MindTrace’s software will predict outcome from neurosurgery

CMU spinoff company harnesses machine learning and AI to improve neurosurgery

MindTrace Technologies Inc. (MindTrace), a Carnegie Mellon University spinoff company, is innovating neurosurgery by providing clinical teams with forecasts about how a given surgical plan will affect a patient’s future mental function.

The brain child of co-founders Brad Mahon, Max Sims and Hugo Angulo (SCS 2018), the company’s software is akin to upgrading your navigation on a road trip from a dusty road atlas to taking turn-by-turn GPS directions with updates about hazards along the way. This more refined and personalized predictive map can help doctors reduce the risk of harming regions of the brain that house a person’s core functions, like speech or memory.

There are several patient populations requiring brain surgery that may ultimately benefit from MindTrace’s technology, including brain tumor and epilepsy.

“At present, MindTrace is laser-focused on the very specific clinical challenges of mapping language function during awake brain tumor surgery,” said Mahon, associate professor in the Department of Psychology. “We will start by building a dataset around awake language mapping, with the goal of being able to predict post-operative language abilities during the neurosurgical procedure. From there, our platform can grow to address other critical functions (memory, perception, action) and be extended to other patient populations, such as epilepsy.”

This company grew out of a need to bridge cognition and anatomy. Every brain is unique, and the exact location where cognitive functions reside may vary by a few centimeters from person to person. This real estate matters, because neurosurgeons navigate on the millimeter scale while making decisions during surgery.

The medical software provides an evidence-based approach to predict how removing brain tissue will affect a patient’s post-operative functions. The process begins by obtaining scans of the brain before surgery using functional Magnetic Resonance Imaging (MRI). These scans are paired with anatomical or structural MRI scans and computerized tomography (CT) to develop a 3D model of a person’s brain. This information helps the team plan a surgical path based on predicted outcomes, all before the first incision. The team can also monitor the results in real time during surgery.

“Resection of brain tissue in ‘eloquent’ brain regions presents a complex surgical challenge,” said Webster H. Pilcher, M.D., Ph.D., the Ernest & Thelma Del Monte Distinguished Professor of Neuromedicine at the University of Rochester. Pilcher has used early versions of MindTrace’s technology, in a research capacity, during neurosurgical procedures. “MindTrace is changing the game by offering clinical teams a software tool that provides a ‘menu’ of tasks to customize pre-operative and intraoperative neurocognitive mapping for each patient. We are excited for MindTrace’s future development, as its software will be able to not only summarize the results of an awake mapping procedure, but also provide a new type of evidence that can support real-time decisions during the surgery.”

To date, two medical centers have used the MindTrace software in a research capacity, assisting neurosurgeons with more than 60 surgeries, including a widely discussed surgery of a young musician who had a benign tumor removed from a region of his brain associated with listening to and making music. 

Using the data gathered from each surgery, MindTrace is constantly iterating and improving. According to Mahon, the data generated from one patient can inform care decisions about future patients in any operating room around the world.

“We will be able to make a prediction, using machine learning and artificial intelligence, before surgery happens,” said Mahon. “The data we collect when we measure how that patient actually did after their surgery is then fed back into the model to improve its ability to predict outcomes in future patients.”

To scale the process and assist more patients, the team aims to begin prospective data collection across a consortium of medical centers that span a wide geographic footprint of the country in 2023. This consortium will be anchored by Mahon’s long-standing relationship with the University of Rochester Medical Center. In parallel, MindTrace is continuing with regulatory submissions to the U.S. Food and Drug Administration to identify the appropriate regulatory pathway and ensure the software is safe and effective to aid in clinical decision making.

CMU has been integral in navigating the challenges for the young startup company.

“CMU has been an incredible resource for MindTrace by creating the conditions to transfer and scale technology from Brad’s lab,” said Sims, CEO of MindTrace. “[The university] has also acted as a force multiplier in the Pittsburgh ecosystem by connecting us with local health science and medical technology resources to de-risk the company.” 

The team hopes that in five years patients will be aware of mindtracing and that this technology will become a standard of care in neurosurgery to protect a patient’s cognitive functions and ensure the best possible outcomes.

“The expansion of MindTrace from a specific area of awake brain surgery to neurosurgery writ large is a [long-term] goal,” said Mahon. “As our clinical studies and predictive modeling progress, the resulting datasets will create new opportunities for expansion of MindTrace’s use.”

MindTrace was recently highlighted in a Pittsburgh INNO article, featuring Pittsburgh’s most promising tech companies.