INTENSIVE TREATMENT OF DYSARTHRIA IN AN ADULT WITH A TRAUMATIC BRAIN INJURY

Objective: This study investigated the impact of an intensive articulation treatment on acoustic and perceptual measures of speech in an individual with spastic dysarthria acquired from a traumatic brain injury (TBI). Method: A single-subject A-B-A-A experimental design was used to measure the effects of an intensive articulation treatment that incorporated principles of motor learning to evaluate the impact on speech and communication. The primary dependent variables were single word intelligibility and vowel space area. Additional dependent variables included vocal sound pressure level (dB SPL) during a variety of speech tasks, acoustic measures of voice, and listener perceptual ratings of voice quality and speech. Results: Multiple comparisons with t-tests were used to determine statistically significant changes in primary and secondary dependent variables. Statistically significant (p<0.05) changes were present immediately post-treatment with single word intelligibility (p=0.00), vowel prolongation duration (p=0.00), and lip pressure exerted (0.04) and six months following treatment with vowel duration prolongation (p=0.01) and Noise-to-Harmonics Ratio (p=0.02). There were no statistically significant (p<0.05) changes with listener preference studies, vowel space area, and vocal dB SPL across vowel prolongation and speech tasks immediately post-treatment and six months following treatment. Conclusions: These data demonstrate that this individual with spastic dysarthria secondary to a traumatic brain injury responded positively to an intensive articulation treatment on selected variables, particularly on tasks practiced directly in treatment. Generalization of a treatment effect outside of treatment was not demonstrated. Further research is needed to determine whether the lack of generalization was due to the treatment or specific characteristics of the individuals who are treated.

Dysarthria tends to be more persistent and stable following the acute phase of TBI and may result in decreased social participation, reduced quality of life, and depression after discharge from rehabilitative services (Brady et al, 2011;McAuliffe et al, 2010).
Therefore, treatment studies to ameliorate dysarthria secondary to TBI are needed to identify the potential of specific individuals to benefit from treatment (Yorkston, 1996).
Dysarthria is characterized by abnormalities in strength, speed, range, timing, and/or accuracy of articulatory movements caused by damage to the nervous system that can result in reduced communicative intelligibility, comprehensibility, and naturalness (Mackenzie & Lowit, 2007). Intelligibility refers to how accurately a speaker's acoustic signal is received by a listener (Hustad, 2008). Comprehensibility refers to how accurately a speaker's acoustic signal is received when paired with speaker, listener, and environmental support (Mackenzie & Lowit, 2007). Reduced intelligibility and comprehensibility can result in disrupted and unsuccessful 3 communicative interactions, which diminishes an individual's quality of life secondary to limited social and vocational participation, coupled with acquired communicative avoidance strategies (Brady, Clark, Dickson, Paton, & Barbour, 2011;Walshe, Miller, Leahy, & Murray, 2008).
The literature on dysarthria treatment provides evidence of behavioral, medical, and prosthetic approaches used to improve functional communication (Duffy, p. 405, 2012;Mackenzie & Lowit, 2007). Behavioral management is the most frequently published approach for increasing intelligibility in individuals with dysarthria (Duffy, p. 415, 2012, Mackenzie & Lowit, 2007Mahler & Jones, 2012;Mahler, Ramig, & Fox, 2009). Increasing intelligibility is a critical component of dysarthria management because of its relationship with improved functional communication, cognition, and quality of life (Mackenzie & Lowit, 2007). However, there is little evidence available in the literature describing specific treatment approaches for improving intelligibility in individuals with dysarthria, due to the heterogeneity of the disorder (Sellars, Hughes, & Langhorne, 2005;Yorkston, 1996).
Furthermore, few reports analyze treatment efficacy of dysarthria in individuals, particularly for individuals with a TBI, because it is often accompanied by other complex cognitive-linguistic disorders (Centers for Disease Control and Prevention, 2013). There is growing evidence regarding the relationship between dysarthria management and the motor-learning literature (Maas, Robin, Austermann Hula, Freedman, Wulf, Ballard, Schmidt, 2008). Behavioral speech treatments based on principles of motor learning have potential to improve the treatment of dysarthria in individuals with TBI (Mahler & Jones, 2012;Mahler & Ramig, 2012;Wenke, 4 Theodoros, Cornwell, 2008). Therefore, the purpose of the current investigation is to determine the impact of an intensive articulation treatment that incorporates principles of motor learning on the intelligibility of an individual with spastic dysarthria secondary to a TBI. It is hypothesized that an individual with spastic dysarthria secondary to TBI will improve intelligibility for functional communication following an intensive articulation treatment.
1. It is hypothesized that this individual's single word intelligibility will improve secondary to an intensive articulation treatment.
2. It is hypothesized that this individual's vowel space area will increase secondary to an intensive articulation treatment.

