Yale Ventures has awarded a grant to an interdisciplinary team of Yale researchers for a project titled, “Dynamic 3D Morphometric Analysis.” The two-year $150,000 grant, funded through the Blavatnik Fund for Innovation, was awarded to associate research scientist, Frank Buono, PhD, associate professor, Daniel Wiznia, MD, and research scientist, Steven Tommasini, PhD, at the Yale Innovation Summit in recognition of their impact through life-science acceleration.

Buono, who is appointed in psychiatry and orthoapedics, has collaborated on this effort for the past several years with Wiznia and Tommasini, both of whom have dual appointments in orthopaedics and engineering.

This support comes after the team received a two-year $50,000 grant that was funded through the 2022 Yale Center for Clinical Investigation (YCCI) Translational and Interdisciplinary Research Pilot Program for their work on utilizing machine learning to model the volumetric growth of vestibular schwannomas.

The endeavor began with the creation of a novel, AI-based, 3D evaluation tool for clinicians to evaluate vestibular schwannoma (VS) tumors. A defining characteristic of Neurofibromatosis type II (NF2) is the presentation of VS tumors. While benign, VS tumors may impinge on vital structures such as cranial nerves causing morbidity or pressing on the brainstem, which can be life-threatening. It is clinically important to accurately measure and evaluate the growth of VS tumors, especially for neurologists, neurosurgeons, and radiation oncologists to better understand the regions and rates of growth and structures that are at risk.

Magnetic resonance imaging (MRI) is currently considered the gold-standard for assessing the size, location, and evolution of VS tumors. Traditionally, tumors are analyzed linearly in two dimensions on planar views, but the methodology has been demonstrated to be inaccurate and time intensive. The 3D volumetric technique proposed by these researchers is more sensitive in identifying tumor progression.

In developing dynamic 3D morphometric analyses, the team of researchers aims to quickly and accurate utilize this new quantitative image analysis system to precisely map the size and shape of VS tumors through the utilization of AI to better understand the intricate details of each unique biological structure.

“This will be a game changer in brain tumor treatment,” Buono said. “This software has the ability to facilitate care for both clinicians and patients.”

The interdisciplinary team automated the labor-intensive process with machine learning to develop 3D diagnostics for medical providers of NF2 to visualize and quantify VS tumors in 3D. To accomplish this goal, the team will build a ground truth dataset of NF2 VS MRI based 3D models, develop an artificial intelligence algorithm to automate the rapid creation of 3D models of VS tumors, and create visualization tools to analyze growth and highlight at risk structures.

This innovative approach to personalized 3D modeling of brain tumors with the integration of AI has the potential to significantly increase the understanding of spatial characteristics of brain tumors, and the researchers will translate these advanced tools into the clinical, academic, and surgical settings to facilitate care for these diseases. This initial project will lay the groundwork for expansion of a volumetric approach to other CNS tumors.