Performance of the Piecewise Rate Modulation Strategy to Convey Mandarin Tonal Information for Cochlear Implant Using Neural Network Simulation

2011 ◽  
Vol 474-476 ◽  
pp. 1049-1052
Author(s):  
Tian Guan ◽  
Qin Gong ◽  
Tong Zhou
Keyword(s):  
Neural Network ◽  
Cochlear Implant ◽  
Spectral Information ◽  
Rate Variation ◽  
Tonal Language ◽  
Rate Modulation ◽  
Stimulus Rate ◽  
Monosyllabic Words ◽  
Electric Stimulator ◽  
Mandarin Tones

In order to improve the pitch perception of cochlear implant (CI) users speaking tonal language, it has been suggested to frequency-modulate the electric stimulus rate by the spectral information of the tonal language. A piecewise CI rate modulation strategy has been recently proposed, which not only encoded the spectral information but also took account of the psychological perception feature for the stimulus rate variation by CI users. This paper further examines its performance to convey Mandarin tonal information by a neural-network-based simulation. The experimental results shown that the correct rates to identify the four Mandarin tones of 80 Mandarin monosyllabic words were 95%, 95%, 100% and 100%, respectively, indicating that the piecewise rate modulation strategy might efficiently convey Mandarin tonal information. Therefore, the piecewise rate modulation strategy could help to design novel CI electric stimulator and enhance the speech perception ability of CI users speaking tonal language, such as Mandarin.

Speech Recognition as a Function of the Number of Electrodes Used in the SPEAK Cochlear Implant Speech Processor

1997 ◽  
Vol 40 (5) ◽  
pp. 1201-1215 ◽  
Author(s):  
Kim E. Fishman ◽  
Robert V. Shannon ◽  
William H. Slattery
Keyword(s):  
Speech Recognition ◽  
Cochlear Implant ◽  
Speech Processing ◽  
Recognition Performance ◽  
Spectral Information ◽  
Processing Strategy ◽  
Speech Processor ◽  
Average Performance ◽  
Speech Stimuli ◽  
Monosyllabic Words

Speech recognition was measured in listeners with the Nucleus-22 SPEAK speech processing strategy as a function of the number of electrodes. Speech stimuli were analyzed into 20 frequency bands and processed according to the usual SPEAK processing strategy. In the normal clinical processor each electrode is assigned to represent the output of one filter. To create reduced-electrode processors the output of several adjacent filters were directed to a single electrode, resulting in processors with 1, 2, 4, 7, 10, and 20 electrodes. The overall spectral bandwidth was preserved, but the number of active electrodes was progressively reduced. After a 2-day period of adjustment to each processor, speech recognition performance was measured on medial consonants, vowels, monosyllabic words, and sentences. Performance with a single electrode processor was poor in all listeners, and average performance increased dramatically on all test materials as the number of electrodes was increased from 1 to 4. No differences in average performance were observed on any test in the 7-, 10-, and 20-electrode conditions. On sentence and consonant tests there was no difference between average performance with the 4-electrode and 20-electrode processors. This pattern of results suggests that cochlear implant listeners are not able to make full use of the spectral information on all 20 electrodes. Further research is necessary to understand the reasons for this limitation and to understand how to increase the amount of spectral information in speech received by implanted listeners.

Perceptual Learning of Vocoded Speech With and Without Contralateral Hearing: Implications for Cochlear Implant Rehabilitation

2020 ◽  
pp. 1-10
Author(s):  
Martin Chavant ◽  
Alexis Hervais-Adelman ◽  
Olivier Macherey
Keyword(s):  
Cochlear Implant ◽  
Perceptual Learning ◽  
Speech Signal ◽  
Training Phase ◽  
Monosyllabic Words ◽  
Low Pass ◽  
Contralateral Ear ◽  
Number Of Individuals ◽  
Insight Into ◽  
Vocoded Speech

Purpose An increasing number of individuals with residual or even normal contralateral hearing are being considered for cochlear implantation. It remains unknown whether the presence of contralateral hearing is beneficial or detrimental to their perceptual learning of cochlear implant (CI)–processed speech. The aim of this experiment was to provide a first insight into this question using acoustic simulations of CI processing. Method Sixty normal-hearing listeners took part in an auditory perceptual learning experiment. Each subject was randomly assigned to one of three groups of 20 referred to as NORMAL, LOWPASS, and NOTHING. The experiment consisted of two test phases separated by a training phase. In the test phases, all subjects were tested on recognition of monosyllabic words passed through a six-channel “PSHC” vocoder presented to a single ear. In the training phase, which consisted of listening to a 25-min audio book, all subjects were also presented with the same vocoded speech in one ear but the signal they received in their other ear differed across groups. The NORMAL group was presented with the unprocessed speech signal, the LOWPASS group with a low-pass filtered version of the speech signal, and the NOTHING group with no sound at all. Results The improvement in speech scores following training was significantly smaller for the NORMAL than for the LOWPASS and NOTHING groups. Conclusions This study suggests that the presentation of normal speech in the contralateral ear reduces or slows down perceptual learning of vocoded speech but that an unintelligible low-pass filtered contralateral signal does not have this effect. Potential implications for the rehabilitation of CI patients with partial or full contralateral hearing are discussed.

