High-grade brain gliomas, the most common of which is glioblastoma multiforme (GBM), have terrible prognosis and a median patient survival of about 12 months. Although combinations of surgical removal, radiotherapy and chemotherapy are used in the clinical practice, a fundamental and persistent limitation in treating these aggressive tumors is that they tend to infiltrate into normal tissue well beyond margins visible via imaging. Since assessing the spatial extent of tumor infiltration is nearly impossible using current radiologic reading practices, clinical treatment tends to be restricted to parts that are deemed to be clearly malignant, frequently only the enhancing tumor. This failure to aggressively treat the infiltrating tumor accelerates tumor recurrence, and eventually patient death. This proposal aims to develop computational modeling and image analysis methods that will improve our ability to estimate GBM infiltration, as well as to predict tissue that is likelyto present fastest tumor recurrence, thereby eventually opening the way for more aggressive, yet targeted, treatment, such as targeted aggressive surgical removal and/or radiosurgery. To achieve our goal, we will integrate information from several sources: 1) advanced multi-parametric imaging, which captures many aspects of tumor anatomy and physiology~ 2) computational modeling of tumor growth and infiltration~ 3) machine learning methods which, after appropriate training, can learn subtle and potentially complex imaging phenotypes of infiltrating tumors~ 4) statistical atlases, which capture population-based trends that can offer additional insights into tumor growth, such as relationship of infiltration to vasculature and to white matter fiber pathways~ 5) data from one of the largest patient populations having advanced imaging, genotyping, follow-up till tumor recurrence, and histological analysis.
This project will develop advance computational imaging and informatics methods for analysis of high-grade gliomas (brain tumors). It will compile a unique database of data from several hundred patients and will construct predictive models of infiltrating malignant tumor and of later recurrence. Therefore, it will pave the way for more refined and targeted treatments of peritumoral brain tissue, which is where most tumor recurrence occurs.
|Gholami, Amir; Mang, Andreas; Biros, George (2016) An inverse problem formulation for parameter estimation of a reaction-diffusion model of low grade gliomas. J Math Biol 72:409-33|
|Macyszyn, Luke; Akbari, Hamed; Pisapia, Jared M et al. (2016) Imaging patterns predict patient survival and molecular subtype in glioblastoma via machine learning techniques. Neuro Oncol 18:417-25|
|Akbari, Hamed; Macyszyn, Luke; Da, Xiao et al. (2016) Imaging Surrogates of Infiltration Obtained Via Multiparametric Imaging Pattern Analysis Predict Subsequent Location of Recurrence of Glioblastoma. Neurosurgery 78:572-80|
|Bai, Harrison X; Lee, Ashley M; Yang, Li et al. (2016) Imaging genomics in cancer research: limitations and promises. Br J Radiol 89:20151030|
|Bilello, Michel; Akbari, Hamed; Da, Xiao et al. (2016) Population-based MRI atlases of spatial distribution are specific to patient and tumor characteristics in glioblastoma. Neuroimage Clin 12:34-40|
|Mang, Andreas; Biros, George (2015) An Inexact Newton-Krylov Algorithm for Constrained Diffeomorphic Image Registration. SIAM J Imaging Sci 8:1030-1069|
|Gaonkar, Bilwaj; Macyszyn, Luke; Bilello, Michel et al. (2015) Automated tumor volumetry using computer-aided image segmentation. Acad Radiol 22:653-61|
|Akbari, Hamed; Macyszyn, Luke; Da, Xiao et al. (2014) Pattern analysis of dynamic susceptibility contrast-enhanced MR imaging demonstrates peritumoral tissue heterogeneity. Radiology 273:502-10|
|Kwon, Dongjin; Shinohara, Russell T; Akbari, Hamed et al. (2014) Combining generative models for multifocal glioma segmentation and registration. Med Image Comput Comput Assist Interv 17:763-70|
|Da, Xiao; Toledo, Jon B; Zee, Jarcy et al. (2014) Integration and relative value of biomarkers for prediction of MCI to AD progression: spatial patterns of brain atrophy, cognitive scores, APOE genotype and CSF biomarkers. Neuroimage Clin 4:164-73|
Showing the most recent 10 out of 27 publications