Neurosurgery Research Projects
The academic emphasis of NYU Langone’s Department of Neurosurgery is a big draw for many potential PhD, MD, MD/PhD, and postdoctoral trainees. We have maintained continuous funding from the National Institutes of Health (NIH) for more than 30 years. Sought after for peer review, our faculty sit on NIH study sections and serve as editors or referees for numerous academic journals.
Our department receives research grant support from federal and state agencies, as well as private foundations. Over the past decade, our faculty members have garnered 20 research grants from the NIH, as well as awards from the U.S. Department of Veterans Affairs and the New York City Department of Health and Mental Hygiene. Governmental support has been complemented by 14 private grants from organizations, including the Parkinson’s Foundation, the American Association of Neurological Surgeons, the American Brain Tumor Association, the Tuberous Sclerosis Alliance, the Child Neurology Foundation, and the Children’s Brain Tumor Foundation.
Mentoring students, residents, and junior faculty is a priority of our research program. Our goal is to produce independent principal investigators who can ultimately attract their own grant support and trainees. We emphasize individual mentoring, along with a program of journal clubs, seminars, and joint laboratory meetings. There are many academic opportunities for neurosurgical residents, who can spend their fifth year entirely engaged in laboratory or clinical research. This emphasis has enabled our residency program to send graduates on to academic careers. The great majority of our graduates ultimately work in university departments of neurosurgery.
Research in the Placantonakis Lab
The laboratory of Dimitris G. Placantonakis, MD, PhD, studies glioma, a primary brain tumor with a poor survival prognosis and limited treatment options. The two major projects in the laboratory are heterogeneity in the cancer stem cell population in glioblastoma, and modeling the origins of glioma with human embryonic stem cells.
Glioblastoma is the most aggressive form of glioma. Over the past decade, several lines of evidence have indicated that not every cell in these tumors is equal. In biological terms, this inequality translates into a cellular hierarchy, with cancer stem cells at the apex.
Throughout the body, tissue homeostasis is maintained by stem cells that generate defined lineages of specialized differentiated cells. Not surprisingly, glioblastoma tumors are remarkably complex at the histologic level, which begs the question: Does a single cell type with stemlike properties produce the entire spectrum of tumor lineages and histologies, or is the cancer stem cell population heterogeneous?
Using human glioblastoma cultures, we recently discovered that within any given glioblastoma tumor there are multiple tumor cell types that fulfill stem cell criteria. These cell types manifest striking transcriptional and metabolic differences, in addition to discrete differentiation programs, that allow them not only to adapt to diverse microenvironmental conditions but also to shape the niches where they reside. Our work focuses on elucidating molecular and cellular mechanisms that govern heterogeneity in glioblastoma’s stem cell population.
When gliomas are found in younger patients, they are often less aggressive, or “low-grade,” gliomas, but they inevitably transform into more aggressive tumors. The early steps in gliomagenesis—such as the process whereby a normal brain cell turns into a glioma cell—are not well understood, and there are no good mouse models for low-grade gliomas. In our laboratory, we use human embryonic stem cells and their neural progeny to understand how cocktails of oncogenes and inactivated tumor suppressors promote oncogenic transformation. Learn more about the Placantonakis Lab.
Research in the Rice Laboratory
Research in the laboratory of Margaret E. Rice, PhD, focuses on factors that regulate release of dopamine, a neurotransmitter that plays crucial roles in movement and motivation. Dopamine dysregulation has been implicated in movement disorders such as Parkinson’s disease and addiction to natural rewards like food, as well as to drugs of abuse. Learn more about the Rice Lab.
Research in the Kondziolka Radiosurgery Program
Douglas Kondziolka, MD, MS, together with investigators in radiation oncology, neuroradiology, pathology, and medical physics, studies imaging and clinical responses to stereotactic radiosurgery for a wide variety of clinical indications. A key focus of investigation is imaging-defined responses—anatomic, metabolic, and cellular—to targeted tumors, vascular malformations, and structures in functional disorders. We utilize radiomics techniques for both benign and malignant tumors. The program not only studies responses to care, but also natural history studies of benign tumors such as vestibular schwannomas and meningiomas to evaluate growth rates. Another key area of research interest is the common oncologic challenge of metastatic brain tumors.
Investigators in our program worked to develop one of the largest prospective registries for data collection. Used daily, the registry maintains key data on outcomes to facilitate research. We work with researchers in information sciences to develop predictive analytics and other software tools.
Artificial Intelligence Research
Eric K. Oermann, MD, and his colleagues study intelligence—what it is and how we can protect it as neurosurgeons and artificial intelligence (AI) researchers. This research requires us to build AI systems, and in doing so we focus on building solutions that also solve practical problems in medicine. Specific AI topics that are of interest to Dr. Oermann are data efficiency and k-shot learning, the generalizability of algorithms, and reinforcement learning. Dr. Oermann is also interested in the application of AI to solve problems in clinical neurosurgery and oncology.
Dr. Oermann’s research has a strong focus on biomedical imaging and working with both pathology and radiology data. Problems that he investigates include the classification and semantic segmentation of tumors and vascular lesions in medical imaging, and using radiomics to predict tissue type and gene expression directly from imaging data. Another area of investigation is questions of data efficiency for medical AI problems using both generative adversarial networks as well as transfer learning to learn from limited amounts of data. This line of applied research has focuses not only on developing these tools, but investigating novel methods of assessing them in clinical trials—using machine learning to run clinical trials for machine learning.
Recent work by Dr. Oermann’s team has delved into basic mechanisms of machine learning and human learning. They have studied weakly supervised learning, and novel methods for improving optimization algorithms using momentum and contrastive learning. They have also investigated the relationship between weakly and strongly supervised computer vision models and ways of improving weakly supervised learning. As a means of simulating human learning, his team has begun to investigate the use of techniques from reservoir computing to model human intracranial electrophysiology data, and attempt to associate computer representation learning with human representation learning in the hopes of better understanding how the human brain encodes and accesses data.