Test Of Temporal Processing on CAPD

Introduction: 

  The term temporal processing refers to time related aspects of the acoustical signal. Temporal processing is critical to a wide variety of everyday listening tasks, including speech perception and perception of music (Hirish 1959).

  Several authors have discussed temporal processing as it relates to the perception of speech and specifically speech discrimination.

  Disorders in the discrimination of temporal (timing) or spectral cues of speech can lead to a breakdown in phonemic discrimination, and consequent disorders in receptive and expressive language and reading.

Temporal processing refers to time-related aspects of acoustic processing. It encompasses a wide range of auditory skills including: temporal resolution (i.e., gap detection, fusion), masking (i.e., backward and forward masking), temporal integration, and temporal ordering (i.e., temporal sequencing) (ASHA, 1996), as well as localization and pitch perception. These auditory skills are seen in a wide range of listening behaviors including: rhythm perception, periodicity pitch discrimination, segregation of auditory figure and ground, perception of a gap between two successive acoustic stimuli, and duration discrimination (Phillips, 2002). Problems arise when deficits in any one or more of the temporal auditory skills are required to decode time related aspects of speech. The manifestation of temporal processing deficits is thought to result in the typical speech comprehension problems associated with APD (Gordon-Salant and Fitzgibbons, 1993). Studies that show reduced temporal processing abilities in patients with brain lesion compared to normal populations allow us to confirm that temporal processing is indeed a central function. Downie, Jakobson, Frisk, and Ushycky (2002) administered a temporal order judgment task to children who had been born with extremely low birth weights who experienced mild, severe, or no periventricular brain injuries. The children were tested at ages 8 years 10 months to 14 years 5 months. The results indicated significantly lower (poorer) temporal order judgment scores for those children with brain injuries relative to  those who did not experience brain injuries. The known presence of a central lesion along with the significantly poor results from the temporal order judgment task suggest that these children may possibly carry a diagnosis of auditory processing disorder and its associated problems. In another lesion study, Musiek and Pinheiro (1987) studied the specificity of the Frequency Pattern Test, another temporal processing test, on its ability to rule out cerebral, brainstem, and cochlear lesions. The Frequency Pattern Test, which is similar to the tone order judgment task, demonstrated high specificity for detecting cerebral lesions when compared to normative data. This is interpreted to suggest normal central function is necessary for successful completion of this task.Strengthening the evidence of temporal processing as a central auditory nervous system function, Baran, Bothfeld and Musiek (2004) administered a battery of central auditory tests on a 46 years old woman who suffered a cerebrovascular accident (CVA) with damage involving a large portion of the primary auditory area of the left hemisphere. Threshold testing post-CVA revealed normal hearing sensitivity bilaterally; however, auditory complaints included: difficulty hearing in the presence of background noise, difficulty understanding in the presence of multiple speakers, and difficulty comprehending speech of people who speak fast. Performance on the Duration Patterns Test and the Auditory Fusion Test-Revised resulted in poor temporal processing function. The scores on the Duration Patterns Test indicated severe dysfunction in this task in both ears. The scores on the Auditory Fusion Test-Revised indicated poor temporal resolution ability bilaterally, with significantly poorer temporal resolution function in the right ear.Phoneme differentiation, syllabic rhythm, varying rates of speech, and perception of pitch as related to the varying rates of vocal fold vibration are a few examples of timing related tasks that require intact temporal processing systems in order to comprehend speech (Phillips, 2002). In other words, intact temporal processing is thought to be required at all levels of language: phonemic, morphologic, and syntactic. An example of a temporal processing dependent element of speech includes comparisons between voiced and voiceless consonants such as /b/ versus /p/. In this example, the slight difference in voice onset time may not be processed clearly, and results in incorrect comprehension (Tallal, Miller, Bedi, Byma, Wang, Nagarajan, Schreiner, Jenkins and Merzenich 1996). Another example is seen in natural pauses between syllables: “They saw the snowdrift.” versus “They saw the snow drift.” Once again, incorrect processing of the time duration between various pauses throughout speech may result in inaccurate comprehension (Chermak, 2001). At the syntactic level, an example of a temporal processing dependent sentence is: “Look out the door!” versus “Look out! The door!” (Lucker and Wood, 2000).

