Motor unit number estimation (MUNE) for the assessment of neuromuscular function

1. A method for assessing neuromuscular function by estimating motor unit numbers, the method comprising:

providing an electrode array for delivering one or more controlled stimuli to a subject using at least one stimulator electrode and for acquiring electric activity signals of one or more motor units of a subject using at least one detector electrode;

positioning said electrode array on a subject, delivering a stimulus of varying characteristics to the subject and acquiring a set of electric activity signals of one or more motor units of the subject;

identifying and eliminating any acquired signals due to alternation;

automatically estimating motor unit number using an automation algorithm,

wherein the automation algorithm consists of:

(i) pre-processing acquired responses so as to consolidate and enhance the acquired signals; and

(ii) identifying plausible new motor unit activation events and corresponding response changes; and

(iii) determining representatives of single motor unit potential features, and computing approaches; and

(iv) estimating motor unit number and distribution statistics associated with the motor unit number;

wherein pre-processing the acquired signals comprises:

(a) estimating background noise level; and

(b) determining motor unit activity region in the acquired signal; and

(c) ordering waveforms according to activity level, and removing noise-only waveforms; and

(d) identifying and combining waveforms with insignificant morphological variations in relation to background noise;

wherein a background noise level is determined by:

(a) using waveform segments known to have no motor unit activity; and

(b) estimating statistics of noise level, the estimated statistics comprising at least one of mean absolute deviation, standard deviation, and mean square error.

2. A method according to claim 1 wherein the electrode array consists of individual electrodes placed independently on a subject according to anatomical landmarks.

3. A method according to claim 1 wherein the electrode array consists of electrodes whose positions, relative to each other, are fixed for at least some of the electrodes.

4. A method according to claim 1 wherein the electrode array consists of one or more pairs of stimulator electrodes, wherein the relative positions of the stimulator electrodes are selected from the group consisting of: (i) fixed, and (ii) adjustable.

5. A method according to claim 1 wherein the varying characteristics of the stimuli are controlled and adjusted according to one selected from the group consisting of: (i) directly and manually by a user, and (ii) automatically by a computer algorithm.

6. A method according to claim 5 wherein the characteristics of the stimuli include at least one selected from the group consisting of: duration, intensity, polarity, frequency, and other features of electric current to be delivered via stimulus electrodes.

7. A method according to claim 5 wherein the control and adjustment of the stimuli is based on responses acquired from the detector electrodes.

8. A method according to claim 7 wherein responses acquired from the detector electrodes are analyzed for features, wherein the features include at least one selected from the group consisting of: onset, total duration, peak duration, peak amplitude, and peak area.

9. A method according to claim 8 wherein said features also include at least one from the group consisting of: an absolute value of the features, and changes in the absolute value from response to response.

10. A method according to claim 1 wherein an activity-free segment is identified as:

(a) a segment of waveform acquired with zero or negligible stimulation strength; and/or (b) a segment of waveform occurring before CMAP activity.

11. A method according to claim 1 wherein a motor unit activity region is determined by:

(a) calculating a median value for all acquired signals; and

(b) determining a deviation of each individual signal from a calculated median value for each time instance; and

(c) constructing an activity profile by averaging the deviations over all of the signals; and

(d) finalizing an activity region by comparing an activity profile with background noise level.

12. A method according to claim 1 wherein waveforms are windowed with the activity region, and windowed waveforms are used for:

(a) calculating a power level of each waveform within an activity region;

(b) ordering all waveforms according to the power levels of the waveforms; and

(c) removing noise-only waveforms if the power level of the waveforms is below a noise threshold.

13. A method according to claim 1 wherein waveform variations are quantified by a similarity measure based on at least one selected from the group consisting of: Euclidian distance, correlation, mean absolute difference, and other comparable measure.

14. A method according to claim 1 wherein waveforms with insignificant morphological variations are:

(a) identified when e similarity measures of the waveforms are above a threshold value related to noise level; and

(b) combined to yield a combined waveform by means of at least one selected from the group consisting of: arithmetic averaging, weighted averaging, other linear combination approaches and nonlinear combination approaches.

15. A method according to claim 1 wherein identifying plausible new motor units from a collection of acquired signals comprises:

(a) optimizing a global search of motor units; and/or

(b) identifying alternation motor unit responses directly.

16. A method according to claim 15 wherein optimizing a global search of true motor units comprises:

(a) establishing a ranking criteria for quantitative evaluation of SMUP waveforms at least one of individually, and in groups; and

(b) establishing a global search strategy to extract SMUPs; and

(c) extracting SMUPs; and

(d) optimizing the extracted SMUPs to determine solutions.

17. A method according to claim 16 wherein an SMUP ranking system is established based upon a consideration part of some or all of the features of at least one from the group consisting of: SMUP responses, CMAP responses, response power, response onset, response end, response duration, response turn, response phase, response area, response amplitude, response peaks, and response peak-distance.

18. A method according to claim 16 wherein a SMUP group is defined as one possible combination set of SMUPs extracted from consolidated MU waveforms.

