Group leaders: Dr. Dominique Duncan and Dr. Amir Jaberzadeh
Medical sciences have benefited from emerging open source web services and virtual technologies in recent years. These tools provide a common virtual platform for data collection, processing pipelines, and cross validation of research outcomes. Using these virtual services facilitates collaboration among remote sites and allows separate teams to reproduce study findings. Several large blueprint NIH grants followed this trend and opened their brain databases as well as their image processing pipelines, which enabled researchers with different backgrounds to test their clinical hypotheses more easily and collaboratively. However, many researchers are using common datasets and processing pipelines, assessing quality of the processed data and manual correction of the failures are existing challenges for those researchers. Several neurologists or radiologists are required to go through thousands of images, overlay processed data, browse each patient’s data, and manipulate the data to prepare them for their specific studies. It is a tedious and time-consuming task, which can take up to a year for a study with a medium sample size. This creates a major bottleneck in the experimental process. Remote access to processed data is a challenge due to large file sizes, lack of software support, and data protection policies. To address these challenges, we have developed a collaborative web service for quality control as well as virtual reality (VR) software for manual manipulation of neuroanatomical magnetic resonance imaging (MRI) data.
This group aims to promote the aforementioned tools to reach a wider community of researchers who study brain structural MRI. Additionally, they hope to recruit researchers, neurologists, and radiologists to use these platforms in their own circle of collaborations. We plan to bring together both neuroimaging experts as well as non-experts from other fields who can work together to expedite the progress in the field of neuroimaging research.
Learn more about their work:
- “Quality control of MRI segmentation using virtual reality and crowdsourcing” dynamic poster presentation at the Society for Neuroscience Annual Meeting in San Diego, CA, November 2018
- “Correcting segmentation errors in neuroimaging data using virtual reality” invited talk and demo at the first annual USC Virtual Technologies for Health Symposium in Los Angeles, CA, September 2018
- “Visualization and manual correction of brain segmentation failures using virtual reality” poster presentation at the Organization for Human Brain Mapping (OHBM) Annual Meeting in Singapore, June 2018
- “Virtual Brain Segmenber (VBS)” exhibit booth at the annual American Academy of Neurology (AAN) meeting in Los Angeles, CA, March 2018
- “Virtual reality and neuroscience” invited talk at the 2018 Los Angeles Brain Bee, Los Angeles, CA, February 2018
Dominique Duncan, MD
Assistant Professor, Neurology