Study overview
A single subject A-B-A-A experimental design was selected for this Phase I research study. Single subject designs are critical in determining treatment effectiveness with one individual, as well as, providing pilot data to justify group treatment efficacy studies (Robey, 2004). A single subject A-B-A-A experimental design was important for making an initial determination of response to an intensive articulation treatment for an individual with dysarthria secondary to TBI. The primary dependent variables were listener intelligibility scores based on 50 single words (Bunton, Leddy, & Miller, 2007) and vowel space area analysis calculated through first and second formant frequencies. Additional dependent variables included vocal sound pressure level measured in dB SPL during reading of sentences, picture and task descriptions, and maximal vowel prolongation; acoustic measures of phonatory stability during maximal vowel prolongation; and listener perceptual ratings of speech comparing pre-and post-treatment samples and pre-and follow-up treatment samples.
The participant received repeated measures of the dependent variables under control conditions during the A phases of the study. Four individual one-hour treatment evaluations were completed the week immediately prior to treatment (A1), after treatment (A2), and six months post-treatment (A3) to allow for trend analysis and visual inspection of data (Beeson & Robey, 2006;Parsonson & Baer, 1992).
Repeated evaluation tasks under controlled conditions included: sustained vowel prolongation, single word intelligibility (Bunton et al., 2007), sentence reading (five 6 repetitions of "The boot on top is packed to keep.") and paragraph reading (The Farm Passage, Crystal & House, 1982), picture description (picnic scene from the Western Aphasia Battery (WAB), Kertesz, 1982), task description (e.g. Describe your favorite sport.), and lip and tongue pressure measures (three repetitions of each that varied no more than 10%). The Bunton, Leddy, & Miller single-word intelligibility test (2007) was administered during Pre4, Post1, and Follow-up 1 evaluation sessions to assess single word intelligibility. The Bunton et al. (2007) (1989), which was designed to examine the acousticphonetic contrasts that contribute to speech intelligibility in individuals with dysarthria (Bunton et al., 2007;Kent, Weismer, & Kent, 1989;Mackenzie & Lowit, 2007). This intelligibility test (Bunton et al., 2007)

Equipment and recording procedures
Each pre-, post-, and follow-up evaluation session occurred in an IAC soundtreated booth at the University of Rhode Island Speech and Hearing Center. A headmounted microphone, model Isomax B3, was adjusted to a mouth-to-microphone distance of 8 cm. A sound level meter (SLM -Radio Shack 33-2055) was 40 cm away from TBI02's lips and level with his mouth to collect vocal intensity data during speech tasks in real time (Matos, 2005). Mouth-to-microphone and SLM distances remained constant across the three weeks of evaluations for reliable data collection.
The head-mounted microphone and SLM signals were digitized directly to a computer and simultaneously recorded onto a flash recorder, Marantz PMD670. A pre-amplifier (Universal Audio 4110) was used to assure quality signal acquisition with the microphone. Speech was sampled at 44.1 kHz using Adobe Audition 2.1 software and standard speech and voice analysis procedures, which were previously 9 discussed in the literature (Ramig, Countryman, Thompson, & Horii, 1995). Each evaluation session was recorded by a HandyCam DCR-DVD92 digital video camera.

