Sharpness Recognition Based on Synergy between Bio-inspired Nociceptors and Tactile Mechanoreceptors&nbsp;

doi:10.21203/rs.3.pex-1325/v1

Method Article

Sharpness Recognition Based on Synergy between Bio-inspired Nociceptors and Tactile Mechanoreceptors

https://doi.org/10.21203/rs.3.pex-1325/v1

This work is licensed under a CC BY 4.0 License

This protocol has been posted on Protocol Exchange, an open repository of community-contributed protocols sponsored by Nature Portfolio. These protocols are posted directly on the Protocol Exchange by authors and are made freely available to the scientific community for use and comment.

Version 1

posted

You are reading this latest protocol version

Touch and pain sensations are complementary aspects of daily life that convey crucial information about the environment while also providing protection to our body. Technological advancements in prosthesis design and control mechanisms assist amputees to regain lost function but often they have no meaningful tactile feedback or perception. In the present study, we propose a bio-inspired tactile system with a population of 23 digital afferents: 12 RA-I, 6 SA-I, and 5 nociceptors. Indeed, the functional concept of the nociceptor is implemented on the FPGA for the first time. One of the main features of biological tactile afferents is that their distal axon branches in the skin, creating complex receptive fields. Given these physiological observations, the bio-inspired afferents are randomly connected to the several neighboring mechanoreceptors with different weights to form their own receptive field. To test the performance of the proposed neuromorphic chip in sharpness detection, a robotic system with three-degree of freedom equipped with the tactile sensor indents the 3D-printed objects. Spike responses of the biomimetic afferents are then collected for analysis by rate and temporal coding algorithms. In this way, the impact of the innervation mechanism and collaboration of afferents and nociceptors on sharpness recognition are investigated. Our findings suggest that the synergy between sensory afferents and nociceptors conveys more information about tactile stimuli which in turn leads to the robustness of the proposed neuromorphic system against damage to the taxels or afferents. Moreover, it is illustrated that spiking activity of the biomimetic nociceptors is amplified as the sharpness increases which can be considered as a feedback mechanism for prosthesis protection. This neuromorphic approach advances the development of prosthesis to include the sensory feedback and to distinguish innocuous (non-painful) and noxious (painful) stimuli.

Neuroscience

Biotechnology

Computational biology and bioinformatics

Touch sense

E-skin

Nociceptor

Mechanoreceptor

Neuromorphic circuit

Tactile afferents

FPGA

One of the main functions of the somatosensory system is to respond to the various types of tactile stimuli¹. Touch sense provides valuable and essential contact information and allows us to interact with the environment and perform daily tasks². Meissner corpuscles, Merkel cells, Ruffini endings, and Pacinian corpuscles are the primary skin mechanoreceptors that transmit tactile information to the upper layers of the nervous system. The Merkel cells and Ruffini endings are labeled as slowly adapting (SA) and respond to the sustained tactile stimuli. Meissner and Pacinian corpuscles which are known as rapidly adapting (RA) mechanoreceptors, respond to the onset and offset of the tactile stimulation^1,3. These mechanoreceptors are innervated by the first-order neurons of the tactile pathway. The innervation pattern enables individual afferents to encode a portion of the geometric characteristics of the touched objects⁴. More recently, it is demonstrated that tactile information coding⁵ and tactile features extraction⁴ are also done by fingertip. Indeed, activation of tactile afferents spatially encodes the contact stimuli and sends the tactile information to the upper layers of the somatosensory pathway.

Free nerve endings (nociceptors) are placed in the exterior layer of the skin (epidermal layer) and are widely distributed over the body. They convey the tactile stimuli to the spinal cord leading to the perception of a painful experience⁶. Free nerve endings innervate the skin, bones, muscles, heart, and most of the internal organs. Nociceptors behave as high-threshold mechanoreceptors (HTMR) and respond to harmful stimuli through Aβ, Aδ, and C nerve fibers³. The mechanism of pain perception has its own peripheral receptors and includes a complex and chemically unique set of central circuits⁷. It has been demonstrated that painful perception increases when nociceptors are active⁷. In this way, we can perceive a range of innocuous and noxious feelings.

