This Year’s Model
Perhaps more than any other medical field, neurotechnology involves a vast array of variables and parameters that can have a profound effect on the viability of a new therapy or diagnostic test. This panorama of parameters such as target location, intensity, pulse width, and contact material creates vast opportunities for neurotech investigators and entrepreneurs to devise new products. But it also can make it harder for the community as a whole to obtain a robust understanding of the mechanisms of action for a particular neuromodulation approach.
This is an issue that two neurotech researchers took up in a perspective article in a recent issue of the journal Bioelectronic Medicine. The authors, Marco Capogrosso from the Rehabilitation Neural Engineering Laboratories at the University of Pittsburgh, and Scott Lempka from the University of Michigan biomedical engineering department, argue in favor of developing computational models of how neuromodulation therapies achieve their effect.
“The neuroscience community is too quickly focusing on ‘translational applications’ (i.e. the translation of scientific discoveries in neuroscience to clinical settings),” they wrote. “Fostered by the urge to solve the impelling needs of an aging society, funding bodies provide ever-increasing support to this type of research. Given the stakes, as members of the scientific community and information-era human beings, we should question the very concept of translation, and approach this task with the most rigorous scientific attitude.” They added that a computational model offers other investigators a “virtual testing platform” that can be exploited even before animal models are introduced.
Models developed by researchers such as Warren Grill at Duke and Cameron McIntyre at Case have been highly useful in understanding the potential therapeutic effect of changes in stimulation parameters. The authors also cited their own work, in cooperation with Gregoire Courtine at EPFL to develop a model of the interaction and alternate recruitment of agonist and antagonist muscles in locomotion. This work proved helpful in devising therapeutic regimens to restore locomotion after spinal cord injury.
Although efforts such as these help advance basic neuroscience, they also can have positive effect on commercial development. Indeed, the FDA has signaled its intention to incorporate computational modeling and virtual testing in the device approval process, as a way to expedite commercialization of new devices and therapies.
“Ultimately, we could provide the impatient information-era public community with hopes of accelerating the slow pace of clinical technology development and validation,” the authors conclude. We tend to agree with that assessment.
Editor and Publisher