2. Project Overview Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work Our product: Rehabilitation device  Recumbent stationary bicycle Electromyography of quadriceps muscles  Feedback User-friendly interface  Autonomous recovery Our customers: Post-ACL repair patients  Phase II and III of rehabilitation  Assist in at-home exercises

3. 1) EMG signals differ between patellofemoral pain syndrome patients and control Cowan SM, et al. Arch Phys Med Rehabil. (2001) 82:183-189. 2) Literature on EMG acquisition on bicycles Garrett WE, Kirkendall DT. Exercise and Sport Science. Lippincott Williams & Wilkins. (2000) 3) Quadricep EMG signals differ between ACL patients and control when cycling Hunt MA, et al. Clinical Biomechanics. (2003) 18: 393-400. Clinical Relevance Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work

4. 1) EMG signals differ between patellofemoral pain syndrome patients and control Cowan SM, et al. Arch Phys Med Rehabil. (2001) 82:183-189. 2) Literature on EMG acquisition on bicycles Garrett WE, Kirkendall DT. Exercise and Sport Science. Lippincott Williams & Wilkins. (2000) 3) Quadricep EMG signals differ between ACL patients and control when cycling Hunt MA, et al. Clinical Biomechanics. (2003) 18: 393-400. Clinical Relevance Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work

5. 1) EMG signals differ between patellofemoral pain syndrome patients and control Cowan SM, et al. Arch Phys Med Rehabil. (2001) 82:183-189. 2) Literature on EMG acquisition on bicycles Garrett WE, Kirkendall DT. Exercise and Sport Science. Lippincott Williams & Wilkins. (2000) 3) Quadricep EMG signals differ between ACL patients and control when cycling Hunt MA, et al. Clinical Biomechanics. (2003) 18: 393-400. Clinical Relevance Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work

6. I.Signal Acquisition A.Collect EMG data while cycling B.Correlate crank angle with EMG signal II.Signal Processing A.Filtering noise B.Algorithms to analyze signal C.Developing user-friendly interface III.Testing A.Protocol optimization B.Proof of concept Design Goals Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work

7. Components and Setup Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work

8. EMG Electrode Placement Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work Cowan SM, et al. Arch Phys Med Rehabil. (2001) 82:183-189. Electrode placement on the quadriceps muscles

9. Hall Effect Sensor Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work 90 o 180 o 270 o 0o0o Magnets

10. Voltage (volts) EMG Acquisition Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work Time (ms) VMO, ~50 rpm, Hall effect sensor at 180 o

11. Voltage (volts) EMG Acquisition Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work Time (ms) VMO, ~40 rpm, Hall effect sensor at 0 o and 180 o

12. The Butterworth Filter Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work

13. Unfiltered Data Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work

14. Filtered Data Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work Cowan SM, et al. Arch Phys Med Rehabil. (2001) 82:183-189.

15. Integrated Data Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work Garrett WE, Kirkendall DT. Exercise and Sport Science. Lippincott Williams & Wilkins. (2000)

16. Experiments Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work ExperimentTest Minimum muscle warm-up time for repeatable EMG signals With warm-up prescribed from Cowan SM, et al. (2001) vs. no warmup Effect of changing electrode position on EMG signal Change electrode placement at set distances Effect of individual differences on EMG signals for non-injured volunteers Measure EMG signals amongst several people Minimum resistance/speed for even pedaling AND VMO/VL stimulation Increase the resistance (e.g. low, medium, high settings)

17. 1.Develop a user-friendly GUI 2.Correlate position with crank angle more precisely 3.More signal processing for de-noising 4.Process the signal and plot as a function of crank angle or time 5.Proof of concept studies with current recovering ACL patients Future Work Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work

18. Bloopers Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work

19. Acknowledgements Overview | Relevance | Design Goals | Signal Acquisition | Signal Processing | Future Work Dr. William Macaulay Orthopedic Surgeon, Columbia University Medical Center Director of Center for Hip and Knee Replacement Dr. Ranjan Gupta Department Chair of Orthopaedic Surgery, UC Irvine Professor of Orthopaedics, Anatomy & Neurobiology, and BME James Gossett Associate Athletic Director, Columbia University Dr. Evan Johnson Director of Physical Therapy at the Spine Center Administrative Director of the Spine Center Julianne Costa Occupational Therapist Registered Physical Therapist Dr. Tim Wright Orthopaedic Biomechanics and Biomaterials Hospital of Special Surgery Dr. Clark Hung Associate Professor of Biomedical Engineering Dr. Gordana Vunjak-Novakovic Professor of Biomedical Engineering Dr. Paul Sajda Associate Professor of Biomedical Engineering Dr. Elizabeth Hillman Assistant Professor of Biomedical Engineering Keith Yeager Senior Staff Associate, Laboratory Manager Sean Burgess Teaching Assistant Robert Maidhof Teaching Assistant Viktor Gamarnik BME Senior, SMArtView