Chuck Larson 1 , Donald Robin 2 , Jack Lancaster 2 , Peter Fox 2

1 Northwestern University, Communication Sciences and Disorders, Evanston

2 University of Texas Health Science Center, Research Imaging Center, San Antonio

Question:
The goal of the present study was to test the hypothesis that the laryngeal motor cortex is involved in generating vocal responses to perturbations in auditory feedback. It is well established that vocalizing subjects respond to pitch-shifted voice auditory feedback with a compensatory change in voice fundamental frequency (F0).  A previous study showed that a pitch-shift stimulus elicits a cricothyroid muscle and voice F0 response in 50 and 80 ms respectively. Based on these timing constraints, it was predicted that the cortex should become active in less than 50 ms following a pitch-shift stimulus. 

Methods:
In the present study, the laryngeal area of motor cortex was first identified from a sparse-sampling fMRI scanning protocol as subjects vocalized and received pitch-shifted voice feedback. A structural image of the brain was also created using fMRI techniques.  The coordinates of laryngeal motor cortex were then fed to a robotically controlled TMS system. The TMS stimulator coil was then positioned over laryngeal motor cortex, the position of which was verified by recording EMG potentials from the cricothyroid muscle with latencies of 10 – 12 ms following cortical TMS. Subjects then sustained vowel phonations while receiving short duration (200 ms) increases in pitch-shifted voice feedback in real time. Increases in pitch-shifted voice feedback generally cause a decrease in cricothyroid EMG and voice F0. Single TMS pulses timed to occur at specific intervals following pitch-shifted voice feedback (0, 10, 20, 30, 40, 50, 60 ms), caused an initial increase in voice F0 followed by a decrease. For each pitch-shift --- TMS interval condition, the average magnitude of the initial increase in the F0 was measured.

Results:
Reductions in the magnitude of the F0 response occurred with pitch-shift --- TMS intervals of 30 – 40 ms.  The reduction is thought to result from an interaction between the TMS pulse and the decrease in neural activity elicited by the upward pitch-shift stimulus.

Conclusions:
This interval coincides with the predicted time of laryngeal motor cortex excitation necessary to produce an increase in voice F0 following a pitch-shift stimulus. These results confirm the hypothesis that the pitch-shift reflex involves the cerebral cortex. 

Ken Roßlau 1 , Sibylle Herholz 2 , Arne Knief 1 , Christian Dobel 2 , Christo Pantev 2 , Antoinette am Zehnhoff-Dinnesen 1

1 Department of phoniatrics and pediatric audiology, university hospital, Münster

2 Institute of Biomagnetism and Biosingalanalysis, university hospital, Münster

Question

It is a speciality of professional actors and singers to learn text fragments very fast and in addition for singers to do that in melodic context. We suppose that there could be a difference in recognition of faults in sungen or spoken rhymes between singers and actors. So the actors should be able to recognize mainly semantic mistakes even in the sung and spoken version, while singers split their attention and can detect the semantic and melodic incongruities. Combinated incongruity of semantic and pitch mistakes in french operatic songs elicited both an increased N400 and P600 eeg-potential as evidence for an independent processing of melodic and semantic aspects in language (Besson et al., 2001). We suppose differences between actors and singers in elektrophysiological and behavioral data during processing fragments of spoken and sung german SCHUBERT songs.

Methods

We prepared 240 short fragments from german SCHUBERT songs in 4 different conditions of the last word in sung and spoken version. In the first condition the semantic and melodic ending was correct, in the second and third either there was a semantic or melodic fault and in the fourth there was a double incongruity. The subjects have to detect the kind of mistake in the ending. We collect behavioral data and measure EEG/MEG potentials to show cortical processing strategies.

  

Results

A pilot study with 8 non-musician psychology students showed better results in all conditions of the spoken version compared to the sung version. Only the condition with double incongruity showed that there was a problem to detect weak pitch increases in the spoken form. In the sung version best results were reached during semantic and melodic correctness. The most difficult condition to recognize for the subjects was a semantic fault during melodic correctness in the sung version. The different 4 conditions in both versions were rated normal current.

