Nil.vp

Response time as an index for
selective auditory cognitive deficits
Reza Nilipour1,2, Stephanie Clarke3, Behrad Noudoost2,
Golbarg Tarighat Saber2 and Abdolrahman Najlerahim4
1Department of Speech Therapy, University of Welfare and RehabilitationSciences, Kudakyar Ave., Evin, 19834 Tehran, Iran; 2School of CognitiveSciences, IPM, Niavaran Square, P.O. Box 19395-5746, Tehran, Iran;3Division de Neuropsychologie, CHUV, Rue de Bugnon 46, 1011 Lausanne,Switzerland; 4Division of Neurology, Shohada Hospital, Medical Universityof Shahid Beheshti, Tehran, Iran Abstract. The full or partial recovery of cognitive functions following brain
lesions is believed to rely on the recruitment of alternative neural networks.
This has been shown anatomically for selective auditory cognitive functions
(Adriani et al. 2003b). We investigate here behavioral correlates that may
accompany the use of alternative processing networks and in particular the
resulting increase in response times. The performance of 5 patients with right
or left unilateral hemispheric infarction and 6 normal control subjects in
sound identification, asemantic sound recognition, sound localization, and
sound motion perception was evaluated by the number of correct replies and
response times for correct and wrong replies. Performance and response times
were compared across patients and normal control subjects. Two patients with
left lesions were deficient in sound identification and sound motion
perception and normal in sound localization and asemantic sound recognition;
one patient with right lesion was deficient in sound localization and sound
motion perception and normal in sound identification and asemantic sound
recognition; deficient performance was associated with increased response
times. The remaining 2 patients (1 with left, 1 with right lesion) had normal
performance in all 4 tasks but had significantly longer response times in some
(but not all) tasks. Patients with normal or deficient performance tended more
often than normal subjects to give faster correct than wrong replies. We
propose that increased response time is an indication of processing within an
alternative network.
The correspondence should beaddressed to R. Nilipour, Key words: response time, sound identification, sound localization, parietal
lesion, auditory "what" and "where" processing streams INTRODUCTION
(Adriani et al. 2003a). Follow-up evaluations showedthat most of the deficits associated with these small le- Several lines of evidence indicate that sound recogni- sions recovered within the next months (Rey et al. 2003).
tion and sound localization are processed in two at least Activation studies on motricity (see Calautti and partially independent processing streams. In non-hu- Baron 2003 for review) and language (Blasi et al. 2002, man primates the non-primary auditory areas of the belt Cao et al. 1999, Karbe et al. 1998, Ohyama et al. 1996, were shown to contain predominantly neurons respond- Rosen et al. 2000, Warburton et al. 1999, Weiller et al.
ing to animal-cry like stimuli (area AL) or to the spatial 1995) have shown that full or partial recovery of func- location of the sound (area PL) (Rauschecker and Tian tion following brain lesions relies on the recruitment of 2000). In man fMRI (Alain et al. 2001, Maeder et al.
alternative neural networks. The processing within 2001) and electrophysiological investigations (Alain et these networks is very likely to be less efficient than in al. 2001, Anourova et al. 2001) suggest a similar dichot- normals and be associated thus with subnormal perfor- omy. In particular the recognition of significant audi- mance and/or increased processing times. The latter can tory stimuli activated selectively bilaterally regions on be assessed by means of response times in specific cog- the temporal convexity and sound localization bilater- nitive tasks. Most neuropsychological studies charac- ally parietal and frontal foci (Maeder et al. 2001). Rela- terize deficient performance by the insufficient number tively large lesions centered on either of these networks of correct replies as compared to the normal population were shown to be associated in the chronic stage with and/or by excessively slow responses (Lezak 1995, the corresponding deficit both after right (Clarke et al.
