Wearables from Lab to Bedside: The Stanford Wearable Health Lab Story by Matthew Smuck, MD, Chief, Physical Medicine & Rehabilitation, Associate Professor, Orthopaedics, Director, Wearable Health Lab, Stanford University. Presented at the IoT, Big Data Healthcare Summit Western Canada on February 1, 2017.

1. Wearables from Lab to Bedside:
The Stanford Wearable Health Lab Story
www.iotevents.ca -# Io T Healthc areW
Matthew Smuck, MD
Chief, Physical Medicine & Rehabilitation
Associate Professor, Orthopaedics
Director, Wearable Health Lab
Stanford University
IoT, Big Data Healthcare Summit
Feb 1, 2017

2. Thanks to!
Wearable Health Lab
FACULTY AND STAFF
 Matthew Smuck, MD (Director)
 Christy Tomkins-Lane, PhD
 Agnes Martinez-Ith
 Ming-Chi Kao, MD, PhD
STUDENTS & POST-DOCS
 Justin Norden
 Aman Sinha
 Vibhu Agarwal
 Amir Muaremi, PhD
 Patricia Zheng, MD
COLLABORATIONS WITH:
 Scott Delp, PhD (Stanford Mobilize Center)
 William Haskell, PhD (Stanford Prevention Center)
 Nigam Shah, PhD (Stanford Bioinformatics)
Supported by: NIH U54EB020405 Stanford Mobilize Center

3. PRECISION MEDICINE –
http://med.stanford.edu/precisionhealth.html

4. PRECISION MEDICINE –
Definition:

5. PRECISION MEDICINE –
Definition:
The right treatment.

6. PRECISION MEDICINE –
Definition:
The right treatment.
For the right person.

7. PRECISION MEDICINE –
Definition:
The right treatment.
For the right person.
At the right time.

8. WEARABLES & PRECISION MEDICINE –
Zio-Patch
http://www.irhythmtech.com/patients/how-it-works

9. MY MOTIVATION –
THE BIGGEST PROBLEM IN ORTHOPEDICS
RESEARCH

10. MY MOTIVATION –
THE BIGGEST PROBLEM IN ORTHOPEDICS
RESEARCH
Imprecise & Subjective Outcomes

11. PROBLEM
Imprecise & Subjective Outcomes
As a result, we lack a common language to discuss
with patients the context and consequences of treatment
choices

12. The exception that proves the rule.
Reliance on Subjective Outcomes

13. What we communicate to patients is
defined by what we measure

14. What we communicate to patients is
defined by what we measure
THE SOLUTION STARTS WITH

15. Outcomes Measurement
Pain
Human Function

16. MOTIVATION –
http://www.everymac.com/systems/apple/iphone/iphone-faq/differences-between-the-original-iphone-and-iphone-3g.html

17. SOLUTIONS FROM MOBILE WEARABLE
TECHNOLOGY
 PHYSICAL ACTIVITY MONITORING
• Problem Solved ??
– A ubiquitous, objective FUNCTIONAL outcomes measure

18. Surprising Fact
Current activity monitors are useless in my
clinic

19. PHYSICAL ACTIVITY MONITORING IN MSK
RESEARCH
Results are counterintuitive – WHY?

20. Physical Performance Analysis:
A new approach to assessing free-living physical activity in
musculoskeletal pain and mobility limited populations
Kao MCJ, Jarosz R, Tomkins-Lane C, Smuck M. Publication Pending.
Existing Intervals (VO2 based)
STANDARD MEASUREMENT OF PHYSICAL ACTIVITY
– Steps
– Energy expenditure

21. Physical Performance Analysis:
A new approach to assessing free-living physical activity in
musculoskeletal pain and mobility limited populations
Kao MCJ, Jarosz R, Tomkins-Lane C, Smuck M. Publication Pending.
RESULTS
 Empirically derived novel intervals for analysis of physical performance
Physical Performance Intervals
Existing Intervals (VO2 based)

22. - First applications of this novel tool
- Measure differences between spinal stenosis subjects &
controls
- Evaluate post-op changes in PHYSICAL PERFORMANCE
Activity Monitoring with Accelerometry Outperforms Self-
Reported and Laboratory Assessments of Function in
Patients with Lumbar Spinal Stenosis
Smuck M, Buman M, Ith MA, Haskell WL, Kao MCJ. PM&R. 2013 Sept:5(9):S292.
Lumbar Spinal Stenosis Decompression Normalizes
Free-Living Physical Activity Impairment
Smuck M, Buman M, Ith MA, Haskell WL, Kao MCJ. Spine J. 2013 Sept;13(9S):S41-S42.

23. Phenotypes of Physical Performance for
Knee Osteoarthritis and Lumbar Spinal Stenosis
Norden J, Singah A, Ith MA, Tomkins-Lane C, Smuck M.
LSS vs. NHANES OAI vs. NHANES LSS vs. OAI
# of Features
in Sparse Set 4 9 10
Classification
Accuracy* 0.80 0.83 0.72
LEVERAGING 3 EXISTING DATASETS
• Knee OA (OAI)
• Lumbar Spinal Stenosis (LSS)
• Population control (NHANES)
RESULTS

24. BENEFICIAL SIDE EFFECTS
- New insights into disease mechanism

25. WHAT HAVE WE LEARNED FROM THIS ?

26.  Biosensor data analysis has to match the clinical
condition
WHAT HAVE WE LEARNED FROM THIS ?

27.  Biosensor data analysis has to match the clinical
condition
WHAT HAVE WE LEARNED FROM THIS ?
 Solution to Problem – Access to an objective and
quantitative functional outcomes measure (physical
performance) using accelerometry

28.  Biosensor data analysis has to match the clinical
condition
WHAT HAVE WE LEARNED FROM THIS ?
 Solution to Problem – Access to an objective and
quantitative functional outcomes measure (physical
performance) using accelerometry
 Solution to Problem – Opportunity for meaningful context
in the clinical conversation

29. WHAT HAVE WE LEARNED FROM THIS ?
 Solution to Problem – Opportunity for meaningful context
in the clinical conversation
 Performance Phenotypes are digital biomarkers that
can be used to develop predictive models for Precision
Health
 Biosensor data analysis has to match the clinical
condition
 Solution to Problem – Access to an objective and
quantitative functional outcomes measure (physical
performance) using accelerometry

30. Thank You!
Matthew Smuck, MD
Chief, Physical Medicine & Rehabilitation
Associate Professor, Orthopaedics
Director, Wearable Health Lab
msmuck@stanford.edu