an implantable system for angles measurement in

Instrumented Prosthesis for Knee Implants Monitoring A. Arami, M. Simoncini, O. Atasoy, W. Hasenkamp, S. Ali, A. Bertsch, E. Meurville, S. Tanner, H. Dejnabadi, V. Leclercq, P. Renaud, C. Dehollain, P.A. Farine, B. M. Jolles, K. Aminian and P. Ryser Ecole Polytechnique Fédérale de Lausanne (EPFL), CH Symbios Orthopédie SA, CH University Hospital of Lausanne (CHUV), CH

Outline • Background • Objective • Methods and Material – Measurement system – Electronic system – Validation in large scale demonstrator

• Results • Discussion and conclusion Instrumented Prosthesis for Knee Implants Monitoring

2

Background • Total knee arhtroplasty: one of the most common treatment in knee injuries and diseases • The necessity to: – Assess the surgery – Evaluate the functionality of different prostheses – Anticipate the need of revision

Instrumented Prosthesis for Knee Implants Monitoring

3

Background Measure kinetics and kinematics of joints, conventional lab. measurement system: Camera-marker based, force plate Inertial sensors

Physilog® 3D gyro-accelero


4

Background Measure kinetics and kinematics of joints, conventional lab. measurement system: Camera-marker based, force plate Inertial sensors

Physilog®

• Drawbacks

3D gyro-accelero

– Impossibility of measurement of the contact force an moments – Soft tissue artifact (STA)

Recently x-ray measurement systems Instrumented Prosthesis for Knee Implants Monitoring

5

Previous works (instrumented prosthesis): mostly changed the design of the prosthesis Focus only on force and moments

[Graichen, Arnold, Rohlmann, Bergmann 2007]

For joint function evaluation, kinematic measurement is also necessary. We need highly accurate STA free measurements: Accuracy of 1degree for different rotations


6

Objective Devise an implantable measuring system into knee prosthesis: – Design the Measurement systems • Kinematics (STA free and accurate) • Kinetics (actual)

– Design Analog front end for sensor read out (low power) – Design Remote Powering system (efficient)


7

Objective Remote powering and communication

Sensors for kinematics Amp. and ADC

AA and FE angles Processing

Sensors for kinetics


Force and imbalance

8

Project divisions Communication Remote powering

Packaging

Surgical implantation

Force sensors

Biomechanical modeling

Electronics Sensors interface 9

Method & Material AA

• The F.I.R.S.T knee prosthesis (Symbios) was used. • Sensors and electronics in Polyethylene

Femoral Polyeth.

Tibial

– Almost no change in design of other parts – Flexibility of design – Sensor and electronics powering and communication issue


10

Kinematics measurement Type of sensors-Number of sensors-ConfigurationMapping algorithm • Type of sensors Magnet Magnetic sensors – Transparency of body and the prosthesis towards magnetic flux – No need for an active node in Femoral part (a magnet as a passive source)

AMR Sensors

– Low power (possibility of duty cycling) – No considerable drift Instrumented Prosthesis for Knee Implants Monitoring

11

Kinematics measurement • Number of sensors? Pos. & Ori. of a magnet

Biomechanical constraints

5 parameters to estimate

2~3 parameters to estimate


Min number of inf. Sources or sensors one 2D ~ two 2D or one 3D

12

Kinematics measurement • Where to put sensors? A 3-sensor configuration: To have high sensitivity in FE and AA 2 sensors for each rotation (Overall: 3 sensors)


13

Kinematics measurement • Mapping algorithm • ANN is utilized in absence of unifying model for magnet and AMR sensors. Magnetic sensor1 sensor2 sensor3

Preprocessing

Reference system (Vicon)

Artificial Neural Network

Estimated trajectory of markers

+-

A model-based angle estimator

Ref. Angles

Estimated Angles Trajectory Estimation error

Training phase (75% of randomly selected data) Test phase (25% of remaining randomly selected data)

14

Kinetics measurement • Force imbalance in the medial-lateral prosthesis’s condoyle • Total contact force

Type of sensors-Configuration • Type of sensors Strain gauges made of biocompatible materials Sensors were connected to 2 independent Wheatstone bridges (quarter-bridge conf.) • Configuration The design based on finite element analysis Instrumented Prosthesis for Knee Implants Monitoring

