Effect on signal-to-noise ratio of splitting the continuous contacts of cuff electrodes into smaller recording areas
© Ortiz-Catalan et al.; licensee BioMed Central Ltd. 2013
Received: 22 February 2012
Accepted: 5 February 2013
Published: 21 February 2013
Cuff electrodes have been widely used chronically in different clinical applications. This neural interface has been dominantly used for nerve stimulation while interfering noise is the major issue when employed for recording purposes. Advancements have been made in rejecting extra-neural interference by using continuous ring contacts in tripolar topologies. Ring contacts provide an average of the neural activity, and thus reduce the information retrieved. Splitting these contacts into smaller recording areas could potentially increase the information content. In this study, we investigate the impact of such discretization on the Signal-to-Noise Ratio (SNR). The effect of contacts positioning and an additional short circuited pair of electrodes were also addressed.
Different recording configurations using ring, dot, and a mixed of both contacts were studied in vitro in a frog model. An interfering signal was induced in the medium to simulate myoelectric noise. The experimental setup was design in such a way that the only difference between recordings was the configuration used. The inter-session experimental differences were taken care of by a common configuration that allowed normalization between electrode designs.
It was found that splitting all contacts into small recording areas had negative effects on noise rejection. However, if this is only applied to the central contact creating a mixed tripole configuration, a considerable and statistically significant improvement was observed. Moreover, the signal to noise ratio was equal or larger than what can be achieved with the best known configuration, namely the short circuited tripole. This suggests that for recording purposes, any tripole topology would benefit from splitting the central contact into one or more discrete contacts.
Our results showed that a mixed tripole configuration performs better than the configuration including only ring contacts. Therefore, splitting the central ring contact of a cuff electrode into a number of dot contacts not only provides additional information but also an improved SNR. In addition, the effect of an additional pair of short circuited electrodes and the “end effect” observed with the presented method are in line with previous findings by other authors.
KeywordsCuff electrodes Neural recordings Neural prostheses
Cuff electrodes have been widely used in neuroprosthetics with different applications such as footdrop correction by stimulation of the peroneal nerve in patients with hemiplegia ; stimulation of the median and ulnar nerves for pain control ; treatment of obstructive sleep apnea by stimulation of the hypoglossal nerve ; and stimulation of the optic nerve in blind patients as part of a visual prosthesis .
Stimulation has been the main clinical application of cuff electrodes, with few exceptions such as recordings from the sural nerve for footdrop correction , and from the digital nerve for force-controlled hand-grasping . The information provided by these recordings, however, has only been used in an “on/off” manner. In an animal model, offline discrimination between two to three different stimuli using a single channel cuff electrode has been demonstrated . The signal to noise ratio (SNR), however, was reported as the main limiting factor for the classification performance, even in the absence of myoelectric interference which is known to be the major issue in cuff recordings .
In this study, we investigate different cuff configurations that could potentially increase the amount of information retrieved, and at the same time, preserve an acceptable SNR taking into account the main noise sources. When using cuff electrodes, the SNR depends on different factors such as the cuff geometry  and closure ; the recording configuration [10–12]; and the amplification electronics [13–17]. The latter has been approached in different ways, such as introducing audio transformers to match impedance, reduce the common noise (e.g. power lines), and as a passive pre-amplification [13, 15]. Our study was focused on the recording contacts configuration in order to identify the best performing topologies.
Dot configurations have been successfully employed for stimulation purposes [27–30]. It has been shown that dot contacts can be advantageous while using steering currents to better localize the stimulation . In humans, selective activation of upper limb muscles has been shown feasible intraoperatively using this type of contacts .
Although the clinical use of dot contacts for recordings purposes is limited, feasibility has been shown in long-term (>2 years) selective recordings using multi-contact cuff electrodes in canines , where Rozman et al. used 4 out of 11 possible tripolar montages (3 bands of 11 contacts) to directly discriminate afferent signals from 2 muscles (1 source to 1 channel). Using statistical models, Cheng et al. achieved activity discrimination of the tibial and peroneal nerves in recordings obtained in the proximal sciatic nerve of rabbits (2 bands of 4 contacts) . More recently, Zariffa et al. increased the number of contacts to 56 (7 bands of 8 contacts) to proximally discriminate the tibial, peroneal, and sural nerves using an experimental leadfield approach . Although relatively satisfactory in offline signal discrimination under controlled environments, the SNR was admittedly low , which would hinder stability in a clinical implementation.