Methods of assessing temporal processing disorders

  Temporal processing is one of the processes of central auditory processing. Several tests to assess temporal processing are --

  Pitch pattern sequence test (PPST)

  Duration pattern test

  Gap detection test

  Temporal modulation transfer function (TMTF)

  Auditory fusion test, revised (AFT-R)

Pitch Pattern Sequence (PPS)

An Auditory Sequencing Procedure.

Three tones (some combination of high or low frequency, as illustrated by the bars) are presented either monaurally or binaurally. The patient response may be verbal (identi­fication of the frequencies as high or low) or humming the pattern. The primarily decussating (crossing) auditor)' pathways are represented by the dashed arrows. LH = left hemisphere auditor}' cortex and RH = right hemisphere auditor cortex.                     

Protocol

Pitch Pattern Sequence Test (PPS) (see related figure)

  General Information

  First reported by Pinheiro and Ptacek in 1971.

  Utilizes nonverbal stimulus.

   Clinical Validity

  Yields bilateral deficits for lesions limited to one hemisphere, thus does not provide clear iaterality information. This is related to the interaction of both hemispheres in decoding the pattern for verbal response.

  Split-brain patients and other patients with corpus collosum lesions can hum sequences but fail at verbal report.

  Sensitive to cerebral lesions, but less sensitive to brainstem lesions.

  Patients with lesions of either hemisphere or of interhemispheric pathways experience difficulty,

  Results of the PPST may provide information regarding neuro-maturation in a child with learning disability by indicating the degree of myelination of the corpus collosum (Musiek, Gollegly and Baron 1984)

  A comparison of either verbal or manual response with hummed responses is valuable in differentiating impairment in perception from impairment in processing the auditory sequences (Pinheiro and Musiek 1985)

   Test Administration

  Three tones are presented in succession, one of which is a different frequency than the other two, thus six patterns are possible.

  Two test frequencies (pitches):       high (H) = 1122 Hz

   low (L)  =  880 Hz

  Examples:      HHL     LLH      HLH      LHL      HLL      LHH

  30 to 50 pitch pattern sequences are presented.

  Five to 10 practice items precede actual scored patterns.

  A verbal subject response is usually requested initially. Then, if scores are abnormal a humming response is requested.

Scoring and Interpretation

  Percent correct score are derived for the binaural presentation.

  Normal limits vary depending on source of test material. Scores less than 75% are considered abnormal with adults suspected of auditory central nervous system lesions, although pediatric normative data are also available.

  Advantages

  Relatively resistant to cochlear loss.

  Sensitive to cortical and interhemispheric lesions based on large and varied published clinical database.

  Does not use speech as a stimulus, and thus can be used to assess individuals with limited or impaired language skills.

   Disadvantages

  Not sensitive to brainstem lesions.

  Yields bilateral deficits for lesions limited to one hemisphere, and thus cannot provide definitive laterally information.

Normative data were obtained by Bellies and others (1996) using 30 items half lists and collected from 150 listener’s ages 7 through adults. Normative values 2 SD below the mean are as follows.

   8 years –  8 years   11 months                         42%

   9 years –  9 years   11 months                         63%

  10 years – 10 years 11 months                         78%

  11 years – 11 years 11 months                         78%

  12 years to adults                                             80%

Studies in different pathologies

  Musiek et al (1980) studied auditory pattern perception in 3 split brain patients. Pitch pattern consisted of 2 different frequencies 1122 Hz and 880 Hz. Intensity patterns consisted of soft and loud sounds of 1 KHz tone.

   30 frequencies and 30 intensity patterns were presented at 40 dB HL above the SRT. Subjects were asked to respond both verbally and humming.

   Results indicated that sectioning of the corpus collosum dramatically affects the ability to verbally report both intensity and frequency patterns.

  Cranford, Stream, Rye and Slade (1982) studied detection Vs discrimination of brief duration tones in subjects with temporal lobe damage. Absolute detection threshold and difference limens were found for 1 KHz tones over a range of seven signal durations 500, 200, 100, 50, 10 and 5 msec.