Despite substantial progress in the design and control of prosthesis mechanics⁸, sensory perception of prosthetic hands is at the beginning of the road. Due to the importance of the tactile sense and its significant role in prostheses, it has undoubtedly attracted much attention to the development of new tactile sensors and bringing back sensory information in amputees. Recent studies focus on replicating the behavior of biological tactile receptors using sophisticated skin dynamics⁹ and neuromorphic systems¹⁰ to improve the efficiency and performance over traditional techniques. The flexible electronic elements^11-13, self-healing ^14,15 recyclable materials¹⁶, mechanoreceptor-inspired elements^14,17, and optoelectronic strain sensors¹⁸ have been proposed for prosthetic limbs. In this research, a novel neuromorphic system is designed and then tested by taking into account the biological features of mechanoreceptors and nociceptors for interpretation of tactile information.

Neuromorphic systems replicate the biological functions and spike-based neuronal processing and are broadly based on the analog and digital realization¹⁹. Neuromorphic sensory systems have made a great step forward in recent years using a new form of asynchronous output representation which provides timing information similar to the action potentials in the biological neuronal system²⁰. In the last few years, the application of spiking neural networks and neuromorphic implementations in tactile systems have been increased^10,21-23. One of the most effective methods of realizing these computational neural models is digital circuit implementation due to their high performance for practical applications^24-30. Digital execution with Field-Programmable Gate Array (FPGA) offers parallel computations and flexibility for algorithm investigation while filling time and performance limitations. FPGAs have broad applications in the neural network simulations³¹ and motivate further exploration^32,33. An approximate circuit technique was used to implement tactile data processing on FPGA for the e-skin applications³⁴. Furthermore, the spiking neural network implemented on FPGA was proposed for bi-directional interaction with living neurons cultured in microelectrode array³⁵. The spiking model of cutaneous mechanoreceptor is implemented on the digital hardware (FPGA) to identify the distinct pressure stimuli³⁶. For simulation and digital execution of the SA-I and RA-I afferents on the FPGA, the Izhikevich neuron model was frequently used in recent studies due to its rich dynamics which is suitable for tactile sense modeling^36,37. Salimi-Nezhad and his colleagues³⁸ implemented a population of afferents on the FPGA to realize the spatial coding and used a glove covered by pressure sensors to recognize objects during grasping. A neuromorphic system for pain perception and self-protection of a hand prosthesis was introduced by Osborn and his colleagues³⁷. They fabricated a multilayered e-skin which imitates the behavioral characteristics of mechanoreceptors and nociceptors to provide sensory feedback for a prosthesis.

Given the fact that the majority of tactile information collected from the environment is encoded not only in multiple sub-modalities but also through a population of different afferent types, in the present research a bio-inspired digital system for the first layer of tactile sensory pathways including SA-I/RA-I afferents and nociceptors is designed. Specifically, the concept of the nociceptor is functionally implemented on the FPGA for the first time. One of the main features of tactile afferents is that their distal axon branches in the skin, creating complex receptive fields³⁹. Consequently, the innervation concept to form receptive fields is also integrated into the proposed tactile neuromorphic systems. The digital afferents have receptive fields that overlap with each other. To have a bio-inspired model for the SA-I/RA-I afferents and nociceptor, the Izhikevich neuron model is considered. Moreover, similar to the biological afferents which are not synaptically connected and only convey tactile information from the fingertip to the spinal cord for further processing, here, we have implemented a population of afferent circuits while considering the innervation concept to build receptive fields. Next, we investigate how the collaboration of afferents and nociceptors facilities sharpness recognition. It should be pointed out that utilizing the innervation technique in the prosthetic/robotic applications not only reduces the number of implemented afferents which in turn decreases the cost and power consumption of the neuromorphic devices but also high-resolution tactile sensors can also be handled. Furthermore, by implementing the nociceptors in the proposed tactile neuromorphic system, the concept of pain feeling also emerges. This can provide the prosthesis self-protection to avoid injury during haptic exploration. Indeed, sensor arrays are exposed to damage that can adversely affect the performance of the neuromorphic system⁴⁰. Considering the role of nociceptors and mechanoreceptors, simultaneously, makes the system to be robust against damage in taxels or afferents to some extent. The bio-inspired tactile system includes a population of 23 digital afferents (12 RA-I, 6 SA-I, and 5 nociceptors). Using the proposed system, we explore how the collected spike responses can be used for sharpness classification. In particular, first, the impact of the afferent innervation and creation of receptive field on the firing pattern is investigated. Second, the contribution of tactile afferents and nociceptors on sharpness recognition is explored. Third, the fault tolerance characteristic of the biomimetic system is addressed.