The flat interface nerve electrode (FINE) , a cuff-like neural interface, has been used in a combination of ring and dot contacts to selectively record signals from multifasciculated nerves in canines . Spatial filtering or beamforming algorithms have been used to extract neural information with considerable crosstalk from this multi-contact electrode [25, 26]. The contacts selection, however, has not been motivated.
The tripolar configuration allows a noise field reduction by short-circuiting the end electrodes . It has been suggested that the interfering signal picked up by the end electrodes travels through the short-circuited path rather than in the cuff, thus, reducing the interference . The efficiency of the tripole, and its later modifications, resides in the linearization of the extra-fascicular noise field by the physical confinement due to the cuff itself [11, 16]. The main purpose of this study was to investigate if the screening properties of the tripole are degraded by splitting the ring contacts into single dots. At the same time, we extended the investigation to study the effect of displacing the recording contacts inwards the cuff (the “end effect”), as previously analyzed theoretically [8, 12, 39], and tested in an animal model .
In order to compare how different recording configurations affect the SNR, an experiment was designed where the sources of noise and other environmental factors were kept constant. The only variable was the assignment of the electrode contacts to the differential amplifier inputs.
The nerve was entirely submerged in physiological Ringer solution at 24° C (±2° C). In order to simulate the noise induced by surrounding muscles, silver electrodes were placed in the solution approximately 55 mm apart, in opposite corners of the container. These were driven by a 200 mV p p sinusoidal signal at 200 Hz. This noise is further referred to as the simulated myoelectric signal (SMES). The 200 Hz frequency corresponds to a normal peak frequency of myoelectric signals [40–42]. The 200 mV p p amplitude was selected for being high enough to appear in the recordings of all configurations, thus, allowing comparisons to be made. A drawing of the experiment setup is given in Figure 3. The cuff electrodes were finally wrapped around the nerve, with the stimulation electrode near the end of the nerve that would have been proximal in the body.
The stimulation of the nerve was performed with an asymmetric charge balanced pulse (Lilli pulse). In this experiment, however, at the stimulating contact (the one closest to the recording electrodes), the first phase was the anodic charge compensating pulse (70 μ A amplitude and 250 μ s duration) while the second phase represented the actual cathodic phase (-350 μ A and 50 μ s respectively). This pulse shape was used in order to ease the identification of the SA and the actual CAP.
A single experiment consisted in the preparation of the sciatic nerve; placement of the recording and stimulation electrodes; submersion of the nerve in physiological Ringer solution; fixation of the electrodes inducing the SMES; fixation of the reference electrode in a corner opposite to the SMES (not illustrated). The experiments were performed inside a Faraday cage and the recordings were amplified 100 times with a low pass filter at 3 kHz (Tektronix ADA400A), and digitalized at 10 kHz. For each configuration of a single experiment, 6 recordings were obtained and averaged for each of the three signals: the CAP, the SA and the SMES. Peak to peak amplitudes were measured as the study dependent variables. The complete experiment was repeated 16 times in different nerves and with different electrodes.
Since this study is about comparing the amplitude of the CAP against SA and SMES, it was important to make sure the nerve remained healthy for the whole duration of each experiment. For that purpose, the nerve conduction velocity (CV) and the CAPs’ amplitude were recorded at the onset and end of each experiment, and then compared. The CV and CAPs were obtained from two simultaneous bipolar recordings derived from both nerve end pair of contacts. No experiment had to be rejected on the basis of significant nerve degradation as reported in the results section.
