Davide Valeriani, PhD: Using the DystoniaNet Platform for Dystonia Diagnosis

October 6, 2020

The postdoctoral research fellow in the Dystonia and Speech Motor Control Laboratory at Mass Eye and Ear spoke to the clinical translation of the AI-based DystoniaNet tool.

“We specifically took an approach that could be easily clinically translated, which is why we used raw structural MRI. This kind of measure is what comes out directly from the scanner and is available to many clinicians in-clinic.”

Recently, data were published displaying the performance of an artificial intelligence (AI) based diagnosis platform, called DystoniaNet, which suggest that the machine learning platform can accurately diagnose dystonia in a fraction of a second. In addition to its 98.8% accuracy in diagnosis, the tool is also capable of recommending when the diagnosis requires supplementary opinions from the physician.

Davide Valeriani, PhD, postdoctoral research fellow, Dystonia and Speech Motor Control Laboratory, Mass Eye and Ear and Harvard Medical School, an author on the paper, spoke with NeurologyLive in an interview to offer his perspective on the platform’s performance, and to explain how DystoniaNet could translate clinically to practice. While additional studies are needed—some of which are underway—he noted that the way the system was designed would allow for almost seamless translation to in-clinic use.

Valeriani explained that the DystoniaNet platform takes MRI data and learns from the training algorithm developed from a large dystonia dataset, then uses this data-driven approach to identify the condition. It was developed over the last several years by Valeriani and coauthor Kristina Simonyan, MD, PhD, DrMed, director, Laryngology Research, Mass Eye and Ear; associate neuroscientist, Massachusetts General Hospital; and associate professor, Otolaryngology-Head and Neck Surgery, Harvard Medical School, and was previously evaluated in an analysis-driven study.

REFERENCE
Valeriani D, Simonyan K. A novel microstructural neural network biomarker for dystonia diagnosis identified by a DystoniaNet deep learning platform. PNAS. Published September 28, 2020. doi: 10.1073/pnas.2009165117.