Accepted Manuscript

Title: Feasibility study of a game integrating assessment andtherapy of tinnitus

Author: K Wise K Kobayashi G.D. Searchfield

PII: S0165-0270(15)00141-7DOI: http://dx.doi.org/doi:10.1016/j.jneumeth.2015.04.002Reference: NSM 7207

To appear in: Journal of Neuroscience Methods

Received date: 7-1-2015Revised date: 1-4-2015Accepted date: 2-4-2015

Please cite this article as: Wise K, Kobayashi K, Searchfield GD, Feasibility study of agame integrating assessment and therapy of tinnitus., Journal of Neuroscience Methods(2015), http://dx.doi.org/10.1016/j.jneumeth.2015.04.002

This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.

Page 1 of 27

Accep

ted

Man

uscr

ipt

1

Feasibility study of a game integrating assessment and therapy of tinnitus.

Wise K1, Kobayashi K1, Searchfield, G.D.1,2

1Centre for Brain Research & Audiology Section Faculty of Medical and Health Sciences,

The University of Auckland.

2Tinnitus Research Initiative

Correspondence:

Dr Grant D Searchfield

Section of Audiology

Faculty of Medical and Health Sciences

The University of Auckland

Private Bag 92019

Auckland

New Zealand

g.searchfield@auckland.ac.nz

Page 2 of 27

Accep

ted

Man

uscr

ipt

2

Abstract

Background: Tinnitus, head and ear noise, is due to maladaptive plastic changes in auditory

and associated neural networks. Tinnitus has been traditionally managed through the use of

sound to passively mask or facilitate habituation to tinnitus, a process that may take 6-12

months. New Method: A game-based perceptual training method, requiring localization and

selective attention to sounds, was developed and customized to the individual’s tinnitus

perception. Eight participants tested the games usability at home. Results: Each participant

successfully completed 30 minutes of training, for 20 days, along with daily psychoacoustic

assessment of tinnitus pitch and loudness. The training period and intensity of training

appears sufficient to reduce tinnitus handicap. Comparison with Existing Method(s): The

training approach used may be a viable alternative to frequency discrimination based training

for treating tinnitus (Hoare et al., 2014) and a useful tool in exploring learning mechanisms in

the auditory system. Conclusions: Integration of tinnitus assessment with therapy in a game

is feasible, and the method(s) warrant further investigation.

Highlights

A computer game to assess and treat tinnitus was determined to be a feasible approach

to tinnitus treatment.

Twenty days of 30 mins game play resulted in significant reductions in THI scores

and is the recommended training period for a future randomised trial.

The new method enabled tinnitus evaluation consistent with an ecological model of

tinnitus.

Page 3 of 27

Accep

ted

Man

uscr

ipt

3

Daily variance in tinnitus complicates understanding the neuroscientific basis of

tinnitus.

Keywords ��

Tinnitus, assessment, therapy, game, plasticity, psychoacoustics

Abbreviations

THI: Tinnitus Handicap Inventory

MML: Minimum Masking Level

FDT: Frequency Discrimination Training

Page 4 of 27

Accep

ted

Man

uscr

ipt

4

1. Introduction

Tinnitus is the perception of an auditory object in the absence of an acoustic event

(Searchfield, 2014) and is a common experience following hearing loss (Sindhusake et al.,

2003). It is considered to be the result of maladaptive plasticity in central auditory pathways

following changes in input from the periphery (Roberts et al., 2010). Sound based therapies

are often part of a tinnitus management plan (Folmer et al., 2006; Henry et al., 2006;

Searchfield and Baguley, 2010). “Sound therapy” usually attempts to passively mask tinnitus,

making it less audible (Jastreboff and Jastreboff, 2000; Vernon and Meikle, 2000). Sound

therapy may take greater than 12 months to reach full effectiveness (Henry et al., 2006;