All the spiral cuff electrodes  used here were 26 mm long with an inner diameter of 1 mm. The platinum contacts were longitudinally spaced by 5 mm (center to center) and the extreme ones were at 3 mm from the cuff ends (center to edge). Two types of contacts were used, named respectively the dot and the ring contacts. The dot contacts were made of 1x1 mm squares embedded in the silicone rubber with a round opening of 500 μ m diameter facing the nerve. The ring contacts were made of 1x3 mm platinum foils with 3 openings of 500 μ m distributed around the nerve periphery.
The same stimulation electrode was used in all experiments. It consisted of a 5 mm long spiral cuff  with 2 ring contacts of 1x3 mm separated from each other and from the cuff ends by 1 mm space. The opening windows were similar to the ones described for the ring contacts.
The measured CAP, SA and SMES values vary randomly between different experiments due to slight differences in the experimental setup, such as the cuff fitting around the nerve; the nerve dimensions and physiological status; the nerve conduction velocity; the precise placement of the electrodes inducing SMES; and the separation of the stimulation and recording electrodes. However, once the experimental setup is completed, the only intra-study difference between the recording configurations is the connection of the electrode contacts to the differential amplifier. Therefore, we choose to focus on the improvement ratios (IRs), a concept developed by Andreasen and Struijk to compare different recording configurations [10, 39].
In this application, the IRs can be defined as follows:
Similarly, we quantify the SMES ratio as rSMES = SMESj/SMESi, and the SA ratio as rSA = SAj/SAi. The CAP, SMES and SA correspond to the peak to peak values.
The non-parametric, Wilcoxon signed rank test, was used to evaluate the significance of pair-wise differences between configurations in the same electrode design and experiment or session (same nerve, electrode, and environmental conditions). A different non-parametric test, the Mann Whitney U-test, was required to compare configurations from different electrode designs since the pair-wise relationship is non-existing in this case. When comparing recordings from different electrode designs and therefore, different sessions, the effects of the experimental setup were accounted for by using recordings from one of the configurations used in the same experiment as a normalization parameter. In our case, this configuration was the bipolar recording between contact 1 and 5 (+...-), since it is identical in both electrode designs. The significance level used for both statistical tests was p = 0.05.
The effect of reducing the tripolar length by moving the contacts inwards
The end contacts displacement corresponds to a 50% reduction of the tripolar length which incidentally resulted in around 50% reduction of all the signals recorded, as intuitively expected in a proportionally linear relationship but not observed by other authors [10, 39].
Effect of an additional short-circuited pair of electrodes
Short-circuiting a pair of ring electrodes in the cuff ends proved to increase the SNR in both bipolar and tripolar recordings, for both ring and dot configurations (see Figure 9). The SNR of bipolar ring recordings was increased on average by IR(SA) = 1.68 (r, 0+.-0/.+.-.), and IR(SMES) = 1.50 (r, 0+.-0/.+.-.), both with statistical significance. The effect on the tripolar configuration was smaller with IR(SA) = 1.06 (r, 0+-+0/.+-+.), and IR(SMES) = 1.31 (r, 0+-+0/.+-+.). Only the latter reached statistical significance.
The effect of splitting the ring contacts
We found important but statistically non-significant differences between SNRs of ring and dot configurations in bipolar recordings (.+.-.) with IR(SA) = 0.81 (r.+.-./d.+.-.), and IR(SMES) = 0.87 (r.+.-./d.+.-.). Tripolar recordings (.+-+.) also yielded rather large differences with IR(SA) = 0.84 (r.+-+./d.+-+.), and IR(SMES) = 1.67 (r.+-+./d.+-+.), however, again without statistical significance.
The configurations with the best SNR against SMES were the ring short-circuited tripole (ro+-+o) and the mixed tripole (m+.-.+), for ring and mixed electrode designs respectively. The differences between these two configurations yielded an IR(SMES) = 1.30 (m+.-.+/ro+-+o) without reaching statistically significance. The improvement of SNR rejecting SA was higher with IR(SA) = 1.93 (sd+.-.+/ro+-+o), however still not statistically significant. Figure 11 shows the IR of the latter configurations which have the best SNR(SMES).
