Brain temperature differs from body temperature and is an important indicator of health after an injury, but it's difficult to measure without invasive procedures such as the implantation of temperature probes. A group of researchers has proposed a new biophysical model that can predict personalized brain temperature using data from magnetic resonance imaging of individual brain tissue and vessel structures.

A paper published April 15 in Communications Physics outlines the team's effort to improve the ease of brain thermometry, or the measurement of brain temperature, in clinical settings. The temperature of the brain is higher than body temperature on average because of the brain's distinct regulatory mechanisms — the brain has a high level of metabolic activity, which produces heat.

Brain temperature also tends to rise faster than body temperature following head injuries, stroke and cardiac arrest, and previous research has shown that dissociation of brain and body temperatures was a strong predictor of poor prognoses and deaths in neurosurgery patients, according to the paper.

Senior authors Candace C. Fleischer, an assistant professor affiliated with both Emory University School of Medicine and the Georgia Institute of Technology, and Andrei G. Fedorov, a professor of engineering at the Georgia Institute of Technology, explained to The Academic Times that they used computational modeling to predict biophysical behavior, based on the first principles of energy and mass conservation, in order to predict brain temperature.

Computational modeling refers to the use of computers to simulate and study complex systems through math, physics and computer science, while a biophysical model simulates a specific biological system using the mathematical formalizations of the physical properties of that system.

"To address the clear need for an expanded understanding of brain thermoregulation, particularly after injury, biophysical models have been developed to understand drivers of local variations in brain temperature and close the gap between empirical observations and underlying heat transfer mechanisms," the authors said in the paper. 

Previous experimental approaches to brain thermometry have used temperature-sensitive magnetic resonance parameters, but direct validation of them in healthy volunteers is not feasible without a "gold standard" of measurement like a temperature probe, according to the paper. Most of the methods have not been successfully integrated into clinical practice, and none have accounted for individual differences in metabolism, vessel structure and blood flow, all of which can vary widely person-to-person.

Fedorov described the human brain as a "complex biological machine" where many processes managing blood flow take place simultaneously. A network of different-sized blood vessels delivers blood throughout the brain and also moves blood out. Alongside this process, heat is moved in and out of the brain, Fedorov explained; this heat can result from the generation of thermal energy in different brain regions.

The researchers were able to produce a model of all these processes based on the fundamental principles of energy and mass conservation, and supplement it using data from healthy human brains. This process lets them create a tailored tool that can predict how the temperature of an individual person's brain may evolve over time, which has important implications for brain health, Fedorov said.

The model is derived from first principles, which means imposing mass and energy conservation locally and using constitutive equations for linking energy flow and temperature, the authors wrote. And to construct the model equations, they used three heat transfer domains, including arteries, veins and tissue, and three modes of energy transfer, including conduction, advection, which is the transport of thermal energy by a moving fluid, and convection, the combination of conduction and advection that transfers heat from a flowing fluid to a wall.

Magnetic resonance data were collected from three healthy adults for the study. Tissue probability maps, segmented vessel structures and other information was taken from those MR images and used as input data for the team's biophysical model.

"While it is relatively straightforward to generate a simple and generic model of the brain, the challenge is everybody's brain is different, and it's surprisingly more different than people realize," Fleischer said. "We collected brain tissue and vessel structure data using MRI from each healthy adult, and used these data to develop a model that can simulate or predict individual brain temperatures. Each adult had unique 3D brain temperature patterns, emphasizing how important it is to be able to generate personalized temperature maps."

The researchers compared their subject-specific model prediction with generic brain predictions, which were generated using the current state-of-the-art method. Both models predicted brain temperatures that were higher than body temperature. However, the use of individual data on tissue and vessel structure in the subject-specific model revealed unique 3D temperature patterns across the brain of each individual. 

Fedorov and Fleischer noted that their study is a first step — the topic needs to be further researched before this approach can be integrated into clinical settings. But it confirmed what scientists already suspect and assume, which is that everybody's brain is different, and furthers the notion that neurology patients need to be diagnosed and treated on an individual basis. "Personalized medicine is changing the way we're treating patients, and changing the way we're implementing emerging treatments. These data provide more evidence to support the importance of personalized and tailored approaches, particularly for brain health," Fleischer said.

Larger studies with more patients are needed for proper validation of the model, the authors said in the paper. Fedorov and Fleischer are working on a follow-up project that incorporates data from patients who have experienced an injury such as a stroke, in order to try and simulate what their brain temperature evolution may look like over time. 

"The important role of temperature after injury to the brain is undisputed, however, the underlying mechanisms that relate higher temperatures to tissue damage are less clear," the authors said in the paper. "As patient care becomes more personalized, biophysical models that facilitate individual predictions at the subject level may have broad applications in prognosis, treatment stratification, and patient monitoring."

The study, "Personalized predictions and non-invasive imaging of human brain temperature," published April 15 in Communications Physics, was authored by Dongsuk Sung and Candace C. Fleischer, Georgia Institute of Technology and Emory University; Peter A. Kottke and Andrei G. Fedorov, Georgia Institute of Technology; and Benjamin B. Risk, Jason W. Allen and Fadi Nahab, Emory University.