Herraiz et al., 2005). An alternative use of sound requires active listening in the form of

various perceptual training paradigms (Hoare et al., 2010). Perceptual training approaches are

based on evidence that the auditory brain is plastic and can be modified by targeted auditory

stimulation (Hoare et al., 2010; Jepsen et al., 2010; Searchfield, 2007). Several perceptual

training methods have been investigated since the concept was first applied to tinnitus by Flor

et al. (Flor et al., 2004). Frequency discrimination training (FDT) has been the most widely

published training method for tinnitus (Hoare et al., 2010). FDT attempts to reverse

maladaptive over-representation of tinnitus pitch frequencies in the cortical tonotopic map

thought to lead to tinnitus (Hoare et al., 2012). Tonotopic maps refer to the orderly

representation of sound frequencies spatially in the cortex. Following injury to the ear these

cortical maps change (Eggermont and Roberts, 2004). Frequency specific damage to the

cochlea leads to a reduction in input to affected frequencies centrally. Over time this may

lead to an expansion of neighboring tonotopic regions in the now under-stimulated region.

The increased representation, or synchronization, of neural firing in expanded regions may

lead to tinnitus perception (Eggermont and Roberts, 2004). Through FDT the listener learns

Page 5 of 27

Accep

ted

Man

uscr

ipt

5

to distinguish between frequencies of sound. Applied to tinnitus it was thought that by

training patients to discriminate sounds neighboring tinnitus pitch tonotopic map might

functionally expand into the region of tinnitus pitch forcing a reduction in its representation

(Hoare et al., 2010) (Flor et al., 2004). Initial studies investigating FDT for tinnitus were

limited but the majority provided some support for the training concept (Hoare et al., 2010).

A barrier to the implementation of FDT is the repetitive nature of the task and need for

patient compliance to regular (often daily) training of greater than 30 mins; participants may

become bored. Hoare et al (2014) sought a solution through building FDT into games,

however the games did not lead to higher compliance, and resulted in only 7% improvement

in Tinnitus handicap Questionnaire, with no statistically significant change in THI. A further

shortcoming of FDT methods is the frequency to be trained is established once, at training

onset, and does not consider change in frequency overtime. Psychoacoustic evaluation of

tinnitus typically involves matching sound characteristics to tinnitus in order to describe what

it sounds like. Standardised methods for psychoacoustic evaluation of tinnitus were proposed

30 years ago and have been implemented in research and clinics through audiometers and

more recently software applications (Henry et al., 2000). Tinnitus pitch may change

naturally, and might be expected to change with tonotopic reorganisation. Failure to account

for changes in tinnitus pitch may lead to training at incorrect frequencies, negating attempts

to remap tinnitus.

Although FDT has been widely used there is increasing evidence that tonotopic map

reorganization is unlikely to be the direct mechanism of tinnitus (Eggermont, 2015; Langers

et al., 2012) and training is effective when learning to distinguish between frequencies lower

Page 6 of 27

Accep

ted

Man

uscr

ipt

6

from the supposed tinnitus causing areas (Hoare et al., 2012). Hoare et al (2012)

consequently hypothesised that FDT benefit is most likely derived from less attention to

tinnitus, rather than FDT per-se.

Given this evidence we have been developing auditory training methods for tinnitus

treatment, by addressing various perceptual components of tinnitus that have focused on

methods for perceptual training other than FDT (Searchfield et al., 2011). It has been

hypothesized that some of the distress associated with tinnitus is related to an inability to

understand tinnitus in the perception of real-world sounds (Searchfield, 2014). Several of the

proposed perceptual tasks require participants to identify and localize sounds according to

frequency and location. By actively reducing attention to sounds resembling tinnitus it is

expected that the perception of tinnitus will change so that tinnitus can be placed in context

and take on less importance.

The basis of the game described in this study, given the name “Terrain”, was an auditory

attention and localisation task customised to each individual player’s tinnitus perception.