Finally, the effect of asymmetry in SNR(SA) for the tripolar configuration can be observed in Figure 13. This effect was most important for the mixed configuration with IR(SA) = 1.38 (m, +.-.+/+..-+) in comparison with the ring configurations with IR(SA) = 0.96 (r, +.-.+/+..-+), although, both with non-statistical significance. The effect of asymmetry was also higher for the mixed configurations when evaluating for SMES with IR(SMES) = 7.51 (m, +.-.+/+..-+), against IR(SMES) = 2.93 (r, +.-.+/+..-+) of the ring tripole, both with statistical significance. The difference between the asymmetric tripoles, however, was not significant when comparing the ring and mixed designs (m/r, +..-+) with IR(SA) = 1.44 and IR(SMES) = 0.74.
The effect of displacing the end tripolar contacts inwards the cuff
In agreement with our results, Andreasen and Struijk have reported a decrement in CAP and both interfering signals (SA and SEMG) when moving the contacts inwards the cuff. These authors also found a progressively more modest SNR increment when the contacts were moved further inside (the “end effect”). This mainly resulted from the fact that the noise reduction rate approximated that of the CAP. Taking this into account, they concluded that a cuff inwards placement of 14.8% relative to the cuff length was close to optimal. For their cuff of 27 mm length and 2 mm diameter, this meant a distance of 4 mm from the contact’s center to the cuff edge. We placed the end contacts of the tripole 30.7% inwards (.+-+.), and found no significant improvement in SNR when compared with a 9.6% inwards placement (+.-.+). However, there was a statistically significant reduction of CAP, SA, and SEMG. These reductions were close enough not to impact the SNRs significantly as expected from previous research . This suggests that if signal amplitude only is considered, the optimal contact placement is between 10% and 30% inwards a 26 mm long cuff with 1 mm diameter. Note that reducing the cuff length increases the interference pick up close to the ends . Combined with a reduction of the spatial extent of the recording tripole, the inwards displacement of the end contacts used in this study has a negligible effect on SNR which is relevant to the rest of the study. It is worthy of mentioning that this study was performed with a defined cuff length and increasing it together with the tripole width will improve the SNR. However, the cuff length is strongly limited by the anatomy and implantability requirements.
Effect of an additional short-circuited pair of electrodes
Our results, as well as previous research, showed an increment in SNR when using an additional short-circuited pair of electrodes [10, 39]. This improvement was observed in bipolar as well as tripolar recordings, and for both ring and dot configurations. This electrode feature resulted in a 68% (0+.-0/.+.-.) improvement for SA and 50% for SMES. In tripolar recordings (0+-+0/.+-+.), the results are clearly less with 6% improvement for SA and 31% for SMES. As previously observed in ring configurations , we thus found the effect of an additional short-circuited pair of electrodes to be higher in bipolar than in tripolar configurations. Andreasen and Struijk suggested that the larger improvement in bipolar recordings could be explained by considering that the tripolar configuration already includes a short-circuited pair of electrodes. Our results support this argument with the additional observation that the dot tripolar configuration, in which the shorted extremes do not form full rings, had higher IRs (0+-+0 /.+-+.) than tripoles whereby the shorted outer contacts are full rings, 22% (d/r) for SA and 30% (d/r) for SEMG. The latter suggests a negative impact on the noise field reduction as the extend of a contact around the nerve is reduced to a dot. The differences between these two configurations, however, were not statistically significant in our work.
The short-circuited tripole has been shown to be the best performing configuration [10, 12, 39], mainly because it benefits from the “end effect” and the additional short-circuited pair of electrodes [10, 39]. The “end effect” known as the reduction of the interfering noise by moving the contacts inwards the cuff, is in our case not contributing to an improvement since the signal of interest (CAP) is also reduced in similar proportions. Therefore, it is worth noting that the observed improvement in SNR most likely corresponds to the additional short-circuited electrodes only.
As mentioned in the introduction, there are two types of tripolar montages: 1) tripolar, and 2) true-tripolar. These configurations reduce the effect of electrical sources located outside the cuff by using the average of the linearized field as a reference. The tripole reduces the externally induced field by short-circuiting it, which is not the case in true-tripole since it has all three contacts connected to the high input impedance of the amplifiers. This has to be taken into consideration while extrapolating these results to true-tripolar recordings.