  Results indicated that all subjects exhibited normal detection threshold in conjunction with substantially elevated frequency difference lines.

  Tallal (1980) observed that there were no significant differences between scores of reading disabled and controls (4) children in which stimuli were presented at slow rates. However when the stimuli were presented more rapidly, the reading impaired group made significantly more errors than the controls.

The summary of the effects of different sessions on PPST given by Bellies (1996) is given below: -

Site of lesion	Effects on temporal processing
1) Right temporal lobe	1) C/L deficit in two tone ordering or gap detection, bilateral deficit on temporal pattering tasks involving more than 2 stimuli’s
2) Left temporal lobe	2) Significant C/L and or bilateral deficits
3) Corpus collosum	3) B/L deficit on temporal patterning tasks involving more than 2 stimuli
4) Brainstem	4) Variable depending upon site of lesion and type of task.
5) Peripheral	5) Little effect on temporal patterning performance.

Duration pattern test

Protocol

Duration Pattern Test (DPT) 

   General information:

  First reported by Mustek and colleagues in 1990.

  Rationale in development was that the DPT does not require good frequency discrimination as only one frequency is used. Thus tn effects of frequency distortion that could be caused by cochlear hearing loss are minimized.

Clinical Validity

  Yields bilateral deficits for lesions limited to one hemisphere, and thus cannot provide definitive laterality information. This is related to the interaction of both hemispheres in decoding the pattern for verbal report.

  More sensitive to cerebral lesions than the pitch pattern test.

  Previous studies showed that the duration cue is more resistant to cochlear hearing loss than a frequency or intensity cue.

Test Administration

  Three tones are presented in succession, one of which is different in duration than the other two.

  Long (L) = 500 msec; short (S) = 250 msec

  Six patterns are possible:     LLS     SSL     LSL     SLS     LSS     SLL

  30 to 50 patterns are presented, with 5 to 10 practice items preceding the actual testing.

  Subject response can be verbal, manual (pointing), and/or humming.

Scoring

  Percent correct score are derived for binaural presentation.

  Scores less than 70% are considered abnormal by some investigators, although normative data are limited especially for children.

Advantages

  Relatively resistant to cochlear loss.

  Flexibility (any frequency can be used).

  Sensitive to cortical and interhemispheric lesions.

  Does not use speech as a stimulus thus can by used to assess indi­viduals with limited or impaired language skills.

Disadvantages

  Not sensitive to brainstem auditory dysfunction.

  Yields bilateral deficits for dysfunction that is limited to one hemisphere, and thus cannot provide definitive laterality information.

  Lack of published investigations in literature describing clinical applications.

Studies in different Pathologies

  Theoretically and some experiments support that concept that both the hemisphere must interact to decode the pattern and orally report it (Musiek et al 1980, 84). Given that subject with disorders often show B/L deficits and normal show similar performance in both the ears, many clinicians opt for doing the test in a sound field. This shortens the test but may also cause a few patients, with disorders that show only unilateral deficit on pattern perception to be missed.

  Therefore, with patterns often revealing B/L deficits, it is often difficult to determine with hemisphere are involved when verbal report is required.

   Humming the patterns appears to require only the right hemisphere, hence if a subject cannot orally report the patterns, but hum them could mean that the R hemisphere is intact and the problem is either in the L hemisphere or corpus colosum (Musiek, Pinherio and Wilson 1980, Musiek Gollegly & Baran 1984).

  Severe B/L deficits are seen in cases of auditory cortex involvement. According to a study done by Musiek, Baran and Pinherio (1990), three groups of subjects were tested on duration pattern recognition task.

   The groups included fifty Normal hearing subjects; the second group included twenty four subjects who had been topologically diagnosed as cochlear hearing losses in one or both ears and negative neurologic histories.

  The third group of subjects included twenty one individuals with neurologically, radiologically and/or surgically confirmed Central Auditory Nervous system lesions involving but not limited to the auditory areas of the cerebrum.

   Results indicated no significant difference in pattern recognition between the normal subjects and subjects with cochlear hearing loss.

   However, the subjects with cerebral lesions performed significantly more poorly than either the normal subjects or those with cochlear hearing loss. In comparing pattern recognition performance for the ears ipsilateral and contra lateral to the hemispheres, with lesion, no differences were noted. Rather, when a central lesion was present, both ears generally yielded abnormal scores.

