With quality control and data harmonization tools leading the charge
Leading a startup grounded in a technical subject demands a certain style when describing our work to the subject matter to lay audiences, such as government partners, media, or the public. At the same time, leaders in my field face understandable regulatory constraints on what we can say about the products we’re building.
Last week, I was reminded of what it’s like to be part of the lay audience. I attended the ISMRM Diffusion Workshop on Diffusion MRI (dMRI) in Amsterdam. This leading conference, held once every five years, brings together some of this field’s leading-edge researchers to discuss recent advancements in their work.
Thoughtfully, the ISMRM workshop’s theme was, “From Research to Clinic.” The organizers also included a forum for clinicians and researchers to present to and learn from each other.
Much dMRI research focuses on unraveling the routes and connectivity of white matter, the brain tissue that consists of nerve fibers and their myelin sheaths. Some research scanning protocols can take significant time to acquire data, to provide higher fidelity, and more accurate reconstructions and representations of fiber paths.
Imaging access in clinics, however, sits at the other end of the time spectrum. Demand for MR systems far outstrips their availability, which means clinical machines are constantly in use. As a result, the range of available scanning protocols in a hospital setting are often determined by the time they require to complete, a constraint far removed from research contexts.
Accordingly, when surveyed about routine clinical studies, 70 percent of radiologists suggested that five minutes was the maximum time they could allow for a patient scan.
That hard limitation presents a significant challenge to the field. How can we distill complex research findings to fit within that five-minute efficacy window? Given this constraint, it’s unsurprising that clinical practice continues to lag cutting-edge developments.
During the workshops, many clinician talks identified the same recurring themes, gaps, and requests that would help them better take care of their patients.
My takeaways from this year’s ISMRM workshop
- As MRI scanning expands in use, the scans required to make an initial diagnosis and follow-up with the patient will increasingly be conducted on different systems in a mix of hospitals and out-patient clinics. Improved reference standards and harmonization tools are needed to accurately monitor and identify disease progression, or trauma recovery.
- Despite the increases in advanced data acquisition protocols, validation issues mean they cannot consistently and unambiguously distinguish between different substrates and fiber. This limit makes determining fiber directions impossible. It also throws into question their relative importance, which is hard to determine in clinical settings such as neuro-oncology.
- Dependable, accurate dMRI scans promise to significantly benefit patients with traumatic brain injuries. With this data, their clinical teams could more easily determine the severity of a condition, predict symptoms, and efficacious courses of treatment.
- We’ve only begun to tap the number of applications where dMRI can benefit patients. Fields of medicine that would not have considered using dMRI in the past are now seeing its potential. For example, research is showing that heart, liver, and placental imaging are areas where an increasing number of conditions could be identified with dMRI technology. Research is at an early stage with fewer overall researchers.
- Machine learning in medicine remains in its infancy. Many applications are possible and the potential for machine learning to support clinicians as a digital augment is very powerful.
Research, clinical practice, and industry are three legs of the same stool required to discover, heal, and disseminate the tools and technologies that will globally raise the standard of neurological care.
Thank you to the ISMRM Diffusion Study Group and Committee for creating such an interesting, fruitful agenda for the workshop.
