Prediction of expanded disability status scale subscores of motor dysfunction in multiple sclerosis using depth-sensing computer vision
Author(s): ,
M. D'Souza
Affiliations:
Neurology, University Hospital Basel, Basel, Switzerland
,
J. Burggraaff
Affiliations:
Neurology, VU University Medical Center, Amsterdam, The Netherlands
,
P. Kontschieder
Affiliations:
Microsoft Research, Cambridge, United Kingdom
,
J. Dorn
Affiliations:
Novartis Pharma AG, Basel
,
C.P. Kamm
Affiliations:
Neurology
,
S. Seinheimer
Affiliations:
University Hospital Bern, Bern, Switzerland
,
P. Tewarie
Affiliations:
Neurology, VU University Medical Center, Amsterdam, The Netherlands
,
C. Morrison
Affiliations:
Microsoft Research, Cambridge, United Kingdom
,
A. Sellen
Affiliations:
Microsoft Research, Cambridge, United Kingdom
,
A. Criminisi
Affiliations:
Microsoft Research, Cambridge, United Kingdom
,
F. Dahlke
Affiliations:
Novartis Pharma AG, Basel
,
B. Uitdehaag
Affiliations:
Neurology, VU University Medical Center, Amsterdam, The Netherlands
L. Kappos
Affiliations:
Neurology, University Hospital Basel, Basel, Switzerland
(Abstract release date: 09/23/15) ECTRIMS Online Library. D'Souza M. 10/09/15; 115783; 1074
Marcus D'Souza
Marcus D'Souza
Contributions
Abstract
Abstract: P818

Type: Poster

Abstract Category: Clinical assessment tools

Background: Clinical assessment of impairment and disability in Multiple Sclerosis (MS) remains the most important outcome in therapeutic trials, and is commonly assessed with the Expanded Disability Status Scale (EDSS). However, the EDSS exhibits high inter- and intra-rater variability. The ASSESS MS system is being developed as a non-invasive, more consistent and potentially finer grained tool to measure motor dysfunction in MS, by combining recordings of prescribed neurological movements with machine learning methods to assess motor dysfunction based on EDSS subscores.

Objectives: To test the prediction of EDSS subscores from recordings of a depth-sensing video analysed by machine learning algorithms.

Methods: Pre-defined movements from the EDSS assessment were recorded in 300 patients and 200 healthy volunteers. Video recordings of patients were scored by four neurologist from 3 sites based on the Neurostatus/EDSS assessment definitions. These scores were used to train a machine learning algorithm to correctly predict motor dysfunction from depth-sensing video recordings, which capture a 3D view of the patient.

Results: We report that on movements covering upper extremities and trunk, the machine learning algorithm predicts motor dysfunction of patients with MS with an accuracy similar to neurologists' intra-rater retest reliability. For example, the agreement of the upper extremity tremor/dysmetria subscore from the finger-to-nose test of the algorithm with the neurologists' assessment is 73% across scores 0, 1, 2 and 3, while the long-term intra-rater agreement of the neurologists with their previous assessment is 67% in this challenging-to-assess range.

Conclusions: Automated quantification of movement recordings using a depth-sensing camera and image analysis based on a machine-learning algorithm enables an accurate and sensitive quantitative assessment of motor dysfunction in MS patients. ASSESS MS is expected to improve the evaluation of disability progression in clinical studies and clinical practice.

Disclosure:

M. D`Souza received travel support from Bayer AG, Teva and Genzyme and research

support from the University of Basel,

J. Burggraaff received travel support from Novartis Pharma AG,

P. Kontschieder is an employee of Microsoft Research,

J. Dorn is an employee of Novartis Pharma AG,

Ch. P. Kamm received honoraria for lectures and consulting from Biogen idec, Novartis, Teva, Merck-Serono, Genzyme and Bayer Schweiz AG,

S. Steinheimer: nothing to disclose,

P. Tewarie received travel support from Novartis Pharma AG, C.Morrison is an employee of Microsoft Research,

A. Sellen is an employee of Microsoft Research,

B. M. J. Uitdehaag received consultation fees from: Biogen Idec, Novartis, EMD Serono, Teva Pharmaceuticals, Genzyme and Roche. The Multiple Sclerosis Centre Amsterdam has received financial support for research activities from Biogen Idec, Merck Serono, Novartis, and Teva Pharmaceuticals,

A. Criminisi is an employee of Microsoft Research,

F. Dahlke is an employee of Novartis Pharma AG, Ludwig Kappos's institution, the University Hospital Basel, has received research support and payments that were used exclusively for research support for Prof Kappos' activities as principal investigator and member or chair of planning and steering committees or advisory boards in trials sponsored by Actelion, Addex, Bayer Health Care Pharmaceuticals, Bayer Schering Pharma, Biogen Idec, CLC Behring, Genentech, GeNeuro SA, Genzyme, Merck Serono, Mitsubishi Pharma, Novartis, Octapharma, Praxicon, Roche, Sanofi-Aventis, Santhera, Siemens and Teva; royalties from Neurostatus GmbH; research grants from the Swiss MS Society, Swiss National Research Foundation, the European Union, Gianni Rubatto Foundation, and the Novartis and Roche Research foundations.