Treatment
An intensive articulation treatment was administered during four one-hour treatment sessions a week for four weeks for a total of 16 individual treatment sessions. The articulation treatment implemented traditional articulation tasks, including minimal pairs and exaggerated articulation. The actual tasks completed each treatment session are commonly used by speech-language pathologists, but not supported in the literature for treatment of spastic dysarthria secondary to a TBI.
Administration of the treatment was novel because it incorporated principles of motor learning to promote neural restructuring for increased intelligibility for functional communication. The articulation treatment was driven by principles of motor learning for clinical rehabilitation of dysarthria, which has been previously discussed in the literature (Kleim & Jones, 2008;Mahler et al., 2009;Ramig, Sapir, Countryman, Pawlas, O'Brien, Hoehn, & Thompson, 2001a;Ludlow, Hoit, Kent, Ramig, Strand, Yorkston, Sapienza, 2008). The literature discusses neural plasticity, which refers to the ability of the central nervous system to change and/or adapt to environmental influences for both learning in normal brains and relearning in damaged brains (Ludlow et al., 2008;Kleim & Jones, 2008 activation of attentional brain networks for neutral function underlying clear speech production (Ludlow et al., 2008;Maas et al., 2008).
The motor learning principle of complexity was established through the exercise-dependent articulation tasks completed each treatment session. Speech is a complex motor task that can be divided into component parts to practice (Maas et al., 2008). Therefore, the first half of each treatment session utilized tasks that were overlearned, which reduced cognitive-linguistic demands, but were real speech tasks since the goal of treatment was increased intelligibility. This approach was particularly useful for TBI02 because concentration on a single aspect of speech TBI02 sustained "ah" at his habitual pitch for speech to improve the coordination of respiration and phonation, strengthen vocal fold adduction, and increase vocal loudness. Previous intelligibility studies indicated that increased loudness can improve intelligibility and vocal quality (Lam, Tjaden, & Wilding, 2007;Dromey, 2000;Goberman & Elmer, 2005;Pichney, Durlack, & Braida, 1986;Ramig et al., 2001a;Trail, Fox, Ramig, Sapir, Howard, & Lai, 2005). Therefore, duration of sustained phonation and loudness measured in dB SPL were collected. TBI02 then counted to 15, an automatic speech task with low cognitive load, to incorporate high effort training of articulation and vocal loudness. Loudness measured in dB SPL was collected.
TBI02's most salient speech sound errors, final /t/ and /d/ and final /g/ and /k/, were targeted through minimal pair tasks (i.e., pairs of words that differ by only one sound; e.g., "sat" and "sad"). Targeted Table 1.

Listener Intelligibility and Perception Tests
Ten undergraduate and graduate communicative disorder students with normal hearing and no history of neurological disorder served as listeners for intelligibility testing. Listeners signed a consent form following review of the purpose of the study and confidentiality. Listeners were unfamiliar with TBI02 to represent a typical communication situation since the literature has shown that familiarization with a speaker increases intelligibility (Garcia & Cannito, 1996). Listeners were blind to the time of a recording of 50 phonetically balanced words (Bunton et al., 2007) that were collected during Pre 4, Post 1, and Follow-up 1 evaluation sessions. Listeners circled the word that they heard through a multiple-choice format of the target word and three foil words, which were chosen for the interpretation of vowel and consonant errors perceived by listeners. A blank column was provided for listeners to write in a word they heard that was not presented in the list. The total number of words accurately identified by the blind listeners was used to calculate percent single word intelligibility score (Kent et al., 1989).
Blind listeners then listened to pairs of 25 identical sentences (e.g., "The boot on top was packed to keep.") to control for speech content, limit listener bias, and maintain reliability across listeners. The sentence, "The boot on top is packed to keep," which was read five times and collected during each pre-, post-, and follow-up evaluation session, was randomly presented to the listeners for a total of 25 paired sentence comparisons at each evaluation. Sentence pairs were randomized based on presentation (e.g., pre-, post-, follow-up) and sentence token number (e.g., 1-25).
Listeners were presented two speech samples at a time and asked to rate the second 15 sample (B) relative to the first sample presented (A) based on naturalness (e.g., vocal loudness, vocal quality, pitch variability, and speech clarity Listener preference percentages were calculated by dividing the distance between zero and the rating by half of the total length of the line scale.