Saygin et al. 2003). In online paradigms differences in 2002, Fujii et al. 1990, Griffiths et al. 1996, 1997, response time between correct and wrong replies have Spreen et al. 1965), left (Clarke et al. 2000) or bilateral been interpreted as an additional measure of perfor- hemispheric lesions (Jerger et al. 1972, Rosati et al.
mance (Buttet Sovilla and Grosjean 1997). This re- 1982). Smaller lesions were shown to produce in the search seeks to investigate whether increasing response acute stage auditory cognitive deficits, but these deficits time is due to a general slowness in cognitive functions did not always correspond to the specialization of the as proposed by Saygin et al. (2003) or can be interpreted damaged network, e.g., small lesions centered on the as an additional measure of reorganization of a deficient recognition network were found to be associated in the processing in chronic patients. This is a first report using acute stage with a selective deficit in sound localization these two approaches to describe the performance of (M) male; (F) female; (R) right; (L) left, (m) month. Lesions have been assessed on CT scans.
Auditory cognitive deficits and response times 165
brain-damaged patients in spatial and nonspatial audi- locations: extreme right (0.6 ms ITD in favor of the right ear), right (0.3 ms), center, left (-0.3 ms), and extremeleft (-0.6 ms) The subjects were asked to choose one of the five positions marked on a head drawing shown onthe screen by indicating the location of the presented Six healthy control subjects (4 male and 2 female, sound by pushing the proper key on the keyboard. The mean age 42) with no history of neurological or hearing left and right Ctrl keys were assigned for the extreme disease and five right-handed patients with a first unilat- left and extreme right positions respectively, left and eral hemispheric stroke (Table I) were included in this right Alt keys for the left and right hemispace positions study. The patients corresponded to consecutive cases respectively, and the Space-bar key in the center for the from a speech therapy clinic that fulfilled the following stimuli presented in the center. Four samples were criteria: (i) unilateral hemispheric stroke; (ii) no previ- presented as training trials prior to the task.
ous brain damage; (iii) normal hearing; and (iv) absence In the motion task, the sound stimulus, 2.3 s "bumble- of major behavioral disturbances. Two patients sus- bee" sound shaped with 100 ms rising and falling times, tained a unilateral right (MF and KH) and three a left le- moved from one location to another using simulation sion (PA, HA and GT). The onset of the lesion was with progressive changes of ITD. Four types of moving between 7 to 35 months prior to the auditory cognitive paths and two speed conditions were used in this task: testing reported here. They were evaluated for the extreme right to left, extreme left to right, right to center neuropsychological and neurolinguistic deficits using a and left to center. The starting and finishing positions standard aphasia test (Nilipour 1993) as part of their were defined by ITD as in the localization task. The ex- neurorehabilitation program for speech-language ther- treme left to right and extreme right to left were pre- apy. Patients HA, PA and GT, but not MF and KH had a sented with fast speed and the left to center and right to moderate general language deficit profile subsequent to center with slow speed. The speed rate of the fast condi- the lesion and CVA history, but did not suffer from a tion was 40 degrees/s and the slow condition 20 de- grees/s. The subjects were asked to identify the We presented two sets of auditory tasks in this experi- direction and the trajectory of the moving sound they ment: two auditory spatial tasks and two auditory recog- heard via the headphones by pushing the proper key.
nition tasks. The tasks were cultural adaptations of tasks The left and right Ctrl keys were assigned for right to left used in previous studies (Bellmann et al. 2001, Clarke et and left to right stimuli respectively, left and right Alt al. 1996, 2000). The adapted auditory tasks were per- keys for the left to center and right to center positions re- spectively. Four trials were presented prior to the task.
www.neurobs.com) and were run on a Pentium III PC.
The sound recognition tasks included a sound identi- The number of correct replies as well as the response fication task and an asemantic task with a similar design which have explained in previous research (Clarke et al.