1- FPGA (zynq)

2- pressure tactile sensor (Array)

3- microcontroller (Arduino)

4- robotic hand

1. The human tactile system converts the contact events at the fingertip to trains of action potentials (spikes) and then transmits to upper processing layers. The biological SA-I afferents produce a sustained response to a static indentation of the skin, and the biological RA-I afferents respond only to the onset and offset phases of indentation. Similarly, we have developed a new communication architecture for e-skins that can functionally mimic the behavior of mechanoreceptors and nociceptors.

2. Tactile information is collected from the pressure sensor grid and then transmitted to the neuromorphic system through the interface circuit.

3. A population of 23 afferents (12 RA-I, 6 SA-I, and 5 nociceptors) is digitally realized in the FPGA. The ratio of these two afferents is according to previous research and can be scaled up easily based on the applications.

4. The data which are delivered to the FPGA comes from three groups. Individual digital SA-I afferent receives its inputs directly from the specified receptive field. For each digital RA-I afferent, from its receptive field, the derivative of the input signal is first calculated and then is rectified to be applied to the Izhikevich neuron model. This is due to the fact that based on the biological evidence, the RA-I afferents respond to dynamic skin deformations, hence, for the trapezoidal indentation profile, the RA-I afferents are activated during the onset and offset phases.

5. Individual digital nociceptor also receives the sensor data from all taxels. In this case, we detect the number of taxels (‘NoT’) that exceed the predefined thresholds. Next, the maximum current value (‘MCV’) of the 25 taxels is determined and then the division of ‘MCV’ over ‘NoT’ is calculated (shift to the right in the FPGA).

6. Finally, this value is applied to the Izhikevich neuron model to produce spikes. To analyze the tactile data, all 23 obtained spike trains are transmitted to the Personal Computer (PC) through the Universal asynchronous receiver-transmitter (UART) interface.

1 Abraira, V. E. & Ginty, D. D. The sensory neurons of touch. Neuron 79, 618-639 (2013).

2 Johansson, R. S. & Flanagan, J. R. Coding and use of tactile signals from the fingertips in object manipulation tasks. Nature Reviews Neuroscience 10, 345-359 (2009).

3 Vallbo, A. B. & Johansson, R. S. Properties of cutaneous mechanoreceptors in the human hand related to touch sensation. Hum Neurobiol 3, 3-14 (1984).

4 Pruszynski, J. A. & Johansson, R. S. Edge-orientation processing in first-order tactile neurons. Nature neuroscience 17, 1404-1409 (2014).

5 Scheibert, J., Leurent, S., Prevost, A. & Debrégeas, G. The role of fingerprints in the coding of tactile information probed with a biomimetic sensor. Science 323, 1503-1506 (2009).

6 Smith, E. S. J. & Lewin, G. R. Nociceptors: a phylogenetic view. Journal of Comparative Physiology A 195, 1089-1106 (2009).

7 Dubin, A. E. & Patapoutian, A. Nociceptors: the sensors of the pain pathway. The Journal of clinical investigation 120, 3760-3772 (2010).

8 Farina, D. et al. Man/machine interface based on the discharge timings of spinal motor neurons after targeted muscle reinnervation. Nature biomedical engineering 1, 1-12 (2017).

9 Saal, H. P., Delhaye, B. P., Rayhaun, B. C. & Bensmaia, S. J. Simulating tactile signals from the whole hand with millisecond precision. Proceedings of the National Academy of Sciences 114, E5693-E5702 (2017).

10 Oddo, C. M. et al. Intraneural stimulation elicits discrimination of textural features by artificial fingertip in intact and amputee humans. Elife 5, e09148 (2016).

11 Kim, D.-H. et al. Stretchable and foldable silicon integrated circuits. Science 320, 507-511 (2008).

12 Kim, J. et al. Stretchable silicon nanoribbon electronics for skin prosthesis. Nature communications 5, 1-11 (2014).

13 Larson, C. et al. Highly stretchable electroluminescent skin for optical signaling and tactile sensing. science 351, 1071-1074 (2016).

