Research Projects

We developed an AI algorithm to automate the generation of Vestibular Schwannoma 3D volumetric tumor models. To build this algorithm we collected and processed hundreds of images.

This tool has:

• Decreased the time taken to generate 3D tumor models from hours to minutes
• A greater than 88% accuracy

We are developing algorithms for many additional tumors including: meningiomas, chordomas, craniopharyngiomas, germ cell tumors, glomus jugular tumors, pineocytomas, pituitary adenomas, glioblastomas, hemangioblastomas, rhabdoid tumors, neuronal tumors, mixed glioneuronal tumors, optic path gliomas, CNS lymphomas and plexiform neurofibromas.

To develop our algorithms, we require hundreds of MRI scans of CNS tumors. Thus, we work closely with many nonprofit organizations to recruit MRI images of patients who are interested in contributing their scans to help us build new AI algorithms. If you have personal MRI images for any of these tumors and would like to contribute to the development of a ground truth database for this tumor type, please reach out to 3Dtumorlab@yale.edu.