Vowel Space Area
Previous literature has demonstrated that acoustic measures are sensitive to articulatory movements during vowel and consonant production in speakers with spastic and mixed dysarthria (Kent et al., 1989;Kent, Weismer, Kent, Vorperian, & Duffy, 1999;Roy, Leeper, Blomgren, Cameron, 2001). Reduced vowel space area calculated from F1 and F2 of corner vowels has been associated with speakers with dysarthria. The literature shows that centralization of the first and second formant frequencies (F1 and F2) and reduced articulatory movements of vowels account for decreased intelligibility of dysarthric speech (Roy et al., 2001;Mahler & Ramig, 2012). An increase in vowel space area has been correlated with improved intelligibility scores during perceptual studies (Liu, Tsao, & Kuhl, 2005). Therefore, acoustic analysis of vowel space area was performed to determine the impact of the intensive articulation treatment on speech intelligibility, as well as, overcome the limitations of subjective, listener intelligibility studies (Collins, 1984).
Vowel area was calculated from vowel triangles obtained from the sentence, "The boot on top is packed to keep." The sentence was read five times during each pre-, post-, and follow-up evaluation session, resulting in a total of 20 tokens at each evaluation. F1 and F2 values were determined through wideband spectrographic displays and linear predictive coding spectra using Time-Frequency Analysis Software (TF32), a Windows-based version of CSpeech software (Milenkovic, 2001, Madison, WI). F1 and F2 values were obtained from the corner vowels /u/, /a/, and /i/ measured at the temporal midpoint of each vowel production to avoid interference of coarticulation.

Vocal Sound Pressure Level (dB SPL)
Vocal sound pressure level measured in dB SPL was collected during sentence reading (e.g. "The boot on top is packed to keep."), reading of the Farm Passage (Crystal & House, 1982), picture description of the picnic scene from the Western Aphasia Battery-Revised (Kertesz, 2006), and task description, which varied for each evaluation session. Vocal dB SPL was chosen as a dependent variable to determine the impact of vocal loudness on increased intelligibility and comprehensibility across various speech tasks.

Acoustic Measures
Acoustic and highest outliers (i.e., values one standard deviation below and above the mean) for listener ratings for pre-, post-, and follow-up single word intelligibility were omitted to decrease error variance and increase normality of data. In addition, TBI02 did not receive additional speech treatment during participation in the research study.

Single Word Intelligibility
Single word percent accuracy increased from 24% pre-treatment to 73% posttreatment, revealing a 49% increase following treatment. The pre-post t-test was 0.00, which was statistically significant (p<0.05), with a large effect size at 0.97. The follow-up single word intelligibility decreased to 21% follow-up treatment, revealing a 3% decrease. The pre-follow-up t-test was 0.17, which was not statistically significant (p<0.05), with a small effect size at 0.29. Quantitative changes of single word percent intelligibility from pre-, post-, to follow-up treatment are displayed in Table 2.