14 Tee, B. C., Wang, C., Allen, R. & Bao, Z. An electrically and mechanically self-healing composite with pressure-and flexion-sensitive properties for electronic skin applications. Nature nanotechnology 7, 825 (2012).

15 Li, C.-H. et al. A highly stretchable autonomous self-healing elastomer. Nature chemistry 8, 618 (2016).

16 Zou, Z. et al. Rehealable, fully recyclable, and malleable electronic skin enabled by dynamic covalent thermoset nanocomposite. Science advances 4, eaaq0508 (2018).

17 Chun, K. Y., Son, Y. J., Jeon, E. S., Lee, S. & Han, C. S. A Self‐Powered Sensor Mimicking Slow‐and Fast‐Adapting Cutaneous Mechanoreceptors. Advanced Materials 30, 1706299 (2018).

18 Zhao, H., O’Brien, K., Li, S. & Shepherd, R. F. Optoelectronically innervated soft prosthetic hand via stretchable optical waveguides. Science Robotics 1, eaai7529 (2016).

19 Liu, S.-C. & Delbruck, T. Neuromorphic sensory systems. Current Opinion in Neurobiology 20, 288-295 (2010).

20 Laughlin, S. B. & Sejnowski, T. J. Communication in Neuronal Networks. Science 301, 1870 (2003).

21 Kim, S. S. et al. Conveying tactile feedback in sensorized hand neuroprostheses using a biofidelic model of mechanotransduction. IEEE transactions on biomedical circuits and systems 3, 398-404 (2009).

22 Yi, Z. & Zhang, Y. Bio-inspired tactile FA-I spiking generation under sinusoidal stimuli. Journal of Bionic Engineering 13, 612-621 (2016).

23 Friedl, K. E., Voelker, A. R., Peer, A. & Eliasmith, C. Human-inspired neurorobotic system for classifying surface textures by touch. IEEE Robotics and Automation Letters 1, 516-523 (2016).

24 Cassidy, A., Andreou, A. G. & Georgiou, J. Design of a one million neuron single FPGA neuromorphic system for real-time multimodal scene analysis. 45th Annual Conference on Information Sciences and Systems, 1-6 (2011).

25 Escudero, E. C. et al. Real-time neuro-inspired sound source localization and tracking architecture applied to a robotic platform. Neurocomputing 283, 129-139 (2018).

26 Cheung, K., Schultz, S. R. & Luk, W. A Large-Scale Spiking Neural Network Accelerator for FPGA Systems. Artificial Neural Networks and Machine Learning, 113-120 (2012).

27 Wang, R., Hamilton, T. J., Tapson, J. & Schaik, A. v. An FPGA design framework for large-scale spiking neural networks. 2014 IEEE International Symposium on Circuits and Systems (ISCAS), 457-460 (2014).

28 Wang, J., Belatreche, A., Maguire, L. & McGinnity, T. M. An online supervised learning method for spiking neural networks with adaptive structure. Neurocomputing 144, 526-536 (2014).

29 Yang, S. et al. Cost-efficient FPGA implementation of basal ganglia and their Parkinsonian analysis. Neural Networks 71, 62-75 (2015).

30 Merolla, P. A. et al. A million spiking-neuron integrated circuit with a scalable communication network and interface. Science 345, 668 (2014).

31 Nanami, T. & Kohno, T. Simple cortical and thalamic neuron models for digital arithmetic circuit implementation. Frontiers in neuroscience 10, 181 (2016).

32 Wang, R. M., Thakur, C. S. & van Schaik, A. An FPGA-based massively parallel neuromorphic cortex simulator. Frontiers in neuroscience 12, 213 (2018).

33 Zjajo, A. et al. A real-time reconfigurable multichip architecture for large-scale biophysically accurate neuron simulation. IEEE transactions on biomedical circuits and systems 12, 326-337 (2018).

34 Franceschi, M., Camus, V., Ibrahim, A., Enz, C. & Valle, M. Approximate FPGA implementation of CORDIC for tactile data processing using speculative adders. New Generation of CAS (NGCAS), 41-44 (2017).