The auditory spatial tasks included a localization task 2000). In the sound identification task, the auditory and a motion task. Previous studies of brain-damaged sound object lasted 7 s was presented on the headphones patients have tested the ability to localize sound sources with a multiple-choice drawing presented on the screen either in free-field condition, i.e., with loudspeaker simultaneously. The subject was asked to select the vi- placed in different locations around the patient, or with sual object corresponding to the sound from five draw- simulations using interaural time (Altman et al. 1979, Bisiach et al. 1984) or intensity differences (Sterzi et semantically but not acoustically related (saw); an al.1996). As in our previous studies, we have used here acoustically but not semantically related (clock); a se- spatial simulations with interaural time differences mantically and acoustically related (hatchet); and an un- (ITD) (Adriani et al. 2003a, Bellmann et al. 2001, related object (ship). Five keys were labeled with 1 to 5 Bellmann Thiran and Clarke 2003, Clarke et al. 2000, which would correspond to the number of the selected 2002). In the auditory localization task, a stationary au- ditory stimulus, a 2 second broadband bumblebee In the asemantic task, two consecutive non-identical sound, shaped with 100 ms rising and falling times, was environmental sounds were presented to the subject via presented to the subject via headphones at five different headphones. The subjects were asked to decide whether the two consecutive sounds belonged to the same or dif- ied between 116 ms (mean response time to correct re- ferent sound objects by pushing the key labeled "yes" or plies in the asemantic recognition tests) and 647 ms "no" on the keyboard. Four training trials were per- (mean response time to wrong replies in the sound identi- fication test). The response times when giving correct Response times were measured from the offset of the versus wrong replies were significantly different for stimulus (the second stimulus in the asemantic recogni- sound motion perception, almost significantly different tion task). The normal distribution of the response times for sound identification, and not significantly different for each test and each subject was tested with p-p plots.
for sound localization and asemantic sound recognition.
Since response times had normal distribution a paramet- Three patients presented selective deficits as assessed ric t-test was performed to compare response times in by the number of correct replies. Patient MF was defi- each case. The clustering of patients and normal sub- cient in auditory spatial functions (i.e., his scores were jects based on their response times in each task was per- significantly different from those of normal subjects), formed using ANOVA test and Tukey HSD post-hoc both in sound localization (P=0.022) and sound motion perception (P=0.006), while sound identification andasemantic sound recognition were within normal limits (P=0.816 and 0.667, respectively). His response timeswere significantly slower than those of normal subjects The performance of normal subjects in sound identifi- in the deficient domains, i.e., sound localization and cation, asemantic sound recognition, sound localization sound motion perception, as well as in the apparently and sound motion perception is summarized in Table II.
preserved sound identification, but not in asemantic None of the tests had a ceiling effect. Response times var- sound recognition (Fig. 1). Within a given task, re- lm
so
da PA
lm
mm
da ms
ms
ab
so
pm
lm
lm
so
so
da
ms PA
da
ms
Fig. 1. Clustering patients and normal subjects based on response times in the four auditory tasks; (using ANOVA test andpost-hoc Tukey HSD). Initials identify patients (upper case) and normal subjects (lower case). Normal subjects and patients en-tered the same contour are not statistically different from each other.
Auditory cognitive deficits and response times 167
sponse times of correct and of wrong replies did not dif- were significantly slower than those of normal subjects fer significantly (Table II). Detailed analysis of the type in the deficient domain (sound identification; response of errors in sound localization revealed a rightward shift times for motion perception have not been assessed), which was not present in normal subjects.
but not in the preserved sound localization and the rela- Patient HA was deficient in sound identification and tively preserved asemantic sound recognition (Fig. 1).
sound motion perception (P=0.000 for both), while Within a given task, response times of correct and of sound localization was within normal limits (P=0.256) wrong replies did not differ significantly for sound lo- and asemantic sound recognition at the lower limit of calization, sound motion perception and asemantic normal performance (P=0.074). His response times sound recognition. For sound identification correct re- Performance of patients and normal subjects and related response times (RT) Mean of normal subjects
Number of correct replies
Patient MF
Number of correct replies
Patient HA
Number of correct replies
Patient PA
Number of correct replies
Patient KH
Number of correct replies
Patient GT
Number of correct replies
(#) statistically significant differences in response times associated with correct vs. wrong replies; (*) deficient performance or significantly slower response times in brain-damaged patients as compared to normal subjects plies were significantly faster than wrong replies. De- identification, correct replies were significantly faster tailed analysis of the type of errors in the sound identification task revealed a specific weakness in thesemantic domain. This patient gave 25 wrong replies of DISCUSSION
which 9 were semantically but not acoustically relatedto the target, 3 acoustically but not semantically, 9 se- Our results demonstrate that deficient performance in mantically and acoustically and 4 unrelated (the range terms of correct replies tends to be accompanied by in- of these errors in normal subjects was, respectively, 0-2, creased response times. A similar inverse relationship be- 0-1, 3-8, and 0-2). Furthermore, he gave wrong replies tween the number of correct replies and response time has which were never given by normal subjects and which been recently described for both verbal and non-verbal revealed confusion between acoustic and semantic cate- auditory recognition in aphasic patients (Saygin et al.