Listener Perception Tasks
The

Vowel Space Area
Pre-, post-, and follow-up vowel triangles were obtained by analyzing F1 and F2 values of vowels /u, a, i/ to calculate vowel space area. Vowel space area for pretreatment was 193,802 Hz 2 and for post-treatment was 214,463 Hz 2 , indicating a 20,661 Hz 2 change. Vowel space area for follow-up treatment was 253,886 Hz 2 , indicating a 60,084 Hz 2 change compared to pre-treatment. The pre-post t-test was 0.52, which was not statistically significant (p<0.05), with a small to medium effect size of 0.38. The pre-follow-up t-test was 0.05, which was not statistically significant (p<0.05), with a medium to large effect size of 0.74. Table 5 illustrates quantitative changes for pre-, post-, and follow-up treatment vowel space area. Figures 1 and 2 present a visual depiction of pre-post and pre-follow-up vowel space areas.  Table 6 illustrates quantitative changes in F1 and F2 across pre-, post-, and follow-up evaluation session for /u/, /a/, and /i/.

Vocal dB SPL
Visual inspection of mean vocal dB SPL data for sustained vowel phonation and speech tasks (e.g. reading of sentences and paragraphs, picture and task description) indicated stability across pre-, post-, and follow-up evaluation sessions.
Appendices C and D present visual depictions of pre-post and pre-follow-up mean vocal dB SPL data for sustained vowel phonation and speech tasks, respectively. The slopes of dB SPL data for pre-, post-, and follow-up treatment fluctuated for each evaluation session with overlapping values for each speech task. Pre-post and prefollow-up t-tests were not statistically significant for any speech tasks (p< 0.05) and the effect size was small for sustained vowel phonation, reading of paragraphs, and task description and medium for reading of sentences and picture description. The pre-follow-up effect size was small to medium for sustained vowel phonation, reading of paragraphs, picture description, and task description and medium for reading of sentences. A summary of quantitative changes in vocal dB SPL from pre-, post-, and follow-up evaluations is displayed in Table 7.

Phonatory Stability
The pre-post t-tests for relative average perturbation (RAP), pitch perturbation quotient (PPQ), and noise-to-harmonics ratio (NHR) revealed no statistically significant changes (p<0.05) in phonatory stability, with a small effect size for RAP and PPQ and a medium to large effect size for NHR. The pre-follow-up t-tests for RAP and PPQ were not statistically significant (p<0.05), with small effect sizes.
However, the pre-follow-up t-test for NHR was 0.02 and was statistically significant (p<0.05), with a large effect size at 0.89. The RAP and PPQ pre-, post-, and follow-up treatment means were within the normative range for the participant's gender and age.
The NHR pre-treatment mean was slightly above the normative range, but fell within the normative range for both post-treatment and six months following treatment.
Quantitative changes in MDVP values during vowel prolongation are displayed in Table 8.

Lip and Lingual Pressure Exerted
The pre-post t-test for lip pressure exerted was 0.04 and was statistically significant (p<0.05), with a large effect size at 0.85. However, the pre-follow-up t-test was not statistically significant (p<0.05), with a medium effect size at 0.54. The pre-27 post and pre-follow-up t-tests for lingual pressure exerted were not statistically significant (p<0.05), with a medium effect sizes of 0.47 and 0.48, respectively.
Quantitative changes in lip and lingual pressure exerted kPa values are displayed in Table 9. Figures 5 and 6 illustrate mean lip and lingual pressure exerted pre-post treatment and pre-follow-up treatment, respectively.  revealed that individuals with nonprogressive dysarthria acquired from TBI can improve speech intelligibility following intensive treatment. Clinically significant improvements were evident with tasks that were directly trained within each treatment session with little generalization to stimuli not directly trained. The first hypothesis that this individual would improve single word intelligibility was supported by the data immediately following treatment, but not at the six-month evaluation. TBI02 had hip replacement surgery performed three weeks following post-treatment evaluations.
Lack of maintenance of statistically significant improvements in single word intelligibility may have been related to TBI02's shift in focus from clear speech for functional communication to physical mobility. The second hypothesis that there would be an increase in vowel space area was supported immediately following treatment and at the six month evaluation, with the most significant increase occurring at follow-up. This increase in vowel space area was related to improvements in F1 across /u/, /a/, and /i/ towards normative values, which was indicative of increased lingual height. However, this increase in lingual movement did not have an impact on single word intelligibility six months following treatment.

