Correcting a robotic surgery user interface device tracking input
A sequence of tracking input samples that are measures of position or orientation of a user interface device, UID, being held by a user, are received. In a prediction phase, a current output sample of a state of linear quadratic estimator, LQE, is computed that is an estimate of the position or orientation of the UID. The current output sample is computed based on i) a previously computed output sample, and ii) a velocity term. In an update phase, an updated output sample of the state of the LQE is computed, based on i) a previously computed output sample from the prediction phase, and ii) a most recent tracking input sample. Other embodiments are also described and claimed.
1. A method for correcting a user interface device, UID, tracking input in a robotic surgery system, the method comprising:
receiving a sequence of tracking input samples that are measures of position or orientation of a user interface device, UID, in a robotic surgery system;
in a prediction phase of a linear quadratic estimator, LQE, computing a current output sample of a state of the LQE that is an estimate of the position or orientation of the UID, wherein the current output sample is computed based on i) a previously computed output sample, and ii) a velocity term, wherein computing the current output sample comprises fitting a line to a plurality of previously computed output samples that are within a window and the velocity term comprises a slope of the line; and
in an update phase of the LQE, computing an updated output sample of the state of the LQE, based on i) a previously computed output sample from the prediction phase, and ii) a tracking input sample.
2. The method of claim 1 wherein the current output sample is computed based on a tuning parameter that adjusts a sample-to-sample time interval used in the prediction phase.
3. The method of claim 1 wherein in the update phase of the LQE, computing the updated output sample of the state of the LQE comprises computing a Kalman filter gain.
4. The method of claim 1 wherein the sequence of tracking input samples has a sampling rate that is lower than that of computing the output samples of the state of the LQE in the prediction phase.
5. The method of claim 1 wherein the UID tracking input comprises an orientation vector, and the state of the LQE is an estimate of the orientation of the UID.
6. The method of claim 1 further comprising in the prediction phase:
computing a next output sample of the state of the LQE based on i) another previously computed output sample, and ii) another velocity term, by
moving the window forward so that another plurality of previously computed output samples are within the window; and
repeating a) and b) with the moved window to compute said another velocity term.
7. The method of claim 1 wherein the sequence of tracking input samples is processed as a live stream of data and is processed in real time.
8. The method of claim 2 wherein the sample-to-sample time interval used in the prediction phase comprises a sum of i) the tuning parameter and ii) a nominal output sampling period of the LQE.
9. The method of claim 6 wherein the next output sample is computed based on a tuning parameter that adjust a sample-to-sample time interval used in the prediction phase.
10. An article of manufacture comprising:
a machine-readable medium having stored therein instructions that when executed by a processor correct a user interface device, UID, tracking input, by producing an output sequence that is an estimate of the position or the orientation of the UID, wherein the processor is to
receive a sequence of tracking input samples that are measures of position or orientation of a user interface device, UID, in a robotic surgery system,
in a prediction phase of a linear quadratic estimator, LQE, compute a current output sample of a state of the LQE that is an estimate of the position or orientation of the UID, wherein the current output sample is computed based on i) a previously computed output sample, and ii) a velocity term, wherein the processor computes the current output sample by fitting, a line to a plurality of previously computed output samples that are within a window and the velocity comprises a slope of the line, and
in an update phase of the LQE, compute an updated output sample of the state of the LQE, based on i) a previously computed output sample from the prediction phase, and ii) a tracking input sample.
11. The article of manufacture of claim 10 wherein the current output sample is computed based on a tuning parameter that adjusts a sample-to-sample time interval used in the prediction phase.
12. The article of manufacture of claim 10 wherein in the update phase of the LQE, the processor computes the updated output sample of the state of the LQE by computing a Kalman filter gain.
13. The article of manufacture of claim 10 wherein the sequence of tracking input samples has a sampling rate that is lower than that of computing the output samples of the state of the LQE in the prediction phase.
14. The article of manufacture of claim 10 wherein the UID tracking input comprises an orientation vector, and the state of the LQE is an estimate of the orientation of the UID.
15. The article of manufacture of claim 10 , wherein in the prediction phase the processor:
computes a next output sample of the state of the LQE based on i) another previously computed output sample, and ii) another velocity term, by
moving the window forward so that another plurality of previously computed output samples come within the window; and
repeating a) and b) with the moved window to compute said another velocity term.
16. The article of manufacture of claim 10 wherein the sequence of tracking input samples is processed as a live stream of data and is processed in real time.
17. The article of manufacture of claim 11 wherein the sample-to-sample time interval used in the prediction phase comprises a sum of i) the tuning parameter and ii) a nominal output sampling period of the LQE.
18. The article of manufacture of claim 15 wherein the next output sample is computed based on a tuning parameter that adjust a sample-to-sample time interval used in the prediction phase.