A microstructural neural network biomarker for dystonia diagnosis identified by a DystoniaNet deep learning platform.
Proc Natl Acad Sci U S A 2020;
117:26398-26405. [PMID:
33004625 PMCID:
PMC7586425 DOI:
10.1073/pnas.2009165117]
[Citation(s) in RCA: 18] [Impact Index Per Article: 4.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/07/2023] Open
Abstract
This research identified a microstructural neural network biomarker for objective and accurate diagnosis of isolated dystonia based on the disorder pathophysiology using an advanced deep learning algorithm, DystoniaNet, and raw structural brain images of large cohorts of patients with isolated focal dystonia and healthy controls. DystoniaNet significantly outperformed shallow machine-learning pipelines and substantially exceeded the current agreement rates between clinicians, reaching an overall accuracy of 98.8% in diagnosing different forms of isolated focal dystonia. These results suggest that DystoniaNet could serve as an objective, robust, and generalizable algorithmic platform of dystonia diagnosis for enhanced clinical decision-making. Implementation of the identified biomarker for objective and accurate diagnosis of dystonia may be transformative for clinical management of this disorder.
Isolated dystonia is a neurological disorder of heterogeneous pathophysiology, which causes involuntary muscle contractions leading to abnormal movements and postures. Its diagnosis is remarkably challenging due to the absence of a biomarker or gold standard diagnostic test. This leads to a low agreement between clinicians, with up to 50% of cases being misdiagnosed and diagnostic delays extending up to 10.1 y. We developed a deep learning algorithmic platform, DystoniaNet, to automatically identify and validate a microstructural neural network biomarker for dystonia diagnosis from raw structural brain MRIs of 612 subjects, including 392 patients with three different forms of isolated focal dystonia and 220 healthy controls. DystoniaNet identified clusters in corpus callosum, anterior and posterior thalamic radiations, inferior fronto-occipital fasciculus, and inferior temporal and superior orbital gyri as the biomarker components. These regions are known to contribute to abnormal interhemispheric information transfer, heteromodal sensorimotor processing, and executive control of motor commands in dystonia pathophysiology. The DystoniaNet-based biomarker showed an overall accuracy of 98.8% in diagnosing dystonia, with a referral of 3.5% of cases due to diagnostic uncertainty. The diagnostic decision by DystoniaNet was computed in 0.36 s per subject. DystoniaNet significantly outperformed shallow machine-learning algorithms in benchmark comparisons, showing nearly a 20% increase in its diagnostic performance. Importantly, the microstructural neural network biomarker and its DystoniaNet platform showed substantial improvement over the current 34% agreement on dystonia diagnosis between clinicians. The translational potential of this biomarker is in its highly accurate, interpretable, and generalizable performance for enhanced clinical decision-making.
Collapse