gories: e.g., when he heard a "hammer" he chose a "saw" 2003). In this study poor performance in verbal and non-verbal domains tended to be associated, an observa- Patient PA was deficient in sound identification and tion which was interpreted in terms of shared processing sound motion perception (P=0.003 and 0.000, respec- networks for the two functions. In our study, three out of tively), but normal in asemantic sound recognition five patients presented selective deficits in auditory spatial (P=0.138) and at the lower limit of normal performance tasks (patient MF) or in sound identification and sound sound localization (P=0.089). Her response times were motion (patients PA and HA), confirming previous find- significantly slower than those of normal subjects in the ings of such dissociations in other patients (Adriani et al.
two deficient domains (sound identification and sound 2003a, Clarke et al. 1996, 2000, 2002). Response times motion perception), but not in the preserved asemantic were significantly slower, as compared to normal subjects, sound recognition and the relatively preserved sound lo- in deficient domains. Each of the three patients had at least calization (Fig. 1). Within a given task, response times one domain in which normal performance in terms of cor- of correct and of wrong replies did not differ signifi- rect replies was associated with response times within the cantly for sound identification, asemantic sound recog- normal range, speaking against a general slowness.
nition and sound localization. For sound motion For some tasks in a given patient, normal performance perception, correct replies were significantly faster than was, however, also associated with increased response times. We argue that this profile reflects less efficient Two patients had normal performance in all four au- processing within an alternative network. The use of al- ditory cognitive functions (patient KH; P<0.05 for all ternative cortical networks for sound recognition and four functions) or in all except asemantic sound recogni- sound localization in cases of brain lesions has been dem- tion, which was at the lower limit of normal perfor- onstrated both in the ipsi- and contralesional hemispheres mance (patient GT; P=0.074 for asemantic sound (Adriani et al. 2003b) Two out of five patients in this recognition, P<0.05 for the other three functions). The study had normal performance in all domains, as assessed response times of patient KH were significantly slower with the number of correct replies. The statistically sig- than those of normal subjects in sound localization and nificant increase in their respective response times sug- sound motion perception, but not in sound identification gested, however, a relative weakness in auditory spatial and asemantic sound recognition (Fig. 1), suggesting a processing (patient KH) or in sound recognition (plus weakness in the auditory spatial processing. Within a sound motion perception; patient GT). The interpretation given task, response times of correct and of wrong of the increase in response times as a measure of reorga- replies did not differ significantly (Table II).
nization of the corresponding processing stream has ad- The response times of patient GT were significantly ditional support from the fact that patient KH sustained a slower in sound identification, asemantic sound recog- right parietal lesion, known to disrupt mainly processing nition and sound motion perception, but not in sound lo- within the auditory "where" stream (Clarke et al. 2002).
calization (Fig. 1), suggesting a main relative weakness This interpretation is in agreement with previous studies in sound recognition processing. Within a given task, by others. The crossed visuo-motor task of the response times of correct and of wrong replies did not Poffenberger paradigm, i.e., motor response to a visual differ significantly for sound localization, sound motion stimulus executed with the hand which is contralateral to perception and asemantic sound recognition. For sound the visual hemifield which has been stimulated, can be Auditory cognitive deficits and response times 169
successfully performed by callosotomized patients, but ABBREVIATIONS
their reaction times are significantly longer than those ofnormal subjects; activation studies have shown that - anterolateral auditory area in macaque monkeys callosotomized patients execute this task with different neural networks than normal subjects (Marzi et al. 1999).