Effect of using dot contacts over ring contacts
Although the proposed mixed tripole (m+.-.+) yielded higher SNRs than the previously best known configuration, namely the ring short-circuited tripole (ro+-+o) [10, 11, 39] by 30% for SMES and 93% for SA, these differences do not reach statistical significance. These findings, however, suggest that an additional pair of short-circuited electrodes might not be necessary when employing a mixed tripole. Reduction of the required number of contacts could be a practical advantage of such scenario. Two interpretations can be considered here: 1) the reduction in noise rejection is compensated by an increment in signal amplitude due to the longer distance between the tripole contacts; or, 2) some of the CAP amplitude is picked up by the reference contacts but this is smaller with ring contacts thus yielding larger potentials when a dot contact records against two ring contacts. Since it has been shown that the effect of longer distances between contacts falls short of balancing against the benefit of the additional short-circuited electrodes , the second explanation should thus be considered. The SMES amplitude was on average 32% larger (p<0.05) in the mixed tripole, however, the increment in CAP was even higher with 129% (p<0.05), therefore yielding an improved SNR. It is worth noting, as mentioned before, that the outer tripole contact displacement inwards the cuff did not contribute to a higher SNR in this specific design. The question as to whether an optimal distance to the cuff ends could incline the balance towards the short-circuited tripole configuration has still to be investigated. Our work, however, aimed at investigating whether splitting ring contacts into discrete dots would negatively affect the SNR. Our results suggest that if only the central contact is reduced to a dot, the SNR of the short-circuited tripole would actually improve. This is because on average, the mixed tripole yields an IR(SA) of 100% and IR(SMES) of 87%, when compared with its ring counterpart (p<0.05). Furthermore, Chu et al. has recently proposed an additional middle ring contact as an enhancement of the ring tripolar configuration . This “revised-tripole” was incidentally studied in our experiment (+-.-+). SNR improvements for SA and MES were observed in both studies independently. The mixed tripole, however, still produced higher SNR, but more importantly, splitting the central contacts of the revised-tripole was also found beneficial (see Figure 7 and 13).
The selectivity of a discrete contact is conditioned by its distance to the source. In the case where a signal of particular interest is conducted in a fascicle situated far from the dot contact, a higher SNR for that particular signal might be achieved using a ring contact instead. On the other hand, Rozman et al. have shown that having sufficient, and shortly spaced contacts (0.5 mm), selection of the best placed electrodes based in SNR is possible .
Reducing only the central ring to a dot contact has favorable practical implications. A smaller contact area is more easily obstructed from the site of recording interest (e.g. if placed over a blood vessel during implantation). When using end ring contacts, this problem is only of concern for the central contacts. Fewer contacts also reduce the number of wires required.
Despite the large values (16% reduction (r/d) for SA, but 67% improvement (r/d) for SMES) the difference in SNR between all ring and all dot montages (.+-+.) remain statistically non-significant. Similar contrasting findings result from the comparison between short-circuited tripoles (o+-+o) with 29% reduction (r/d) and 50% improvement (r/d). These results indicate that the interfering signals are rejected differently depending on the source location and the recording configuration. This is in line with the finding that the best SNR against SA corresponds to the modified mixed tripole (+—+). In any case, the tripole and short-circuited tripole have the best SNR tradeoff for both kinds of noise.
Nerve conduction velocity and experimental model
The nerve conduction velocity determines the optimal inter-electrode distance for recording CAPs. It is thus obvious that with very long cuffs and increasing distances between contacts, there is a point where the CAP amplitude will no longer increase while the noise figure, independent of the conduction velocity. This is expected to happen at shorter lengths and distances with slow conducting fibers. Here, we use a nerve with a rather slow conduction velocity compared to the later foreseen human applications while long cuffs quickly become surgically unrealistic. The conclusions given hereafter thus seem broadly applicable but it should also be pointed out that in a follow up study, variations of the frog model temperature will be used to fully explore the effects of conduction velocity.