Customisation in this context means the psychoacoustic measurement of the individual’s

tinnitus, then using this information to generate sounds similar (distracter) and dissimilar

(target) to their tinnitus. Subsequent distracters were designed to be dissimilar to each

individual player’s subjective tinnitus sensation. The end-result was a straight-forward

selective attention task with frequency and localisation cues. Participants attended to a target

different to their tinnitus percept, while attempting to ignore distracters. Positive

reinforcement was provided by allocating game points and progression to the next level of the

game; responding to a sound similar to tinnitus was not rewarded.

Page 7 of 27

Accep

ted

Man

uscr

ipt

7

The game was the result of a heuristic exercise to develop a game-based training approach

integrating assessment and tinnitus localisation, attention and potentially conditioned

responses. The feasibility study described here was needed to further evaluate the usability

and effect of the game before entering larger-scale clinical trials. This study was undertaken

to ascertain if an attention training programme, daily customised to individual’s tinnitus and

day-to-day changes in tinnitus perception, might be possible. The small open trial would

ascertain appropriate methods for a future RCT. The practicality and usefulness of obtaining

daily measures of tinnitus using the computer interface were ascertained. Following the trial

the intention was to improve the training methods, based on feedback of effectiveness and

usability.

1.1 Aims and Hypotheses

The aims of this study were to determine if a game-based training programme, customised to

account for daily changes in tinnitus percept was feasible and ascertain appropriate methods

for a future controlled trial.

2.0 Methods

This study was approved by the University of Auckland’s Human Participants Ethics Committee.

2.1 Participants

Seven males and one female aged 39 – 64 years (mean 56, SD 8.1) participated in the study.

They were recruited from a research database of adult participants (> 18 years of age) with

Page 8 of 27

Accep

ted

Man

uscr

ipt

8

bilateral hearing loss (hearing thresholds > 20 dB HL, for 0.25 – 8 kHz) and chronic tinnitus

(> 6 months; small problem or greater). Participants with severe or greater (> 70 dB HL)

hearing loss for the subjective tinnitus pitch-match frequency and octave side-bands were

excluded, as were those volunteers with conductive hearing impairment or insufficient

tinnitus complaint. Those planning to join the study were instructed that participation

required hearing and tinnitus assessment, completing questionnaires concerning cognitive

state, depression, anxiety and tinnitus severity. Participants had to have daily (30-minutes-

per-day, for 20 consecutive days) access to a computer (Microsoft® Windows® XP or

Windows® 7) compatible with the LabVIEW™ Terrain programme. Twelve individuals

expressed interest in participating, 4 were excluded or withdrew due to: a death in the family,

an inability to adhere to the daily training, a back injury, severe-to-profound high-frequency

loss.

All participants completed an initial case history, audiological and tinnitus assessment (pitch

and loudness matches). Hospital Anxiety Depression Scale (HADS, Zigmond & Snaith,

1983) and Adult ADHD Self-Report Scales (ASRS-v1.1, Adler, Kessler & Spencer, 2003)

were administered to ascertain significant depression, anxiety or attention deficits. Minimum

masking levels (MML) and the Tinnitus Handicap Inventory (THI, Newman, et al., 1996)

were completed 3 times, upon recruitment, immediately before and immediately after 20 days

training. MML was determined by presentation of narrow-band noise at tF and centre

frequencies of 0.5 kHz, 1 kHz, 2 kHz, and 4 kHz, to determine the sensation level that just

masked the tinnitus. An ascending technique was employed, initially using 5 dB steps to gain

a course estimate of intensity required to just mask tinnitus. The procedure was then

repeated twice, using 1 dB step-sizes and an average MML was calculated.

Page 9 of 27

Accep

ted

Man

uscr

ipt

9

Tinnitus was psychoacoustically characterised using tinnitus assessment software (© The

University of Auckland). The Tinnitus assessment software was developed using

LabVIEW™ 8 installed on a Dell desktop computer work station with an Intel® Xeon®

processor. Stimuli presentation was via an M-Audio, Fast Track Pro, mobile USB

audio/MIDI interface with preamps, to ensure the stimuli were audible. The software

obtained: audiometry, tinnitus pitch-match and bandwidth, environmental sound-match to

tinnitus, tinnitus localisation (3-Dimensional) and tinnitus loudness. The data were saved to a

computer file on an Apacer USB (AH325 4GB) flash drive for each individual to install on

their PC as the initial calibration point for the game.