Differences in Children’s Sound Production When Speaking With a Cochlear Implant Turned On and Turned Off

1996 ◽  
Vol 39 (3) ◽  
pp. 604-610 ◽  
Author(s):  
Nancy Tye-Murray ◽  
Linda Spencer ◽  
Elizabeth Gilbert Bedia ◽  
George Woodworth
Keyword(s):  
Cochlear Implant ◽  
Speech Production ◽  
Cochlear Implants ◽  
Sound Production ◽  
Proprioceptive Feedback ◽  
Speech Sample ◽  
Speech Language Pathologists ◽  
Monosyllabic Words ◽  
Consonant Voicing ◽  
Vowel Height

Twenty children who have worn a Cochlear Corporation cochlear implant for an average of 33.6 months participated in a device-on/off experiment. They spoke 14 monosyllabic words three times each after having not worn their cochlear implant speech processors for several hours. They then spoke the same speech sample again with their cochlear implants turned on. The utterances were phonetically transcribed by speech-language pathologists. On average, no difference between speaking conditions on indices of vowel height, vowel place, initial consonant place, initial consonant voicing, or final consonant voicing was found. Comparisons based on a narrow transcription of the speech samples revealed no difference between the two speaking conditions. Children who were more intelligible were no more likely to show a degradation in their speech production in the device-off condition than children who were less intelligible. In the device-on condition, children sometimes nasalized their vowels and inappropriately aspirated their consonants. Their tendency to nasalize vowels and aspirate initial consonants might reflect an attempt to increase proprioceptive feedback, which would provide them with a greater awareness of their speaking behavior.

scholarly journals Speech intelligibility enhancement for Thai-speaking cochlear implant listeners

2019 ◽  
Vol 13 (3) ◽  
pp. 866
Author(s):  
Siriporn Dachasilaruk ◽  
Niphat Jantharamin ◽  
Apichai Rungruang
Keyword(s):  
Cochlear Implant ◽  
Speech Enhancement ◽  
Speech Intelligibility ◽  
English Language ◽  
Single Channel ◽  
Spectral Subtraction ◽  
Monosyllabic Words ◽  
Listening Environments ◽  
Babble Noise ◽  
Vocoded Speech

Cochlear implant (CI) listeners encounter difficulties in communicating with other persons in noisy listening environments. However, most CI research has been carried out using the English language. In this study, single-channel speech enhancement (SE) strategies as a pre-processing approach for the CI system were investigated in terms of Thai speech intelligibility improvement. Two SE algorithms, namely multi-band spectral subtraction (MBSS) and Weiner filter (WF) algorithms, were evaluated. Speech signals consisting of monosyllabic and bisyllabic Thai words were degraded by speech-shaped noise and babble noise at SNR levels of 0, 5, and 10 dB. Then the noisy words were enhanced using SE algorithms. The enhanced words were fed into the CI system to synthesize vocoded speech. The vocoded speech was presented to twenty normal-hearing listeners. The results indicated that speech intelligibility was marginally improved by the MBSS algorithm and significantly improved by the WF algorithm in some conditions. The enhanced bisyllabic words showed a noticeably higher intelligibility improvement than the enhanced monosyllabic words in all conditions, particularly in speech-shaped noise. Such outcomes may be beneficial to Thai-speaking CI listeners.

Activation of type-identified motor units during centrally evoked contractions in the cat medial gastrocnemius muscle. III. Muscle-unit force modulation

1996 ◽  
Vol 75 (1) ◽  
pp. 51-59 ◽  
Author(s):  
K. E. Tansey ◽  
A. K. Yee ◽  
B. R. Botterman
Keyword(s):  
Firing Rate ◽  
Motor Unit ◽  
Muscle Force ◽  
Motor Units ◽  
Medial Gastrocnemius ◽  
Rate Variation ◽  
Unit Force ◽  
Rate Modulation ◽  
Force Modulation ◽  
Whole Muscle

