Omar Feix Do Nascimento

We tested the feasibility for amputees to control artificial limbs using non-invasive electroencephalography (EEG). Thirteen participants engaged in attempts of isometric ankle plantar-flexions using their phantom or intact limb at slow or ballistic speeds. EEG data were analyzed for movement-related cortical potentials (MRCPs), the slow negative potentials related to the planning and execution of movements. We focused on temporal profiles and single-trial classification at electrode location Cz where MRCPs are most prominent. Distinctly different MRCP morphologies were observed for both movement speeds and phantom versus intact limbs. Crucially, time since amputation correlated significantly with classification errors for distinguishing tasks performed with the intact limb from those of the phantom limb (R = 0.36, p =0.004) and movement speed during trials of only the phantom limb (R = -0.33, p = 0.01). Here we show the persistent capacity of amputees to plan and attempt to execute limb motions at varying speeds using their phantom limb. This has implications for understanding neural adaptations over extended post-amputation periods and for the practical implementation of the MRCP in the design of brain-computer interfaces to control prosthetic devices using single-electrode EEG recordings.

Urinary tract infections (UTIs) are common and troublesome complications of clean intermittent catheterisation (CIC) in individuals suffering from incomplete bladder emptying, which may exacerbate the underlying disease and lead to hospitalisation. Aside from the design of the intermittent catheter and its handling, a recent review highlighted residual urine as one of several UTI risk factors. A new urinary intermittent catheter with multiple micro-holes has been developed for improved bladder emptying. In a controlled crossover study, adult male CIC users were randomised for a health care professional-led catheterisation with the new micro-hole zone catheter (MHZC) and a conventional eyelet catheter (CEC) in two individual test visits to compare the number of flow-stops and the residual urine at the first flow-stop as co-primary endpoints. In 42 male CIC users, the MHZC resulted in significantly fewer flow-stop episodes compared to the CEC (mean 0.17, 95% CI [0.06, 0.45] vs. mean 1.09, 95% CI [0.75, 1.6], respectively; p < 0.001) and significantly less residual urine at the first flow-stop (mean 5.10 mL, SE [1.14] vs. mean 39.40 mL, SE [9.65], respectively; p < 0.001). No adverse events were observed in this study. The results confirm the enhanced performance of the MHZC compared to a CEC, ensuring an uninterrupted free urine flow with no need to reposition the catheter until the bladder is thoroughly empty.

The study investigates the accuracy in discriminating rate of torque development (RTD) and target torque (TT) (task parameters) from electroencephalography (EEG) signals generated during imaginary motor tasks. Signals were acquired from nine healthy subjects during four imaginary isometric plantar-flexions of the right foot involving two RTDs (ballistic and moderate) and two TTs (30 and 60% of the maximal voluntary contraction torque), each repeated 60 times in random order. The single-trial EEG traces were classified with a pattern recognition approach based on wavelet coefficients as features and support vector machine (SVM) as classifier. Average misclassification rates were (mean ± SD) 16 ± 9% and 26 ± 13% for discrimination of the two TTs under ballistic and moderate RTDs, respectively. RTDs could be discriminated with misclassification rates of 16 ± 11% and 19 ± 10% under high and low TT, respectively. These results indicate that differences in both TT and RTD can be detected from single-trial EEG traces during imaginary tasks.

The aim of this study was to discriminate on a single-trial basis the cortical activity associated to two rates of torque development (RTDs) in imaginary isometric plantar flexions. Electroencephalographic (EEG), electrooculographic (EOG), and electromyographic (EMG) signals were recorded while ten healthy subjects imagined right-sided isometric ankle plantar-flexion tasks at moderate [from 0% to 60% of the maximal voluntary contraction (MVC) in 4 s] and ballistic (from 0% to 60% MVC as fast as possible) RTDs. The EEG signals were classified using feature extraction based on the marginal distribution of a discrete wavelet transform with optimization of the mother wavelet. The classifier was based on support vector machines (SVMs). Minimum misclassification rate for the best case was 8.3%. Average minimum misclassification rate over the ten subjects was (17.4 plusmn 8.4)%. The two RTDs could be best differentiated from channel C4 on average. In conclusion, different RTDs could be differentiated in imaginary isometric plantar-flexion by only using cortical potentials recorded with surface EEG. This result constitutes the first step for the development of a new type of brain-computer interfaces that rely on kinetic parameters of a single limb rather than movements of opposite limbs.

This paper summarizes the brain-computer interface (BCI)-related research being conducted at Aalborg University. Namely, an online synchronized BCI system using steady-state visual evoked potentials, and investigations on cortical modulation of movement-related parameters are presented.

The aim of this study was to investigate the contribution of proprioceptive feedback to the amplitude modulation of the soleus muscle activity during human walking. We have previously shown that slow-velocity, small-amplitude ankle dorsiflexion enhancements and reductions applied during the stance phase of the step cycle generate, respectively, increments and decrements on the ongoing soleus activity. We have also shown that the increments in soleus activity are at least partially mediated by feedback from group Ia fibres. In the present study, we further investigated the afferent-mediated contribution from muscle group II afferents, cutaneous and proprioceptive afferents from the foot, and load-sensitive afferents to the soleus EMG. Slow-velocity, small-amplitude ankle trajectory modifications were combined with the pharmaceutical depression of group II polysynaptic pathways with tizanidine hydrochloride, anaesthetic blocking of sensory information from the foot with injections of lidocaine hydrochloride, and modulation of load feedback by increasing and decreasing the body load. The depression of the group II afferents significantly reduced the soleus response to the ankle trajectory modifications. Blocking sensory feedback from the foot did not have an effect on the soleus muscle activity. Changes in body load affected the ongoing soleus activity level; however, it did not affect the amplitude of the soleus EMG responses to the ankle trajectory modifications. These results suggest that the feedback from group II afferents, and possibly from load-sensitive afferents, contribute to the amplitude modulation of the soleus muscle activity during the stance phase of the step cycle. However, feedback from cutaneous afferents and instrinsic proprioceptive afferents from the foot does not seem to contribute to this muscle activation.

A multitude of studies have demonstrated a clear activation of the motor cortex during imagination of various motor tasks; however, it is still unclear if movement-related parameters (movement direction, range of motion, speed, force level and rate of force development) specifically modulate cortical activation as they do during the execution of actual motor tasks. Accordingly, this study examined whether the rate of torque development (RTD) and/or the torque amplitude modulates cortical potentials generated during imaginary motor tasks. Fifteen subjects imagined four different left-sided isometric plantar-flexion tasks, while EEG and EMG recordings were being performed. The averaged EEG activity was analyzed in terms of movement-related potentials (MRPs), consisting of readiness potential (RP), motor potential (MP) and movement-monitoring potential (MMP). It was demonstrated that RTD and torque amplitude indeed modulate cortical activity during imaginary motor tasks. Information concerning movement-related parameters for imaginary plantar-flexion tasks seems to be encoded in the supplementary motor area (SMA) and the primary motor cortex (M1). A comparison between MRPs of imaginary and actual motor tasks revealed that early MRPs were morphologically similar, but differed significantly in amplitude. One of the possible suggestions to explain such a difference may be an “abortion” of ongoing motor programs.