2.2 Usability

Participants were asked to complete a daily log of use. The log requested answers to two

open-ended questions asking: What issues were experienced with the game? Do you have

any comments? Feedback was received from all participants.

2.3 Perceptual Training

The training game was demonstrated to participants using a Dell desktop computer with an

Intel® Xeon® Processor. The game was then downloaded to the USB flash drive and

participants were supplied with iPOD® ear-bud phones (response range 20 Hz - 20 kHz, 32

Ω impedance) for use with their home PC. After installing and opening the software but

before proceeding to actual playing, participants adjusted the intensity level of a broadband

noise to threshold of detection. This ensured the participant’s computer audio controls were

Page 10 of 27

Accep

ted

Man

uscr

ipt

10

properly enabled and sounds would be comfortable. The perceptual training game “Terrain”

(Searchfield et al., 2011) began with a series of seven “Daily Tinnitus Calibration” (DTC)

tasks to account for day-to-day shifts in tinnitus perception relative to the initial assessment

(Pitch-match [Is your tinnitus pitch lower or higher?]; Sound Bandwidth [Is your tinnitus

more hissing or ringing?]; Loudness-match [Does your tinnitus sound louder or quieter?];

Laterality [Does your tinnitus localise more left, right, or centre of your head?]; 3D location

front-back [Does your tinnitus localise more to the front, back, or centre of your head?]; 3D

location up-down [Does your tinnitus localise more above/up, below/down, or centre of your

head?]; Final loudness-match [Does your tinnitus sound louder or quieter?]) All calibration

screens were similarly formatted with a straight-forward slider with parameters placed at each

end of the scale (e.g. “quiet verses loud” for the DTC loudness-match adjustment) (Figure 1).

All calibration data acquired prior to each day of game play was automatically saved to the

flash drive allowing changes in tinnitus characterisation over the course of gaming to be

analysed.

Figure 1. Screen shot of one of the seven calibration screens participant’s had to navigate

through, prior to training.

Page 11 of 27

Accep

ted

Man

uscr

ipt

11

The DTC was followed by a sample of the five stimuli that comprised the game sounds. One

target sound was generated to be dissimilar to the tinnitus sound characterisation (data up-

loaded from the assessment), one “distracter” sound generated to be similar to the tinnitus

percept, and four other distracters dissimilar to the tinnitus sound. The target stimulus was

presented and the participants used the slider adjustment to ensure it was comfortably

audible. The target stimulus was presented at a comfortable level ± 7dB (level varying with

perceived proximity to target location), tinnitus distracter at comfortable level -6dB, other

distracters at comfortable level. The frequency of the target stimulus was 1/3 octave below

tinnitus pitch, the first distractor (D1) was at tinnitus pitch, D2 at 2/3 octave below tinnitus

pitch, D3 was 1/3 octave higher than tinnitus pitch, and D4 was + 2/3 octave higher than

tinnitus pitch. The directions of stimuli were determined randomly in a horizontal plane (0,

45, 90, 135, 180, 225, 270, 315 azimuth) except D1 that was fixed at the perceived location

of the tinnitus.

Once the DTC and stimuli intensity tasks were completed, game play initiated. The game was

programmed to run for 20 consecutive days, providing 30-minutes-per-day of training.

Participants received a pop-up message congratulating them on reaching the target amount of

game-play. Once each gaming session concluded, the game programme automatically

generated text (.txt) files, up-loading the data to an embedded file saved to the USB flash

drive. Data were date-stamped with the date, time, scoring and game play information.

Terrain was designed as an auditory-only game, Participants were instructed to imagine a grid

of concealed squares lying underneath the black, featureless screen (Figure 2).

Page 12 of 27

Accep

ted

Man

uscr

ipt

12

Figure 2. Screen shot of the Terrain Game. Terrain operated solely as an auditory-based

training.