  Performance on temporal patterning tests, such as DPT, typically are non-lateralizing i.e. central lesions tend to result in bilateral deficits, regardless of the site of lesion. Therefore, laterality information cannot be obtained by these tests.

  Thompson and Abel (1992) reported that the listener required a greater response time than did the other two groups. In all cases, listeners with lesions of the left temporal lobe tended to demonstrated greater performance deficits.

  Interhemispheric corpus collosum lesions resulted in bilateral deficits on tests of temporal patterning i.e. on DPT, when the listener was required to respond verbally. Bilateral deficits on DPT have been reported (verbal) for patients with deep brain lesions, presumably affecting the Transcortical Auditory Pathway (Musiek, Baran and Pinheiro 1990).

  Split brain patients typically demonstrated bilateral (for verbal report) on pattern perception tests, even though the assessment is a monaural procedure.

  The final portion of the CANS to attain neuro-maturation in the corpus collasum, which is responsible for interaction between cerebral hemispheres.

  Theoretically, it is found that a neuro-maturational delay will result in findings on tests of Central Auditory Function, which are similar to corpus collosum involvement.

   Poor performance is seen on tests of temporal patterning, i.e., DPT, although the performance on these tests may improve when the listener is asked to hum the responses, thus removing the linguistic labeling component (Musiek, Kibbe and Baran 1984).

  Musiek and Lamb (1994) used the terms cortical and hemispheric to differentiate between the two types of cerebral lesion. Cortical refers to the gray matter of the brain alone where as hemispheric refers to lesions that affect both the white and gray matter.

   With similar performance in both the ears, many audiologists opt for doing the test in a sound field. This shortens the test but this may also cause a few patients, with disorders that show only unilateral deficit on pattern perception, to be missed.

  Therefore, with patients often revealing bilateral deficits, it is often difficult to determine which hemisphere is involved when verbal report is required.

Gap Detection Test

  One psychophysical test which has recently been introduced to measure temporal resolution is gap detection test.

   Gap detection is a reasonably well-established method, which measures the ability of the listener to detect brief temporal gap reporting two successive stimuli. Gap detection is probably the most commonly used measure of temporal resolution, i.e. ability to follow rapid changes over time.

  The basic strategy of the gap detection experiment is actually quite straightforward. Suppose we have a continuous burst of noise lasting 500 msec. we could “chop out” a start segment in the center of noise lasting, say 10 msec. we now have a leading noise burst lasting 295msec followed by a 10 msec silent period, followed by a trailing 295 msec noise burst.

   Hence we have a gap lasting 10 msec surrounded in time by leading and trailing raise bursts.

  The duration of the gap is varied according to some psycho physical method in order to find the shortest detectable gap between the two noise bursts which is called the gap detection threshold (GDT).

   Thus the GDT reflects the shortest time interval we can resolve and it is taken as a measure of temporal resolution.

  The duration of the gap is varied according to some psycho physical method in order to find the shortest detectable gap between the two noise bursts which is called the gap detection threshold (GDT).

   Thus the GDT reflects the shortest time interval we can resolve and it is taken as a measure of temporal resolution.

   Divenyi and Hirsh (1974) pointed out that auditory discrimination performance is substantially affected by at least four factors:

1.      No. of stimuli in the sequence of items to be discriminated.

2.      How the sequences are presented (separate presentations or continuously)

3.      Its kind of task the subject must perform

4.      The amount of training the subject has had.

  Differential sensitivity for duration has also been investigated and the general finding is that the difference limen for duration (▲T) becomes smaller as the overall duration decreases (Small and Campbell 1952; Abel 1972, Sinnatt et al 1987, Doaley and Moare 1988).

   Its results were essentially independent of bandwidth and intensity

Advantages

Provides description resolution based on a single threshold, where as other methods require multiple threshold estimates.

Easy to measure is a naïve listener, including infants.

Gap detection threshold obtained from noise listeners are close to those obtained from well trained listeners (Halpen, Spetner and Gillenwater, 1992).

