好色先生

好色先生

Explore the latest content from across our publications
Join The AAN

Log In

Forgot Password?
Create New Account

Loading... please wait

Abstract Details

Classification of ataxic gait by shallow classifiers
General Neurology
General Neurology Posters (7:00 AM-5:00 PM)
073
To assess gait ataxia using accelerometers and to compare accuracy of different data reduction and classification methods for use in clinical practice.
Gait disorders accompanying a number of neurological and musculoskeletal disorders significantly reduce the quality of life. Motion sensors enable high-quality modeling of gait stereotypes. At the same time, however, they produce large volumes of data, the evaluation of which is a challenge.
The recordings of 40 participants (17 ataxic, 23 age-matching reference set) from Perception Neuron device were preprocessed by data reduction algorithms and classified by shallow machine learning algorithms. We compared Principal Component Analysis, Uniform Manifold Approximation and Projection and T-Distributed Stochastic Neighbor Embedding data reduction methods.  The following machine learning methods have been tested: Logistic Regression, Linear Support Vector Machine, Kernel Support Vector Machine, Principal Component Analysis (PCA), Naive Bayes, Quadratic Discriminant Analysis, k-Nearest Neighbor, Decision Tree,  Random Forest,  Neural Net and Adaboost.
Best achieved accuracy between group of healthy individuals and patients with ataxic gait is 98 % by Decision Tree classifier combined with t-Distributed Stochastic Neighbor Embedding (TSNE).
Accelerometric data evaluated by shallow classifiers are highly accurate in ataxia recognition. TSNE is also able to effectively separate subjects according to the severity of the clinical disability measured by SARA scale. These results support future introduction of wearable technology into clinical practise.
Authors/Disclosures

PRESENTER 		No disclosure on file
Ondrej Dostal The institution of Ondrej Dostal has received research support from Ministry of 好色先生 Youths and Sports.
Oldrich Vysata No disclosure on file
Ladislav Pazdera, MD, FAAN (Vestra Clinics S.r.o.) Dr. Pazdera has received personal compensation in the range of $500-$4,999 for serving as a Consultant for Biogen. Dr. Pazdera has received personal compensation in the range of $500-$4,999 for serving on a Speakers Bureau for Teva.
Martin Valis, MD Dr. Valis has received personal compensation in the range of $5,000-$9,999 for serving on a Scientific Advisory or Data Safety Monitoring board for Biogen. Dr. Valis has received personal compensation in the range of $500-$4,999 for serving on a Speakers Bureau for Merck. Dr. Valis has received personal compensation in the range of $500-$4,999 for serving on a Speakers Bureau for Teva.