Previous research has mostly resorted to mammal models for investigating the impact of cuff designs on the SNR [10, 11, 44]. This involves chronic implantation and requires longer times of specialized care. In this work, we have shown that comparable results can be achieved in a simpler in vitro model if the aim of the study is solely to investigate IRs on SNRs. The measured effect of an additional pair of short circuited electrodes and the “end effect” are in line with previous findings [10, 12, 38, 39, 44], thus reassuring the presented methodology.
In this work, different cuff electrode contact configurations were explored by comparing recordings obtained with well-established montages, namely the tripole and the short-circuited tripole. Our results show that a mixed tripole (m+.-.+) performs better than a similar configuration including only ring contacts. Therefore, splitting the central ring contact of a cuff electrode into a number of dot contacts, not only results in additional channels to potentially retrieve more information, but more importantly, an improved SNR is obtained with this modification.
This conclusion also holds for the well-known short-circuited tripole, usually considered the best performing montage. The dimensions of the short-circuited tripole used in this study, however, did not allow to take advantage of the “end effect”. Therefore, further research is required to determine the optimal size for short-circuited mixed tripole.
This work was funded by VINNOVA, the Walloon Region Project DEEP (DGTRE N°5992) and Integrum AB.
- Waters R, McNeal D, Faloon W, Clifford B: Functional electrical stimulation of the peroneal nerve for hemiplegia. Long-term clinical follow-up. J Bone Jt Surg 1985,67(5):792-793.
- Nashold B Jr, Goldner J, Mullen J, Bright D: Long-Term Pain Control by Direct Peripheral-Nerve Stimulation. J Bone Jt Surg 1982, 64-A: 1-10.
- Eastwood P, Walsh J, Maddison K, Baker V, Tesfayesus W, Hillman D: Treatment of Obstructive Sleep Apnea with Unilateral Hypoglossal Nerve Stimulation. Am J Respir Crit Care Med 2010, 181: A5393.
- Veraart C, Raftopoulos C, Mortimer JT, Delbeke J, Pins D, Michaux G, Vanlierde A, Parrini S, Wanet-Defalque MC: Visual sensations produced by optic nerve stimulation using an implanted self-sizing spiral cuff electrode. Brain Research 1998, 813: 181-186. 10.1016/S0006-8993(98)00977-9View ArticlePubMed
- Haugland M, Sinkjaer T: Cutaneous Whole Nerve Recordings Used for Correction of Footdrop in Hemiplegic Man. IEEE Trans Rehabil Eng 1995,3(4):307-317. 10.1109/86.481970View Article
- Haugland M, Lickel A, Haase J, Sinkjaer T: Control of FES thumb Force Using Slip Information Obtained from the Cutaneous Electroneurogram in Quadriplegic Man. IEEE Trans Rehabil Eng 1999,7(2):215-227. 10.1109/86.769412View ArticlePubMed
- Raspopovic S, Carpaneto J, Udina E, Navarro X, Micera S: On the identification of sensory information from mixed nerves by using single-channel cuff electrodes. J Neuroengineering and Rehabilitation 2010, 7: 17. 10.1186/1743-0003-7-17View Article
- Rahal M, Taylor J, Donaldson N: The Effect of Nerve Cuff Geometry on Interference Reduction: A Study by Computer Modeling. IEEE Trans Biomed Eng 2000, 47: 136-138. 10.1109/10.817629View ArticlePubMed
- Andreasen L, Struijk JJ: On the importance of configuration and closure of nerve cuff electrodes for recording. Proc 20th Int Conf IEEE EMBS 1998,20(6):3004-3007.
- Andreasen L, Struijk JJ: Artefact Reduction With Alternative Cuff Configurations. IEEE Trans Biomed Eng 2003,50(10):1160-1166. 10.1109/TBME.2003.817633View ArticlePubMed
- Pflaum C, Riso R, Wiesspeiner G: Performance of alternative amplifier configurations for tripolar nerve cuff recorded ENG. J Rehabil Res Dev 1996, 1: 375-376.