Factors affecting gap detection

1.   Type of stimuli

         Band Pass noise

         Wide band noise

         Stimuli with sinusoidal markers

2.   Noise burst duration

3.   Location and uncertainty of gap

4.   Gap on set and offset

  Subject related factors 

         Age

         Hearing loss

  Language disabilities
Gap Detection in Noise Test (GIN)
The GIN test is another experimental temporal resolution test developed by Frank E. Musiek (Shinn, Jirsa, Baran and Musiek, 2004). Gap detection thresholds are the focus of the GIN, however, in this test protocol the presentation stimuli surrounding the IPIs are no longer tone or click pulses. Instead, a constant white noise presented for a duration of 6 seconds is utilized. Interspersed within the six-second white noise are random gaps, ranging in duration from 2 msec to 20 msec with each gap trial occurring six times within each test list. The GIN test consists of a practice section, and 4 forms (variants of the randomized stimuli). Currently, only adult normative data exists. The mean gap detection duration for adults is 4.9 msec, with a standard deviation of 1 msec (Shinn, Jirsa, Baran and Musiek, 2004). The GIN is administered in both monaural conditions at a presentation level of 55 dB SL, reference to PTA. The gap in noise threshold is determined by the shortest gap that is perceived to be present four times out of six presentations. A notable difference between the GIN and the 3 temporal resolution tests previously described is the response tasks required from subjects. In the administration of the AFTR, RGDT, and the BFT, subjects are required to respond verbally by indicating whether one or two presentations are heard. The GIN on the other hand, requires subjects to respond by clicking a button when a gap in the six second continuous white noise is detected. By not requiring a response involving speech or language production, the absence of the language-processing component in the GIN, as compared to the 3 other tests, suggests the GIN is less cognitively demanding and may, therefore results in reduced response times. It has not been shown that either response task is preferable for temporal resolution testing; however, it is assumed that each tests provides sufficient time between stimulus presentations for subjects to process the auditory information and respond accordingly and as such, should not affect this study’s ability to compare tests. The validity of the GIN as a clinical measure of temporal resolution was examined by comparing the performance of eighteen patients with confirmed neurological lesions of the central auditory nervous system (CANS) with fifty subjects with normal hearing (Shinn et al., 2004). The GIN scores of the patients with CANS lesions were statistically larger than the scores of the normal subjects. The GIN demonstrated a sensitivity of approximately seventy to eighty percent to CANS lesions. The GIN was also shown to present strong equivalent forms reliability (Shinn, personal communication, 2004).  

Auditory Fusion Test-Revised (AFTR)