Abstract: P818

Type: Poster

Abstract Category: Clinical assessment tools

Background: Clinical assessment of impairment and disability in Multiple Sclerosis (MS) remains the most important outcome in therapeutic trials, and is commonly assessed with the Expanded Disability Status Scale (EDSS). However, the EDSS exhibits high inter- and intra-rater variability. The ASSESS MS system is being developed as a non-invasive, more consistent and potentially finer grained tool to measure motor dysfunction in MS, by combining recordings of prescribed neurological movements with machine learning methods to assess motor dysfunction based on EDSS subscores.

Objectives: To test the prediction of EDSS subscores from recordings of a depth-sensing video analysed by machine learning algorithms.

Methods: Pre-defined movements from the EDSS assessment were recorded in 300 patients and 200 healthy volunteers. Video recordings of patients were scored by four neurologist from 3 sites based on the Neurostatus/EDSS assessment definitions. These scores were used to train a machine learning algorithm to correctly predict motor dysfunction from depth-sensing video recordings, which capture a 3D view of the patient.

Results: We report that on movements covering upper extremities and trunk, the machine learning algorithm predicts motor dysfunction of patients with MS with an accuracy similar to neurologists' intra-rater retest reliability. For example, the agreement of the upper extremity tremor/dysmetria subscore from the finger-to-nose test of the algorithm with the neurologists' assessment is 73% across scores 0, 1, 2 and 3, while the long-term intra-rater agreement of the neurologists with their previous assessment is 67% in this challenging-to-assess range.

Conclusions: Automated quantification of movement recordings using a depth-sensing camera and image analysis based on a machine-learning algorithm enables an accurate and sensitive quantitative assessment of motor dysfunction in MS patients. ASSESS MS is expected to improve the evaluation of disability progression in clinical studies and clinical practice.

Disclosure:

M. D`Souza received travel support from Bayer AG, Teva and Genzyme and research

support from the University of Basel,

J. Burggraaff received travel support from Novartis Pharma AG,

P. Kontschieder is an employee of Microsoft Research,

J. Dorn is an employee of Novartis Pharma AG,

Ch. P. Kamm received honoraria for lectures and consulting from Biogen idec, Novartis, Teva, Merck-Serono, Genzyme and Bayer Schweiz AG,

S. Steinheimer: nothing to disclose,

P. Tewarie received travel support from Novartis Pharma AG, C.Morrison is an employee of Microsoft Research,

A. Sellen is an employee of Microsoft Research,

B. M. J. Uitdehaag received consultation fees from: Biogen Idec, Novartis, EMD Serono, Teva Pharmaceuticals, Genzyme and Roche. The Multiple Sclerosis Centre Amsterdam has received financial support for research activities from Biogen Idec, Merck Serono, Novartis, and Teva Pharmaceuticals,

A. Criminisi is an employee of Microsoft Research,

F. Dahlke is an employee of Novartis Pharma AG, Ludwig Kappos's institution, the University Hospital Basel, has received research support and payments that were used exclusively for research support for Prof Kappos' activities as principal investigator and member or chair of planning and steering committees or advisory boards in trials sponsored by Actelion, Addex, Bayer Health Care Pharmaceuticals, Bayer Schering Pharma, Biogen Idec, CLC Behring, Genentech, GeNeuro SA, Genzyme, Merck Serono, Mitsubishi Pharma, Novartis, Octapharma, Praxicon, Roche, Sanofi-Aventis, Santhera, Siemens and Teva; royalties from Neurostatus GmbH; research grants from the Swiss MS Society, Swiss National Research Foundation, the European Union, Gianni Rubatto Foundation, and the Novartis and Roche Research foundations.

By clicking “Accept Terms & all Cookies” or by continuing to browse, you agree to the storing of third-party cookies on your device to enhance your user experience and agree to the user terms and conditions of this learning management system (LMS).

Cookie Settings
Accept Terms & all Cookies