Discussion

The mixture of semantic and melodic faults seemed to be a special problem for non-musicians in the sung version. Previous studies showed different cortical correlates  for processing semantic and/or meldodic information in language which depended on musical education. Our study showed that the used complex stimulus material is appropriate to elucidate a N400 response especially in actors.  A professional singer should be best in the condition of double incongruity and present high neuronal correlates because they have have both: The ability to recognize weak melody changes and semantic faults.

 

Katrin Neumann 1 , Christian Kell 2 , Philippe Dejonckere 2,3

1 University of Frankfurt, Dept. of Phoniatrics and Pediatric Audiology, Frankfurt am Main

2 University of Frankfurt, Dept. of Neurology, Frankfurt am Main

3 University Medical Center Utrecht, Institute of Phoniatrics, Utrecht

Objective: To study with fMRI the cerebral pattern of sensorimotor activation in patients with spasmodic dysphonia before and after a Botulinum toxin therapy. Former studies used somewhat unnatural voicing conditions of sustained vowels and found reduced activation of premotor, sensorimotor, and sensory association regions during vocalization before as well as after therapy (Haslinger et al., 2005).

Methods: The authors performed fMRI recordings during (a) reading aloud and (b) silently, as well as (c) during whispering and (d) during a motor task (pressing) in 12 patients with spasmodic dysphonia. Subjects were scanned before and after treatment with injections of botulinum toxin into the vocalis muscle bilaterally. The tasks aimed on the evaluation of cerebral activation during (a) more (pre-treatment) or less (post-treatment) dystonic symptoms  while reading aloud, (b) reading without voicing as contrast condition, (c) symptom-free speech, and (d) imitation of vocal fold spasms during pure motor execution. To avoid movement artifacts during the oral motor tasks, an event-related fMRI design was used involving non-continuous sampling with no data acquisition during task performance.

Results: The data analysis is not yet completed. It can be expected, however, that (1) because of the use of “natural-like” speech material instead of sustained vowels, no reduced activation of premotor, sensorimotor and sensory association regions occur during vocalization in patients with spasmodic dysphonia, neither pre-treatment nor post-treatment, (2) therapy reduces an abnormal cerebral activation of voicing without deleting it completely, (3) activation pattern of the pressing condition resembles that one of speaking before therapy, and (3) activation pattern of whispering still shows deviations from those of normal voicing subjects.

Reference: Haslinger B, Erhard P, Dresel C, Castrop F, Roettinger M, Ceballos-Baumann AO (2005) “Silent event-related” fMRI reveals reduced sensorimotor activation in laryngeal dystonia. 65, 1562-1569.

Matthias Echternach 1 , Nadja Räss 2 , Johan Sundberg 3 , Bernhard Richter 4

1 Institut of Musicians Medicine, Freiburg University Medical Center, Freiburg

2 Jodellehrerin, , Einsiedeln

3 Department of Speech, Music and Hearing, KTH, Stockholm

4 Institue of Musicians Medicine, Freiburg University Medical Center, Freiburg

Introduction: Yodelling is a special kind of vocal performance in traditional music, particularly in Switzerland. It consists of rapidly and repeatedly changes in pitch. It is assumed that these pitch changes are caused by register changes. However, until now only few studies on yodelling were published.  Material and Methods: We analysed 4 professional yodelling teachers (2male, 2 female) performing yodelling on different vowels with Audio and EGG signal using the Laryngograph. Additionally, vocal tract configuration was analysed with dynamic real time MRI. Results:  Preliminary results reveal that pitch changes in yodelling are associated with decrease of EGG contact quotient at the higher pitch, indicating a register shift. Further more, in contrast to non yodelling volunteers, the lower and upper pitches were much more stable and the register change was faster. Vocal tract shape was mainly changed due to vowel shifts, typically performed in yodelling. Additionally, in special yodelling techniques we found very fast tongue movements. Conclusion: Yodeling is associated with very fast changes of both, vocal fold vibration patterns and the vocal tract.