1. The aim of this study was to examine the extent of muscle-unit force modulation due to motoneuron firing-rate variation in type-identified motor units of the cat medial gastrocnemius (MG) muscle, and to investigate the contribution of muscle-unit force modulation to whole-muscle force regulation. The motoneuron discharge patterns recorded from 8 pairs of motor units during 12 smoothly graded muscle contractions evoked by stimulation of the mesencephalic locomotor region (MLR) were used to reactivate those units in isolation to estimate what their force profiles would have been like during the evoked whole-muscle contractions. 2. For most motor units, muscle-unit force modulation was similar to motoneuron firing-rate modulation, in that muscle-unit force increased over a limited range (120-600 g) of increasing whole-muscle tension and was then maintained at a near maximal (> 70%) output level as muscle force continued to rise. Most muscle units also decreased their force outputs over a slightly larger range of declining whole-muscle force before relaxing. This second finding was best explained by the counterclockwise hysteresis recorded in the motor units' frequency-tension (f-t) relationships. 3. In those instances when whole-muscle force fluctuated just above the recruitment threshold of a motor unit, a substantial percentage (10-25%) of the change in whole-muscle force could be accounted for by force modulation in that motor unit alone. This finding suggested that few motor units in the pool were simultaneously simultaneously undergoing force modulation. To evaluate this possibility, the extent of parallel muscle-unit force modulation within the 8 pairs of simultaneously active motor units was evaluated. As with parallel motoneuron firing-rate modulation, the extent of parallel muscle-unit force modulation was limited to unit pairs of the same physiological type and recruitment threshold. In several instances, pairs of motor units displayed parallel motoneuron firing-rate modulation but did not show parallel muscle-unit force modulation because of the nature of the motor units' f-t relationships. 4. The limited extent of parallel muscle-unit force modulation seen in these experiments implies that the major strategy for force modulation in the cat MG muscle, involving contractions estimated to reach 30-40% of maximum, may be motor-unit recruitment rather than motor-unit firing-rate variation with resulting force modulation. Given, however, that the majority of motor units are already recruited at these output levels (< 40%), it is proposed that motor-unit firing-rate variation with resulting force modulation may take over as the major muscle force modulating strategy at higher output levels.

Contributions of motor-unit recruitment and rate modulation of compensation for muscle yielding

1982 ◽  
Vol 47 (5) ◽  
pp. 797-809 ◽  
Author(s):  
P. J. Cordo ◽  
W. Z. Rymer
Keyword(s):  
Motor Unit ◽  
Stretch Reflex ◽  
Muscle Force ◽  
Motor Units ◽  
Muscle Stiffness ◽  
Motor Unit Recruitment ◽  
Reflex Response ◽  
Rate Modulation ◽  
Stimulus Rate ◽  
Wide Range

1. Subdivided portions of the cut ventral root innervation of the soleus muscle were electrically stimulated in 14 anesthetized cats. The stimulus trains imposed on these nerves simulated the recruitment and rate-modulation patterns of single motor units recorded during stretch-reflex responses in decerebrate preparations. Each activation pattern was evaluated for its ability to prevent muscle yield. 2. Three basic stimulus patterns, recruitment, step increases in stimulus rate, and doublets were imposed during the course of ramp stretches applied over a wide range of velocities. The effect of each stimulus pattern on muscle force was compared to the force output recorded without stretch-related recruitment or rate modulation. 3. Motor-unit recruitment was found to be most effective in preventing yield during muscle stretch. Newly recruited motor units showed no evidence of yielding for some 250 ms following activation, at which time muscle stiffness declined slightly. This time-dependent resistance to yield was observed regardless of whether the onset of the neural stimulus closely preceded or followed stretch onset. 4. Step increases in stimulus rate arising shortly after stretch onset did not prevent the occurrence of yield at most stretch velocities, but did augment muscle stiffness later in the stretch. Doublets in the stimulus train were found to augment muscle stiffness only when they occurred in newly recruited motor units. 5. These results suggest that at low or moderate initial forces, the prevention of yield in lengthening, reflexively intact muscle results primarily from rapid motor-unit recruitment. To a lesser extent, the spring-like character of the stretch-reflex response also derives from step increases in firing rate of motor units active before stretch onset and doublets in units recruited during the course of stretch. Smooth rate increases appear to augment muscle force later in the course of the reflex response.

A Recurrent Neural Network Approach for Automated Neural Tracing in Multispectral 3D Images

2018 ◽  
Author(s):  
Yan Yan ◽  
Douglas H. Roossien ◽  
Benjamin V. Sadis ◽  
Jason J. Corso ◽  
Dawen Cai
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Neuronal Cell ◽  
Ground Truth ◽  
Neuronal Morphology ◽  
Spectral Information ◽  
Neural Net ◽  
Learning Approaches ◽  
3D Images ◽  
Neural Tracing

AbstractNeuronal morphology reconstruction in fluorescence microscopy 3D images is essential for analyzing neuronal cell type and connectivity. Manual tracing of neurons in these images is time consuming and subjective. Automated tracing is highly desired yet is one of the foremost challenges in computational neuroscience. The multispectral labeling technique, Brainbow utilizes high dimensional spectral information to distinguish intermingled neuronal processes. It is particular interesting to develop new algorithms to include the spectral information into the tracing process. Recently, deep learning approaches achieved state-of-the-art in different computer vision and medical imaging applications. To benefit from the power of deep learning, in this paper, we propose an automated neural tracing approach in multispectral 3D Brainbow images based on recurrent neural net-work. We first adopt VBM4D approach to denoise multispectral 3D images. Then we generate cubes as training samples along the ground truth, manually traced paths. These cubes are the input to the recur-rent neural network. The proposed approach is simple and effective. The approach can be implemented with the deep learning toolbox ‘Keras’ in 100 lines. Finally, to evaluate our approach, we computed the average and standard deviation of DIADEM metric from the ground truth results to our tracing results, and from our tracing results to the ground truth results. Extensive experimental results on the collected dataset demonstrate that the proposed approach performs well in Brainbow labeled mouse brain images.

Sign in / Sign up

Export Citation Format

Share Document