reviewed articles, and 150 meeting abstracts. He was elected as a Fellow of the American Association of Physicists in Medicine in 2021.