19. A method according to claim 17 wherein an SMUP is scored by:

(a) calculating an offset of a SMUP waveform power over that of the average level of an SMUP group; and

(b) calculating an offset of a SMUP waveform onset over that of a CMAP; and

(c) calculating an offset of a SMUP waveform duration over that of a CMAP; and

(d) calculating an offset of a SMUP waveform maximum negative peak location over that of a CMAP; and

(e) calculating an offset of waveform numbers of the SMUP group over a total number of consolidated MU waveforms; and

(f) weighting, and then summing, all of the offsets so as to generate an SMUP waveform score.

20. A method according to claim 16 wherein a global search strategy is established by:

(a) determining a decision deferment level; and

(b) extracting new SMUPs and considering consolidated MU waveforms both with alternation and without alternation issues at a pre-determined deferment level.

21. A method according to claim 20 wherein a decision deferment level is determined as a number that represents a maximum number of alternating MUs, with overlapping thresholds.

22. A method according to claim 21 wherein the decision deferment level is determined so as to be a value between (i) one, which represents step-by-step sequential manual processing, and (ii) a number of total consolidated waveforms, which considers all possible alternation cases in the acquired responses.

23. A method according to claim 20 wherein new SMUPs are extracted at pre-set decision deferment level by:

(a) considering direct subtraction of sequential consolidated MUs waveforms; and (b) using a MUs waveform to subtract a base waveform; and

(c) repeating steps (a) and (b) until all the alternation possibilities are exhausted.

24. A method according to claim 23 wherein a base waveform is determined as a consolidated MU waveform based on an alternation consideration at the pre-set decision deferment level.

25. A method according to claim 24 wherein total alternation cases at a pre-set decision deferment level are defined as the power of 2 with an exponent that equals the pre-set decision deferment level −1.

26. A method according to claim 16 wherein SMUPs are optimized by:

(a) scoring all the SMUP waveforms at least one of individually and in groups;

(b) identifying a group of SMUPs with highest scores as true SMUPs; and/or

(c) allowing several alternative solutions to co-exist.

27. A method according to claim 26 wherein several alternative solutions represent solutions with SMUP sets that meet the ranking criteria and have high scores.

28. A method according to claim 15 wherein identifying MU alternation directly comprises:

(a) searching alternation equations to identify MUs alternation range and avoid direct subtraction of alternating MUs waveforms; and

(b) extracting potential SMUPs using a combination of IS method and alternation equations; and

(c) validating potential SMUPs; and

(d) identifying a SMUPs solution.

29. A method according to claim 28 wherein alternation equations are established by:

(a) determining a base waveform, from a null signal then first to last consolidated MU waveforms; and

(b) subtracting the base waveform from all the consolidated MU waveforms at a higher level to obtain residual waveforms; and

(c) using any combination of residual waveforms to search an identical one in the same residual waveform group (i.e., any identified waveforms satisfying an equation, namely alternation equation).

30. A method according to claim 29 wherein an MU waveform at a higher level represents a consolidated MU waveform that partly or fully has larger features, wherein the features comprise at least one from the group consisting of: power, areas, amplitude, and duration.

31. A method according to claim 29 wherein an identical waveform is measured using a similarity measurement.

32. A method according to claim 28 wherein SMUPs are extracted by:

(a) identifying alternation equations, if any, until all possibilities are exhausted;

(b) selecting smaller waveforms from alternation equations as potential SMUPs; or (c) using direct subtraction of sequential MUs waveforms as potential SMUPs, beyond an alternation range in a consolidated MU waveform group.

33. A method according to claim 28 wherein SMUPs are validated by:

(a) self-checking to identify any MU waveforms that do not satisfy a matching criteria compared with SMUPs' combinations; and

(b) extracting new SMUPs from MU waveforms that do not satisfy a matching criteria.

34. A method according to claim 33 wherein self-checking uses SMUPs' combinations to compare with consolidated MU waveforms, with criteria being satisfied if a consolidated MU waveform is identical to one SMUPs' combination; otherwise, new information exists in the MU waveforms that do not meet criteria.

35. A method according to claim 33 wherein new SMUPs are extracted from the MU waveforms that do not meet criteria by:

(a) using an MU waveform to subtract lower level MU waveforms sequentially to obtain extra waveforms; and

(b) using an obtained extra waveform for self-checking; if the extra waveform in combination with existing potential SMUPs satisfies criteria for two or more MU waveforms, the extra waveform is added to the potential SMUP group.

36. A method according to claim 28 wherein a SMUPs solution is identified by:

(a) identifying and removing identical waveforms or any composite SMUP waveforms from a potential SMUP group; and

(b) selecting a set of SMUPs from the potential SMUP group, which has a minimum number of SMUPs, but satisfies all consolidated MUs waveform criteria.

37. A method according to claim 1 wherein representative SMUP features are determined to describe SMUPs' characteristics, which include SMUPs' amplitude and/or areas.

38. A method according to claim 1 wherein computing approaches are determined to calculate sizes of SMUPs features, which include methods of averaging, and/or weighting SMUPs' feature values.

39. A method according to claim 1 wherein motor unit numbers are estimated by using a SMUP representative value divided into a feature value of a maximal CMAP.

40. A method according to claim 1 wherein motor unit number statistics, including motor unit number mean value and standard deviation value, are calculated by using motor unit number estimation alternation solutions.