- Rahal M, Winter J, Taylor J, Donaldson N: An Improved Configuration for the Reduction of EMG in Electrode Cuff Recordings: A Theoretical Approach. IEEE Trans Biomed Eng 2000,47(9):1281-1284. 10.1109/10.867963View ArticlePubMed
- Sahin M: A low-noise preamplifier for nerve cuff electrodes. IEEE Trans. Neural Syst. Rehabil. Eng 2005,13(4):561-5. 10.1109/TNSRE.2005.856073View ArticlePubMed
- Demosthenous A, Triantis I: An adaptive ENG amplifier for tripolar cuff electrodes. Solid-State Circuits, IEEE J 2005,40(2):412-421.View Article
- Nikolic Z, Popovic D, Stein R, Kenwell Z: Instrumentation for ENG and EMG recordings in FES systems. Biomed Eng, IEEE Trans 1994,41(7):703-706. 10.1109/10.301739View Article
- Struijk J, Thomsen M: Tripolar nerve cuff recording: stimulus artifact, EMG and the recorded nerve signal. In Eng Medicine and Biology Society, 1995., IEEE 17th Annual Conference, Volume 2. : IEEE; 1995:1105-1106.
- Triantis I, Demosthenous A, Donaldson N: Comparison of three ENG tripolar cuff recording configurations. In Proc. 1st Int. IEEE EMBS Conf. Neural Eng. Capri Island; 2003:364-367.
- Loeb G, Peck R: Cuff electrodes for chronic stimulation and recording of peripheral nerve activity. J Neurosci Methods 1996, 64: 95-103. 10.1016/0165-0270(95)00123-9View ArticlePubMed
- Tong K, Ronga Lia L, Caob J: Effects of consecutive slips in nerve signals recorded by implanted cuff electrode. Med Eng Phys 2008, 30: 460-465. 10.1016/j.medengphy.2007.05.007View ArticlePubMed
- Donaldson N, Rieger R, Schuettler M, Taylor J: Noise and selectivity of velocity-selective multi-electrode nerve cuffs. Med & Biol Eng & Comput 2008,46(10):1005-18. 10.1007/s11517-008-0365-4View Article
- Rozman J, Zorko B, Bunc M: Selective recording of electroneurograms from the sciatic nerve of a dog with multi-electrode spiral cuffs. Jpn J Physiol 2000,50(5):509-14. 10.2170/jjphysiol.50.509View ArticlePubMed
- Cheng HS, Ju MS, Lin CCK: Estimation of peroneal and tibial afferent activity from a multichannel cuff placed on the sciatic nerve. Muscle & nerve 2005,32(5):589-99. 10.1002/mus.20404View Article
- Zariffa J, Nagai MK, Schuettler M, Stieglitz T, Daskalakis ZJ, Popovic MR: Use of an experimentally derived leadfield in the peripheral nerve pathway discrimination problem. IEEE Trans Neural Syst Rehabil Eng 2011,19(2):147-56.View ArticlePubMed
- Tesfayesus W, Durand DM: Blind source separation of peripheral nerve recordings. J Neural Eng 2007,4(3):S157-67. 10.1088/1741-2560/4/3/S03View ArticlePubMed
- Wodlinger B, Durand DM: Selective recovery of fascicular activity in peripheral nerves. J Neural Eng 2011,8(5):056005. 10.1088/1741-2560/8/5/056005PubMed CentralView ArticlePubMed
- Calvetti D, Wodlinger B, Durand DM, Somersalo E: Hierarchical beamformer and cross-talk reduction in electroneurography. J Neural Eng 2011,8(5):056002. 10.1088/1741-2560/8/5/056002PubMed CentralView ArticlePubMed
- Grill W, Mortimer J: Stability of Input-Output Properties of Chronically Implanted Multiple Contact Nerve Cuff Stimulating Electrodes. IEEE Trans Rehabil Eng 1998,6(4):364-373. 10.1109/86.736150View ArticlePubMed
- Sweeney J, Ksienski D, Mortimer J: A Nerve Cuff Technique for Selective Excitation of Peripheral Nerve Trunk Regions. J Appl Physiol 1990,37(7):706-715.