Listener Intelligibility and Perception Tasks
A four-week intensive articulation program appeared to be a feasible intervention with a large treatment effect size for single word intelligibility for the participant in this study. However, improvements in single word intelligibility were not maintained six months following treatment, with percent accuracy declining approximately to baseline. Lack of maintenance of increased intelligibility at the word level may be related to the participant's complex cognitive-linguistic deficits acquired from his TBI, his hip replacement surgery, and lack of consistent completion of homework exercises. He continued to require external cues on untrained single words and conversation outside of the treatment room. Listener perceptual studies using sentence pairs demonstrated little to no carryover of improvement in speech intelligibility at the sentence level across pre-, post-, and follow-up evaluations.
Listeners preferred treated sentence pairs 51.7% of the time immediately posttreatment and 48.2% of the time six months following treatment, which suggested a lack of generalization to sentences that were not directly targeted during treatment.

31
This was expected due to the increased cognitive-linguistic demands associated with a more complex speech task, as well as, the participant's habitual use of telegraphic speech.
Improvements in single word intelligibility had a functional impact on TBI02's daily communication and social participation due to his chronic use of single words and short phrases during conversation. His caregivers, family members, and graduate speech-language pathology clinicians reported increased comprehensibility and reduced communicative breakdowns during conversation immediately post-treatment.
In addition, his caregivers continued to report increased intelligibility and comprehensibility during functional communication six months following treatment.  (Walshe et al., 2008). Therefore, the rating continuum used in this research study may not have been sensitive enough to capture changes in speech perception.

Vowel Space Area
Acoustic analysis of vowel space area was completed to determine acousticarticulatory changes associated with increased single word intelligibility. The magnitude of treatment effect on vowel space area was small to medium immediately post-treatment and medium to large six months following treatment. The change in vowel space area was not statistically significant immediately post or six months following treatment. However, the increase in vowel space area illustrated changes in F1 and F2 values towards the normative range, which was indicative of increased lingual height and advancement (Hillenbrand et al., 1994). An inverse relationship occurs with F1 values and tongue height (e.g., high-low), while a direct relationship occurs with F2 values and tongue advancement (e.g., front-back). For example, F1 is lower in frequency when tongue position is higher in the mouth and F2 is higher in frequency when the tongue is more anterior in the mouth (Liu et al., 2005). A larger vowel space area following treatment was primarily dependent on improvements in F1 values across /u/, /a/, and /i/, which demonstrated critical changes in tongue height.
It may be that the external cue to "speak clearly" prompted TBI02 to use greater articulatory effort, which resulted in improvements in vowel space area (Kim, Hasegawa-Johnson, & Perlman, 2010

Vocal dB SPL
The intensive articulation treatment had no statistically significant effect on vocal dB SPL for all speech tasks immediately post and six months following treatment. This finding was expected because increased loudness was not directly trained during treatment since TBI02 presented with loudness levels within normal limits pre-treatment, which was consistent with his diagnosis of spastic dysarthria.
His spastic vocal quality was a significant contributor to his overall reduced speech intelligibility related to dysarthria. Changes in vocal dB SPL were not expected to occur due to his average normal loudness at baseline and lack of training, which further emphasized the motor learning principles of salience and specificity. These results indicated that improved speech intelligibility was not correlated with increased vocal loudness.
A significant treatment effect was demonstrated for vowel prolongation immediately post-treatment, with significant maintenance of skills six months following treatment. Individuals with spastic dysarthria may have impaired respiration and phonation secondary to increased muscle tone and muscle weakness.
Respiratory training was not indicated for TBI02 because his breath support was adequate for speech across pre-treatment speech tasks and lack of an underlying respiratory disorder. Speech is a submaximal task that does not require maximal respiratory capacity. However, better coordination of respiration, phonation, and articulation may improve the intricate balance of these subsystems and have a positive impact on speech and voice characteristics. Therefore, increased duration of vowel prolongation may have been indicative of improvement in the coordination of respiration and phonation. Sustained vowel prolongation, a speech task with very limited cognitive load, was sensitive enough to capture the improved relationship between respiration and phonation.