15

Analog front end design • Amplifier gain = 3000 • Offset calibration • Temperature calibration Amplifier Output Noise ADC Quantization Noise

ADC Constraints

ADC resolution g m = 100 µS G = 3000

To have an ADC with - high resolution - inherent anti aliasing filtering - robustness to analog imperfections

}

Vno = 60 µV / Hz n = 10

SDM-based ADC is designed (BW:10kHz, SNDR>64dB)

[1] S.Ali, S.Tanner, P-A.Farine “A Novel 1V, 24uW, Sigma Delta Modulator using Amplifier and Comparator based technique, with 64.7dB SNDR and 10KHz Bandwidth”, IEEE Int. Conf. on Circuits and Systems, December 2010

16

Remote Powering Commands & Power

Modulator

Power Amplifier

L1

L2

Demodulator

Rectifier

Regulator

Sensor Data

Reader

Demodulator

Voltage Doubler

Remote Electronics

High V Regulator

Sensors

Modulator

External Part

Internal Part

Reader Antenna Implanted Antenna Image source: http://healthtopics.hcf.com.au/ TotalKneeReplacement.aspx


[Atasoy, Prime2010] 17

Remote Powering Commands & Power

Modulator

Power Amplifier

L1

L2

Demodulator

Rectifier

Regulator

Sensor Data

Reader

Demodulator

Voltage Doubler

Remote Electronics

High V Regulator

Sensors

Modulator

External Part

Internal Part

Reader Antenna Implanted Antenna Image source: http://healthtopics.hcf.com.au/ TotalKneeReplacement.aspx


[Atasoy, Prime2010] 18

Validation in Large Scale Demonstrator A knee mechanical simulator was designed: to simulate contact forces and rotations. Optical motion capture (Vicon, UK)


19

Validation in Large Scale Demonstrator A knee mechanical simulator was designed: to simulate contact forces and rotations.

• A microcontroller was programmed: data acquisition & transmission • The signals are transmitted to an external reader by a Passive Low-Frequency RFID Transponder Interface Instrumented Prosthesis for Knee Implants Monitoring

20

Validation in Large Scale Demonstrator A knee mechanical simulator was designed: to simulate contact forces and rotations.

Force meter – adjustable static force Instrumented Prosthesis for Knee Implants Monitoring

21

Results (kinematics) 40

30

Abduction angle(degree) Abduction angle

Flexion angle Flexion angle (degree)

35

2 Estimated Flexion Angle Reference Flexion Angle

25 20 15 10 5 0 -5

1 0

Estimated Abduction Angle Reference Abduction Angle

-1 -2 -3 -4 -5 -6 -7

-10 4200 21 4300 4400 22 4500 4600 23 4700 4800 24 4900 5000 25 5100 5200 26 Sample Time(s)

-8

4200 21 4300 4400 22 4500 4600 23 4700 4800 24 4900 5000 25 5100 5200 26 Sample Time(s)

Estimation Error (test)

Joint Angle Angle

E(error)

STD(error)

rms(error)

Max(|error|)

FE

-1.54⁰

1.77⁰

2.35⁰

5.11⁰

0.23⁰

0.20⁰

0.31⁰

0.93⁰

ROM: 52⁰

AA ROM: 12⁰


22

Results (kinetics) A preliminary calibration using demonstrator. • The linear fit based LSE criteria: (R2: 0.81, rms Error: 0.2). • The fit is not perfect due to  Leakage of force from other parts of PE  Nonlinear behavior of PE

1 Instrumented Prosthesis for Knee Implants Monitoring

23

Conclusion • An instrumented prosthesis is designed to measure kinematics and kinetics. • The kinematic measurements for two movement is validated toward Vicon, UK system. • Force sensors are calibrated using the mechanical simulator. • The simulation results showed the efficiency of some of designed parts of electronics and remote powering subsystems. • The system is easily adaptable to other prosthesis


24

Perspectives • Further investigation to measure also the laxity and internal external rotations • A new force sensor to also measure an estimated knee moments • Finalizing the large scale demonstrator : – all internal electronic parts on a unite board, – manual knee simulator to a programmable robotic simulator (realistic knee motioncadaver based studies) Instrumented Prosthesis for Knee Implants Monitoring

25

MTS knee simulator

Acknowledgement. The nano-tera for funding this project(SNF20NAN1_123630).

Thanks for your attention. Any question?


26