- McNeal D, Baker L, Symons J: Recruitment Data for Nerve Cuff Electrodes: Implications for Design of Implantable Stimulators. IEEE Trans Biomed Eng 1989,36(3):301-308. 10.1109/10.19851View ArticlePubMed
- Polasek KH, Hoyen Ha Keith, Tyler DJ: Human nerve stimulation thresholds and selectivity using a multi-contact nerve cuff electrode. IEEE Trans Neural Syst Rehabil Eng 2007, 15: 76-82.View ArticlePubMed
- Tyler D, Durand D: Functionally selective peripheral nerve stimulation with a flat interface nerve electrode. Neural Syst Rehabilitation Eng, IEEE Trans 2002,10(4):294-303. 10.1109/TNSRE.2002.806840View Article
- Yoo PB, Durand DM: Selective recording of the canine hypoglossal nerve using a multicontact flat interface nerve electrode. IEEE Trans Bio-medical Eng 2005,52(8):1461-9. 10.1109/TBME.2005.851482View Article
- Ortiz-Catalan M, Brånemark R, Håkansson B, Delbeke J: On the viability of implantable electrodes for the natural control of artificial limbs: Review and discussion. Biomed Eng OnLine 2012, 11: 33. 10.1186/1475-925X-11-33PubMed CentralView ArticlePubMed
- Haugland M, Hoffer J, Sinkjaer T: Skin Contact Force Information in Sensory Nerve Signals Recorded by Implanted Cuff Electrodes. IEEE Trans Rehabil Eng 1994, 2: 18-28. 10.1109/86.296346View Article
- Stein R, Charles D, Davis L, Jhamandas J, Mannard A, Nichols T: Principles underlying new methods for chronic neural recording. Can J Neurol Sci 1975,2(3):235-44.PubMed
- Stein R, Nichols T, Jhamandas J, Davis L, Charles D: Stable long-term recordings from cat peripheral nerves. Brain Res 1977, 128: 21-38. 10.1016/0006-8993(77)90233-5View ArticlePubMed
- Sahin M, Haxhiu M, Durand D, Dreshaj I: Spiral nerve cuff electrode for recordings of respiratory output. J Appl Physiol 1997, 83: 317-322.PubMed
- Struijk J, Thornsen M, Lorsen J, Sinkjaer T: Cuff Electrodes for Long-Term Recording of Natural Sensory Information. IEEE Eng Med Biol Mag 1999,18(3):99-98. 10.1109/51.765195View Article
- Andreasen L, Struijk JJ: Model-based evaluation of the short-circuited tripolar cuff configuration. Med Biol Eng Comput 2006, 44: 404-413. 10.1007/s11517-006-0057-xView ArticlePubMed
- Nikolic ZM, Popovi DB, Stein RB, Kenwell Z: Instrumentation for ENG and EMG Recordings in FES Systems. IEEE Trans Biomed Eng 1994,41(7):703-706. 10.1109/10.301739View ArticlePubMed
- Gordon T, Hoffer J, Jhamandas J, Stein R: Long-Term Effects of Axotomy on Neural Activity During Cat Locomotion. J Physiol 1980, 303: 243-263.PubMed CentralView ArticlePubMed
- Hoffer J, Loeb G: Implantable electrical and mechanical interfaces with nerve and muscle. Ann Biomed Eng 1980, 8: 351-360. 10.1007/BF02363438View ArticlePubMed
- Naples G, Mortimer JT, Scheiner A, Sweeney J: A spiral nerve cuff electrode for peripheral nerve stimulation. IEEE Trans Biomed Eng 1988,35(11):905-916. 10.1109/10.8670View ArticlePubMed
- Chu JU, Song KI, Han S, Lee SH, Kim J, Kang JY, Hwang D, Suh JKF, Choi K, Youn I: Improvement of signal-to-interference ratio and signal-to-noise ratio in nerve cuff electrode systems. Physiol Meas 2012,33(6):943-67. 10.1088/0967-3334/33/6/943View ArticlePubMed
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.