Phonatory Stability
Phonatory stability parameters were selected to determine the impact of the

Lip and Lingual Pressure Exerted
A large treatment effect was evident for lip pressure exerted immediately posttreatment. However, this treatment effect was not maintained six months following treatment, with results decreasing to a medium treatment effect. These results indicated that some improvements in lip pressure exerted were maintained six months following treatment. Improvements in lip pressure exerted were related to direct training through labial tasks using the IOPI that were completed at the start of each session across four weeks. Increased pressure exerted was related to improved awareness of labial movement during speech production. These improvements in lip 36 pressure exerted were especially critical in improvements of vowel space area and particularly F2 values of /u/ post-treatment, which were related to increased speech intelligibility. The corner vowel /u/ is a high-back, rounded vowel, which means the tongue is in a high, back position in the oral cavity and the lips are protruded during production. Lip rounding is a vowel space dimension that is independent of high-low and front-back tongue positioning and has an impact on formant frequencies.  Therefore, it is critical to thoroughly evaluate a patient's level of cognitive-linguistic abilities to determine whether an intensive articulation treatment is an appropriate treatment option. TBI02's cognitive-linguistic abilities appeared to be more severe post-treatment through informal observations due to increased speech intelligibility and comprehensibility. For example, TBI02 demonstrated more severe deficits in orientation and memory when he inaccurately answered a simple question (e.g., "What day is tomorrow?") using his clear speech techniques. This suggested that additional speech exercise may have been warranted to facilitate generalization of clear speech outside of the treatment environment. It is critical that treatment be structured based on the participant's physiology motor speech deficits, as well as, cognitive-linguistic and language abilities. In addition, TBI02's strained-strangled vocal quality associated with spasticity dysarthria may have had a great impact on listener perception. It is recommended that future studies evaluate the effectiveness of an intensive articulation treatment based on principles of motor learning with other types of dysarthria, such as flaccid dysarthria, which is associated with a less distracting vocal quality.
The post-treatment evaluation results indicated significant improvements in single word intelligibility for this individual. However, improvements in single word intelligibility were not maintained six months following treatment, illustrating reduced treatment effect over time and little maintenance of the targeted communicative behavior. This may have been due to the complex cognitive-linguistic deficits associated with TBI02's brain injury. Therefore, it is recommended that future research studies investigate the feasibility and response to treatment of an intensive speech treatment based on the motor learning literature with increased treatment duration. Duration of treatment should increase to four times per week for six weeks to accommodate cognitive-linguistic deficits associated with TBI. People with nonprogressive dysarthria may need to establish new motor programs for speech motor control and it is possible that a longer treatment duration might facilitate internalization of the cue to speak clearly and reduce reliance on external feedback for greater generalization during functional communication and social participation. The current study was a single subject case study, so the findings cannot be generalized to the population of people who have dysarthria secondary to a TBI. Future studies should include more participants and follow-up evaluations at one and three months to determine whether increased duration of treatment facilitates generalization of improved intelligibility across speech tasks over time and the point in which a decline in intelligibility may begin due to cognitive-linguistic deficits and/or lack of consistent completion of homework tasks. The improvements measured immediately post-and six months following treatment cannot be generalized to all individuals with dysarthria secondary to TBI, but his positive response to treatment indicated that individuals with chronic dysarthria can improve speech intelligibility, even 20 years post injury.
Therefore, further studies should be completed to determine whether similar 40 improvements in speech intelligibility and comprehensibility are made with additional individuals with chronic dysarthria acquired from TBI.