The AFTR, by Robert L. Mc Croskey and Robert W. Keith (1996), is essentially a digitized version of the WAFT. The AFTR measures the shortest separation (in milliseconds) between two auditory stimuli that results in a subject to perceive a single stimulus rather than two separate stimuli. This duration is identified as the Auditory Fusion Threshold (AFThreshold) and is measured in milliseconds (msec). In the AFTR, the tone pairs are presented in ascending and descending runs relative to inter-pulse interval (IPI) durations, both in monaural or binaural presentations. The test consists of 3 subtests. Subtest 1 is a screening test with 500 Hz tone pairs ascending from a 0 to 300 msec IPI. Standard subtest 2 contains five frequencies: 500 Hz, 1000 Hz, 4000 Hz, 250 Hz, and 2000 Hz. Each frequency is presented with 18 corresponding frequency pairs with successively larger and smaller IPI. The IPI of each tone pair ascends from 0 msec to 40 msec and then descends back down to 0 msec. Expanded subtest 3 contains 18 pairs of tones at each of the three frequencies: 1000 Hz, 4000 Hz, and 250 Hz. The IPI of each tone pair ascends from 40msec to 300msec and then descends back down to 30 msec. The expanded subtest 3 is only utilized when two consecutive two-tone pulses are not reported until a 60 msec interpulse interval or greater during subtest 1. If the listener reports two consecutive two-tone pulses under a 60 msec IPI during subtest 1, then the standard subtest 2 is administered. Each subtest is administered in both the monaural condition (left and right) and the binaural condition at a presentation level of 50 dB SL, reference pure tone average (PTA) (Mc Croskey and Keith, 1996). The auditory fusion threshold is determined by a two-step calculation. For each frequency run, an auditory fusion threshold is determined by averaging the last ascending IPI that is perceived as a single fused stimulus before consecutive IPI are perceived as two distinct stimuli, with the first of two consecutive descending IPIs that are perceived as a single fused stimulus. Next, all frequency specific auditory fusion thresholds are averaged to determine the final auditory fusion threshold.No reliability studies have been done on the AFTR; however, limited studies of validity have been published. (Reliability refers to the extent that a test’s score remains stable over a period of time when tested and re-tested on the same group of subjects. Validity refers to the extent a test measures what it was designed to measure. Both reliability and validity studies are necessary in order for one to assume with confidence that a given measure does in fact measure what it is testing.) Mc Croskey and Keith utilized predictive validity to determine whether the AFTR does in fact measure the AFThreshold. Predictive validity is the degree to which the scores on two tests taken at different times are correlated. Predictive validity is used to examine how well scores on one test predict accurately findings on already established tests. For example, if a new temperature thermometer was developed, in order for it to have predictive validity, it must be able to predict accurately the temperature reading of an already proven temperature thermometer currently in use when measuring an individual’s temperature. The predictive validity of the AFTR was demonstrated by its ability to accurately predict individuals with poor temporal processing abilities by comparing itself to its predecessor, the Wichita Auditory Fusion Test (WAFT) of 1975 (Mc Croskey and Keith, 1996). McCroskey and Kidder (1980) studied the validity of the WAFT by testing 135 children divided into groups of normal children and children with reading and learning disorders. They found that children with reading and language disorders present with temporal resolution deficits, which they believe confirmed the hypothesis that the auditory fusion threshold is an effective tool in identifying individuals with temporal resolution problems. Isaac, Horn, Keith, and McGrath (1982) confirmed the results of McCroskey and Kidder’s study. Issacs et al. (1982) reported that the children in the group with language and learning disabilities had significantly larger auditory fusion thresholds than the control group. However, no studies of validity have yet shown that the auditory fusion threshold paradigm identifies temporal resolution deficits in individuals diagnosed with APD.

Binaural Fusion Test (BFT)
The BFT is an experimental fusion test developed by Frank E. Musiek (2002). It is designed to identify temporal resolution deficits and binaural interaction problems by utilizing a series of noise burst pairs with randomly assigned IPIs. The BFT requires subjects to attend to stimuli presented to both ears, and respond by counting whether one or two noise bursts were heard. The noise burst pairs are presented dichotically, in contrast to the RGDT and the binaural portion of the AFTR where the stimuli are presented diotically. The BFT differs from the GIN in that the BFT is a binaural, dichotic task that tests for a fusion threshold, whereas the GIN is a monaural gap in noise detection test. In the BFT, each noise burst from each pair is presented asynchronously between each ear. For example, if the initial noise burst is presented in the left ear then the subsequent noise burst, after the IPI, would be presented in the right ear. This modification results in a more complex task and may enable the experimenter to determine not only if there is a temporal resolution dysfunction, but also opens the possibility of exploring a binaural interaction deficit. Binaural interaction is an auditory skill that enables individuals to take different stimuli from both ears and combine them to produce a meaningful auditory percept (Plakke, Orchik and Beasley, 1981). In real world conditions, this ability to merge disparate information from both ears is seen when acoustic information to each ear differs in timing and intensity due to the head shadow effect and spatial separation of the ears (Perrott and Nelson, 1969). In order to separate out possible deficits in temporal resolution and binaural interaction, further testing utilizing binaural interaction specific tests would be required. Tests of binaural interaction include masking level difference and tests of localization/lateralization. The BFT consists of 3 forms. Each form has 54 trials and randomizes its IPIs and noise burst presentations differently. One of the three available forms can be used as a training/practice section to help in training the subject to the task. The initial presentation (left or right) of each dichotic pair of noise bursts are also randomized to reduce any predictability effects. The IPIs utilized range from 0 msec to 100 msec and are presented at 55 dB HL. The binaural fusion threshold is determined by the shortest IPI that is perceived four out of six times. No data is available regarding the validity and reliability of this experimental test.