Play involved using the keyboard up, down, left, right arrow keys to navigate toward a

pulsing target sound. The target sound was described as a “hidden treasure” situated in one

square somewhere within the concealed grid. If movement was in the correct direction

(toward the target) sound intensity would increase. Movement in the opposite direction

would cause the target sound’s intensity to decrease. Participants were encouraged to listen

carefully, move methodically, and use target sound intensity to determine correct navigation.

When the treasure location was successfully identified, the score increased by one and the

play resumed, with the treasure re-hidden in a novel location.

Based on score, Terrain progressed through 5 levels of difficulty. The object of level 1 was

to acclimatise players to the game format, familiarise them with the target sound (removed

from their tinnitus percept) and encourage them to gain the highest score possible. Each new

Page 13 of 27

Accep

ted

Man

uscr

ipt

13

level would introduce a non-target distracter sound. The first distracter introduced in level 2,

was designed to be similar to the player’s tinnitus characterisation. Subsequent distracter

sounds were different to the tinnitus. There were a maximum of 5 distracter sounds possible

(level 5). The object was the same for each level, locate the target treasure, but made difficult

with the addition of non-target stimuli. Repetitive incorrect movement e.g. repeated pressing

of the navigation keys in a non-target direction, would generate a 5 second time penalty,

deducting 5 seconds from the allotted time to play. If an attempt was made to halt or exit the

game without engaging in the game scenario for the recommended 30 minutes, a message

would appear reminding players they would not advance to the next day of play, until the

compulsory 30 minutes was completed. All aspects of daily game play, namely timing, level

and scores achieved, pause or halt sequences, and warnings were saved to the flash drive.

Once 20 days of gaming for 30-minutes-per-day were accomplished, the programme

generated a pop-up window, congratulating participants on completing the training.

2.4 Analysis

An exploratory analysis was carried out on the quantitative data using IBM SPSS Statistics

19 for Windows®. Qualitative measures of usability were collated and coded to themes using

an inductive framework. GDS and KJW independently reviewed and undertook the coding,

before reaching a consensus. KK identified technical solutions were appropriate.

3.0 Results

3.1 Participant’s characteristics

Page 14 of 27

Accep

ted

Man

uscr

ipt

14

The HADS and ASRS-v1.1 results were normal for all participants, ruling out significant

anxiety, depression, stress and attention deficits. Tinnitus Duration was 14.63 (SD 12.01)

years; tinnitus pitch was 5.74 (SD 3.75) kHz and initial THI 38 (SD 21). Hearing thresholds

were normal for the low frequencies up to approximately 1 kHz and then gradually sloped to

a moderately-severe or severe hearing deficit for the higher frequencies (Figure 3). There

was greater variance in results in the right ear, primarily due to a flat severe hearing loss in

one participant.

Page 15 of 27

Accep

ted

Man

uscr

ipt

15

Figure 3. Mean audiometric thresholds (solid line) and mean plus and minus the standard

error (dashed lines) for participants’ right (A) and left (B) ears (N =8).

3.2 Usability

The responses to the daily log identified several usability and technical issues with the game.

Supporting statements from participants for each theme are provided below. Technical issues

and recommended solutions are provided in Table 1.

3.2.1 Difficulty

Several participants had difficulty in performing the task, particularly in the first few days of

training. Some of the difficulties related to navigation in finding the source of sound.

Identifying the target over the distractors was a difficult task:

“I found that it was difficult to locate the exact source of the pulsing”

“Choosing wrong sound…real struggle with concentration and locating the sound”

“Level 3 confused again”

“Needs a dot [visual indicator] to follow – hard to hear”

“Didn’t manage to locate sound source”

Other complaints related to audibility of sounds:

“Turned up volume a little, better”

Page 16 of 27

Accep

ted

Man

uscr

ipt

16

3.2.2 Headphones

Participants observed issues with use of the supplied headphones, being variability with

insertion depth, and use of earphones in incorrect ears:

“When you put the headphone plug right in the ear, the right ear pulsed loudly”