- posterolateral auditory area in macaque monkeys The relatively faster response time associated with correct than wrong replies occurred in normal subjectsfor sound motion perception. In patients this phenome- REFERENCES
non was associated with deficient performance in soundidentification (patient HA) or in sound motion percep- Adriani M, Bellmann A, Meuli R, Fornari E, Frischknecht R, tion (patient PA) or with normal performance in sound Bindschaedler C, Rivier F, Thiran J-Ph, Maeder P, Clarke identification (patient GT). We argue that this addi- S (2003a) Unilateral hemispheric lesions disrupt parallel tional increase in reaction time associated with wrong processing within the contralateral intact hemisphere: Anauditory fMRI study. Neuroimage 20 (Suppl): S66-74.
replies reflects the additional processing load which is Adriani M, Maeder P, Meuli R, Bellmann Thiran A, put on alternative networks in presence of difficult Frischknecht R, Villemure JG, Mayer J, Annoni JM, items. Differences in response times for correct and Bogousslavsky J, Fornari E, Thiran J-Ph, Clarke S (2003b) wrong replies have been described previously for covert Sound recognition and localization in man: specialized face recognition in a prosopagnosic patient (Landis et cortical networks and acute effects of circumscribed le- al. 1988) and covert target detection in hemineglect (Laeng et al. 2002). In both studies, wrong negative re- Alain C, Arnott SR, Hevenor S, Graham S, Grady CL (2001) plies (i.e., the incorrect detection of an absent stimulus "What" and "Where" in the human auditory system. Proc or its incorrect identification) have been significantly Natl Acad Sci U S A 98: 12301-12306.
faster than wrong positive or correct replies. This obser- Altman J, Balanov L, Delgin V (1979) Effects of unilateral disorder of the brain hemispheric functions in man on di- vation was interpreted as correct pre-attentive process- rectional hearing. Neuropsychologia 17: 295-301.
ing followed by a faulty and early aborted attentional Anourova I, Nikouline VV, Ilmoniemi RJ, Hotta J, Aronen processing. Our data, showing a reverse pattern, cannot HJ, Carlson S (2001) Evidence for dissociation of spatial CONCLUSIONS
Bellmann Thiran A, Clarke S (2003) Preserved use of spatial cues for sound segretation in a case of spatial deafness.
Sound recognition, sound localization and sound motion perception were investigated in patients with Bellmann A, Meuli R, Clarke S (2001) Two types of auditory focal unilateral hemispheric lesions. In terms of num- Bisiach E, Cornacchia L, Sterzi R, Vallar G (1984) Disorders ber of correct replies selective deficits were observed of perceived auditory lateralization after lesions of the in 3 patients, in agreement with previously published studies. Deficient performance was associated with in- Blasi V, Young AC, Tansy AP, Petersen SE, Snyder AZ, creased response times. Two patients had normal per- Corbetta M (2002) Word retrieval learning modulates formance in all auditory cognitive functions, but right frontal cortex in patients with left frontal damage.
presented increased response times in some of them.
Anatomo-clinical correlations suggest that increased Buttet Sovilla J, Grosjean F (1997) Contribution des tests en response times reflect processing in alternative net- temps réel B l’étude de la compréhension auditive chez les works, which occurs when the specialized networks aphasiques: a propos d’un cas (in French). Bulletin Suisse de Linguistique Appliquée 66: 167-183.
Calautti C, Baron JC (2003) Functional neuroimaging studies of motor recovery after stroke in adults: a review. Stroke ACKNOWLEDGEMENT
Cao Y, Vikingstad EM, George KP, Johnson AF, Welch KM We gratefully thank collaboration of Pegah Sarkheil (1999) Cortical language activation in stroke patients re- in developing an earlier version of the adapted auditory covering from aphasia with functional MRI. Stroke 30: Clarke S, Bellmann A, de Ribaupierre F, Assal G (1996) Marzi CA, Pernai D, Tassinari G, Colleluori A, Maravita A, Non-verbal auditory recognition in normal subjects and Miniussi C, Paulesu E, Scifo P, Fazio F (1999) Pathways of brain-damaged patients: evidence for parallel processing.
interhemispheric transfer in normals and in a split brain subject. Exp Brain Res 126: 451-458.
Clarke S, Bellmann A, Meuli RA, Assal G, Steck AJ (2000) Nilipour R, Ashayeri H (1989) Alternating antagonism be- Auditory agnosia and auditory spatial deficits following tween two languages with successive recovery of a third in left hemispheric lesions: evidence for distinct processing a trilingual aphasic patient. Brain Lang 36: 23-48.
pathways. Neuropsychologia 38: 797-807.
