3D Tumor Lab

In 2022, we encountered a patient with Neurofibromatosis Type 2 (NF2) who presented with bi-lateral vestibular schwannomas who reported a worsening of symptoms, despite being told that their tumors had not grown over the past 6 years. It became clear that there was an issue with the current methods for measuring tumor growth with the common method being measuring linear distances on a single MRI slice. We questioned if this method was accurately detecting this specific patient's tumor growth over the years. At that time, 3D segmentation technology, which had gained popularity in orthopedics for analyzing bones in multiple planes, caught our attention. We wondered: why couldn't this technology be applied to brain tumors?

Inspired by its success in other fields, we began manually segmenting MRIs of this patient's vestibular schwannomas over several years. Our analysis revealed a startling discrepancy: while the patient had been informed that their tumor had remained stable, it had actually grown to four times its original volume. This unchecked growth rendered the tumors inoperable, significantly impacting the patient's treatment options.

This experience underscored the need for more accurate tumor analysis and led to the creation of the 3D Tumor Lab. We realized that this problem likely affected many other patients who had only received linear tumor assessments. Thus, our mission is to enhance the accessibility and precision of 3D tumor growth analysis. We aim to advance from manual segmentation to an automated system that generates 3D models of tumors with just the click of a button, ultimately offering patients improved chances for recovery and survival.

Our lab is focused on developing software to automate the identification, measurement, and growth of central nervous system (CNS) tumors. Our goal is to generate clinical tools that are easy to use, accessible and fast so that volumetric tumor analysis can be used in clinical settings.

Did you know that commonly used 2D tumor analysis methods drastically underestimate the growth rates of tumors, causing clinicians to make uninformed decisions? Our lab has demonstrated inaccuracies in the current standard of care for measuring tumor growth.

Did you know that the only widely available method to measure tumor volumes involves time intensive manual segmentation that can take 3-4 hours per scan? Our AI can generate 3D tumor models for volumetric analysis within minutes and is highly accurate.

We are a team made up of innovative physicians, students and engineers dedicated to using advanced technologies to deliver faster, more accurate and more accessible information to improve patient outcomes.