Feasibility of spirography features for objective assessment of motor function in Parkinson's disease

• Published in: Artificial intelligence in medicine Vol. 81; no. SI; pp. 54 - 62
• Main Authors: Sadikov, Aleksander, Groznik, Vida, Možina, Martin, Žabkar, Jure, Nyholm, Dag, Memedi, Mevludin, Georgiev, Dejan
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.09.2017
• Subjects: Aged; Antiparkinson Agents - adverse effects; Diagnosis, Computer-Assisted - methods; Dyskinesia, Drug-Induced - diagnosis; Dyskinesia, Drug-Induced - drug therapy; Dyskinesia, Drug-Induced - physiopathology; Feasibility Studies; Female; FLOAT - Flexibel levodopa-optimerings och individanpassningsteknik; FLOAT - Flexible Levodopa Optimizing Assistive Technology; Health Status; Humans; Hypokinesia - diagnosis; Hypokinesia - drug therapy; Hypokinesia - physiopathology; Image Processing, Computer-Assisted - methods; Informatics; Informatik; Internal Medicine; Machine Learning; Male; Middle Aged; Motor Activity - drug effects; Movement disorder; Objective monitoring; Other; Parkinson Disease - diagnosis; Parkinson Disease - drug therapy; Parkinson Disease - physiopathology; Parkinson's disease; Predictive Value of Tests; Retrospective Studies; Severity of Illness Index; Spirography; Spirography features; Sweden; Time Factors; Treatment Outcome; Upper Extremity - innervation; Visualisation; Visualisation; Spirography features; Parkinson's disease; Movement disorder; Objective monitoring; Spirography
• ISSN: 0933-3657; 1873-2860; 1873-2860
• DOI: 10.1016/j.artmed.2017.03.011

Highlights • A method for self-monitoring the motor function in Parkinson's disease is presented. • Slowness of movement (bradykinesia) is typically associated with under-medication. • Involuntary movements (dyskinesia) can be the result of over-medication. • A machine learning model that detects bradykinesia and dyskinesia is proposed. • The model's visual explanatory power is evaluated.