Ohyama M, Senda M, Kitamura S, Ishii K, Mishina M, Clarke S, Bellmann Thiran A, Maeder P, Adriani M, Vernet Terashi A (1996) Role of the nondominant hemisphere and O, Regli L, Cuisenaire O, Thiran J-Ph (2002) What and undamaged area during word repetition in poststroke Where in human audition: selective deficits following fo- aphasics. A PET activation study. Stroke 27: 897-903.
cal hemispheric lesions. Exp Brain Res 147: 8-15.
Rauschecker JP, Tian B (2000) Mechanisms and streams for Fujii T, Fukatsu R, Watabe S, Ohnuma A, Teramura K, Saso processing of "what" and "where" in auditory cortex. Proc S, Kogure K (1990) Auditory sound agnosia without apha- Natl Acad Sci U S A 97: 11800-11806.
sia following a right temporal lobe lesion. Cortex 26: Rey B, Adriani M, Frichknecht R, Maeder P, Clarke S (2003) Sound recognition, sound localisation and sound motion Griffiths TD, Rees A, Witton C, Shakir RA, Henning GB, perception: patterns of recovery following unilateral focal Green GGR (1996) Evidence for a sound movement area lesions. Acta Neurobiol Exp (Wars) 63 (Suppl): 37.
in the human cerebral cortex. Nature 383: 425-427.
Rosati G, De Bastiani P, Paolino E, Prosser S, Arslan E, Griffiths TD, Rees A, Witton C, Cross PM, Shakir RA, Green Artioli M (1982) Clinical and audiological findings in a GGR (1997) Spatial and temporal auditory processing def- case of auditory agnosia. J Neurol 227: 21-27.
Rosen HJ, Petersen SE, Linenweber MR, Snyder AZ, White psychophysical study. Brain 120: 785-794.
DA, Chapman L, Dromerick AW, Fiez JA, Corbetta MD Jerger J, Loverling L, Wertz M (1972) Auditory disorder fol- (2000) Neural correlates of recovery from aphasia after lowing bilateral temporal lobe insult: report of a case. J damage to left inferior frontal cortex. Neurology 55: Karbe H, Thiel A, Weber-Luxenburger G, Herholz K, Saygin AP, Dich F, Wilson SW, Dronkers NF, Bates E (2003) Kessler J, Heiss WD (1998) Brain plasticity in poststroke Neural resources for processing language and environ- aphasia: what is the contribution of the right hemisphere? mental sounds: evidence from aphasia. Brain 126: Laeng B, Brennen T, Epeseth T (2002) Fast reponses to Spreen O, Benton AL, Fincham RW (1965) Auditory agnosia negelcted targets in visual search reflects pre-attentive pro- without aphasia. Arch Neurol 13: 84-92.
cesses: an exploratioon of reponse times in visual neglect.
Sterzi R, Piacentini S, Polimeni M, Liverani F, Bisiach E (1996) Perceptual and premotor components of unilateral Landis T, Regard M, Bliestle A, Kleihues P (1988) auditory neglect. J Int Neuropsychol Soc 2: 419-425.
Prosopagnosia and agnsocia for non-canonical views: an Warburton E, Price CJ, Swinburn K, Wise RJ (1999) Mecha- autopsied case. Brain 111: 1287-1297.
nisms of recovery from aphasia: evidence from positron Lezak MD (1995) Neuropsychological Assessment. Oxford emission tomography studies. J Neurol Neurosurg Psychi- Maeder P, Meuli R, Adriani M, Bellmann A, Fornari E, Weiller C, Isensee C, Rijntjes M, Huber W, Muller S, Bier D, Thiran J-Ph, Pittet A, Clarke S (2001) Distinct pathways Dutschka K, Woods RP, Noth J, Diener HC (1995) Recov- involved in sound recognition and localisation: a human ery from Wernicke’s aphasia: a positron emission tomographic study. Ann Neurol 37: 723-732.
Received 13 November 2003, accepted 20 March 2003

Source: http://www.msuattention.org/files/2004_ANE.pdf