35 Ambroise, M. et al. Biomimetic neural network for modifying biological dynamics during hybrid experiments. Artificial Life and Robotics 22, 398-403 (2017).

36 Salimi-Nezhad, N., Amiri, M., Falotico, E. & Laschi, C. A digital hardware realization for spiking model of cutaneous mechanoreceptor. Frontiers in neuroscience 12, 322 (2018).

37 Osborn, L. E. et al. Prosthesis with neuromorphic multilayered e-dermis perceives touch and pain. Science robotics 3, eaat3818 (2018).

38 Salimi-Nezhad, N., Ilbeigi, E., Amiri, M., Falotico, E. & Laschi, C. A Digital Hardware System for Spiking Network of Tactile Afferents. Frontiers in Neuroscience 13, 1330 (2019).

39 Hay, E. & Pruszynski, J. A. Orientation processing by synaptic integration across first-order tactile neurons. bioRxiv, 396705 (2018).

40 Teja, S., Mekie, J., Cabibihan, J.-J., Thakor, N. V. & Kukreja, S. L. Fault tolerant tactile sensor arrays for prosthesis. 6th IEEE International Conference on Biomedical Robotics and Biomechatronics (BioRob), 31-34 (2016).

41 Knibestöl, M. Stimulus-response functions of slowly adapting mechanoreceptors in the human glabrous skin area. The Journal of Physiology 245, 63-80 (1975).

42 Delhaye, B. P., Long, K. H. & Bensmaia, S. J. Neural basis of touch and proprioception in primate cortex. Comprehensive Physiology 8, 1575-1602 (2011).

43 Kim, E. K., Wellnitz, S. A., Bourdon, S. M., Lumpkin, E. A. & Gerling, G. J. Force sensor in simulated skin and neural model mimic tactile SAI afferent spiking response to ramp and hold stimuli. Journal of neuroengineering 9, 1-14 (2012).

44 Rongala, U. B., Mazzoni, A. & Oddo, C. M. Neuromorphic artificial touch for categorization of naturalistic textures. IEEE transactions on neural networks and learning systems 28, 819-829 (2015).

45 Victor, J. D. & Purpura, K. P. Metric-space analysis of spike trains: theory, algorithms and application. Network: computation in neural systems 8, 127-164 (1997).

46 Izhikevich, E. M. Simple model of spiking neurons. IEEE Transactions on neural networks 14, 1569-1572 (2003).

47 Lumelsky, V. J., Shur, M. S. & Wagner, S. Sensitive skin. IEEE Sensors Journal 1, 41-51 (2001).

48 Mittendorfer, P. & Cheng, G. Humanoid Multimodal Tactile-Sensing Modules. IEEE Transactions on Robotics 27, 401-410 (2011).

49 Farserotu, J. et al. Smart skin for tactile prosthetics. 6th International Symposium on Medical Information and Communication Technology (ISMICT), 1-8 (2012).

50 Cassidy, A. & Andreou, A. G. Dynamical digital silicon neurons. IEEE Biomedical Circuits and Systems Conference, 289-292 (2008).

51 Vargas-Irwin, C. E., Brandman, D. M., Zimmermann, J. B., Donoghue, J. P. & Black, M. J. Spike train SIMilarity Space (SSIMS): a framework for single neuron and ensemble data analysis. Neural computation 27, 1-31 (2015).

The authors would like to thank the esteemed reviewers for their insightful and helpful comments. M. A would like to thank Prof. Nitish Thakor for his valuable discussion. M. A, A. P-F, and N. S-N would like to appreciate Mr. Behnam Rostamian for his assistance in 3D printing and performing the experiments.

Download PDF

Version 1

posted

You are reading this latest protocol version

Sharpness Recognition Based on Synergy between Bio-inspired Nociceptors and Tactile Mechanoreceptors

Status:

Version 1

Abstract

Introduction

Reagents

Procedure

References

Acknowledgements

Associated Publications

Status:

Version 1

Privacy Policy

Terms of Service

Sharpness Recognition Based on Synergy between Bio-inspired Nociceptors and Tactile Mechanoreceptors

Status:

Version 1

Abstract

Introduction

Reagents

Procedure

References

Acknowledgements

Associated Publications

Status:

Version 1

Privacy Policy

Terms of Service

Manage Cookie Preferences