A new myohaptic instrument to assess wrist motion dynamically

Abstract

The pathophysiological assessment of joint properties and voluntary motion in neurological patients remains a challenge. This is typically the case in cerebellar patients, who exhibit dysmetric movements due to the dysfunction of cerebellar circuitry. Several tools have been developed, but so far most of these tools have remained confined to laboratories, with a lack of standardization. We report on a new device which combines the use of electromyographic (EMG) sensors with haptic technology for the dynamic investigation of wrist properties. The instrument is composed of a drivetrain, a haptic controller and a signal acquisition unit. Angular accuracy is 0.00611 rad, nominal torque is 6 N·m, maximal rotation velocity is 34.907 rad/sec, with a range of motion of -1.0472 to +1.0472 rad. The inertia of the motor and handgrip is 0.004 kg·m2. This is the first standardized myohaptic instrument allowing the dynamic characterization of wrist properties, including under the condition of artificial damping. We show that cerebellar patients are unable to adapt EMG activities when faced with an increase in damping while performing fast reversal movements. The instrument allows the extraction of an electrophysiological signature of a cerebellar deficit.

Keywords: ataxia; damping; movement; myohaptic; sensor.

Figure 1.

Experimental device (wristalyzer). The mechatronic myohaptic unit includes a moving unit and its controller. The system comprises a direct drive brushless motor with a high resolution encoder. Sampling rate is 2,048 Hz per channel. A high degree of stability on the floor is ensured by adjustable screws.

Figure 2.

Correlation of the ataxia rating score with the mean movement amplitude in the cerebellar patients during performance of fast pointing movements (FPM) in the basal mechanical condition. Aimed target located at 0.2, 0.3 and 0.4 rad from the starting position in A, B and C, respectively. Long dash: 95% confidence intervals; dotted lines: 95% prediction intervals.

Figure 3.

Relationship between the ataxia rating score and the onset latency of the antagonist EMG activities for fast pointing movements (FPM) (aimed target: 0.2 rad; basal mechanical state of the hand). Long dash: 95% confidence intervals; dotted lines: 95% prediction intervals.

Figure 4.

Movement and EMG bursts in a control subject and in a cerebellar patient (aimed target: 0.3 rad). Top panels (A,B): average of fast pointing movements (FPM). Bottom panels (C,D) refer to fast reversal movements (FRM). Blue line: no damping, black line: + 0.1 N·m·s/rad, red line: + 0.2 N·m·s/rad. Grey areas: 99% confidence interval of control values of movement amplitudes in the basal mechanical state; dotted lines in black and red: 99% confidence interval of control values during addition of 0.1 N·m·s/rad and 0.2 N·m·s/rad, respectively.

Figure 5.

Effects of damping on movement amplitudes for fast pointing movements (FPM). Aimed target: 0.2 rad, 0.3 rad and 0.4 rad, respectively in A, B and C. Values are mean ±SD. Black columns: control subjects, grey columns: cerebellar patients. *: p < 0.05; **: p < 0.01.

Figure 6.

A: mesh plots illustrating the Q_ACC and Q_DEC in control subjects and patients for fast pointing movements (FPM). Mean values are shown for the 3 aimed targets and the 3 mechanical conditions. The scaling of the intensities of both agonist and antagonist EMG activities as a function of the damping and amplitudes of motion is preserved both in control subjects and in cerebellar patients. B: for fast reversal movements (FRM), control subjects scale appropriately the intensity of the activity of the extensor carpi radialis (ECR) muscle when damping is added and when the aimed amplitude is larger. By contrast, there is a failure of cerebellar patients to adapt the EMG activity as compared to controls.

Figure 7.

Representative EMG activities for the flexor carpi radialis muscle (FCR) during rapid stretches of the wrist (extensions) in a control subject (black trace) and a cerebellar patient (red trace). The intensities of the short-latency stretch response (SLSR) are similar. The intensity of the long-latency stretch response (LLSR) is higher in the patient. Each trace corresponds to an average of 45 trials. Stretch responses are calibrated in arbitrary units. Arrowheads at the bottom of the traces indicate the onset latencies of SLSR and LLSR.

Figure 8.

Fitting of the relationship between onset latency of antagonist EMG activity and the ratios LLSR/SLSR during fast pointing movements (FPM) towards an aimed target of 0.2 rad in cerebellar patients. Medium dash: 95% confidence band.