“I mistakenly had left ear plug in right ear” [For 1 day]

3.2.3 Background noise

Undertaking the game when it was noisy at work or home contributed to poor results, while

quiet environments were helpful:

“Sunday afternoon, some disturbing outside noise”

“Hard drive noise distracting, even with headphones”

“Perhaps because I was working from home so no noise all day in a loud office” [Good

score]

3.2.4 Keyboard

In the first few days of training this was the most common technical problem. It was resolved

by the investigators assisting the participants, and was addressed in revisions of the software:

“Keyboard key is hard on laptop”

“Will try different keyboard and desktop PC”

Page 17 of 27

Accep

ted

Man

uscr

ipt

17

“Keys do not seem to be moving the cursor around the grid”

3.2.5 Incorrect game play

In learning early phases of the game some participants’ randomly pressed keys rather than

systematically trying to find the target sound by moving around the grid:

[Day 3] “So far being randomly using keys, may have to sort out some kind of pattern in my

mind”

This had two effects, the first being the intended training paradigm wasn’t initially being

followed, and second was that the penalty imposed for random key strokes reduced the

playing time by 10 seconds each occurrence.

3.2.6 Improvement with time

Although most participants initially found the game difficult, they more often than not

became aware that they were improving their performance, and this seemed to be a

motivating factor in their continued playing of the game.

Example 1:

[Day 10] “Incredibly frustrating I cannot delineate the sounds at level 2!”

Page 18 of 27

Accep

ted

Man

uscr

ipt

18

[Day 11] “Better at level 2, flying through level 1”

[Day 12] “Got to level 3!”

[Day 13] “Got to level 3 quickly, starting to identify sound at [level] 1 & 2 quite quickly”

[Day 18] “Level 4!”

Example 2:

[Day 1] “Level 1 easy…found level 3 almost impossible”

[Day 2] “Better at level 3. Easier to pick target sound”

3.2.7 Tinnitus

Participants were not specifically asked about their tinnitus in the daily log, however

occasional references were made:

“The sounds seem to be blocking the tinnitus when played”

“Interesting effect of the pulsing, quite hypnotic”

“Tinnitus had shifted slightly forward”

3.2.7 Miscellaneous user observations

Rarely did the software or windows fail or result in error messages:

Page 19 of 27

Accep

ted

Man

uscr

ipt

19

“Windows delayed write failure – this message displayed after I removed the USB stick [flash

drive]”

The participants offered thoughts during the training on how the game software could be

improved or where they were unclear on actions to take:

“Make the time in minutes? And show time played?”

“When you score a point you don't no [know] without looking”

“Need a pause button”

“What to do at end with USB [flash drive]”

Not everyone reported any issues or desired changes:

“None, all very straight forward”

3.3 Quantitative outcomes of training

Within-subjects effects of training on the THI over time (on recruitment, pre-training and

after the 20 days of Terrain Game training was completed) were investigated using a

repeated-measures ANOVA. A significant decrease in mean THI score (F(2,14)=5.782,

p=0.015) was evident after the training (Figure 4).

Page 20 of 27

Accep

ted

Man

uscr

ipt

20

Figure 4. Change in Tinnitus Handicap Inventory score over the three time intervals. The

error bars are plus and minus one standard error.

There were no significant differences in narrow band noise MMLs (centre frequency: 0.5

kHz, 1 kHz, 2 kHz, 4 kHz, and tinnitus pitch-match) before and after training. A novel aspect

of the program was the DTC and recording of each day’s tinnitus loudness and pitch (Figure

5). There was considerable intra-subject variability, especially in loudness.

Page 21 of 27

Accep

ted

Man

uscr

ipt

21

Figure 5. Tinnitus loudness (A) and pitch (B) for each participant (thin lines) and mean

(thick line) for the 20 days of training. Session 1 represents the initial tinnitus assessment,

session 21 is the DTC prior to final day’s training.

4.0 Discussion

This feasibility study supports the further investigation of the training game “Terrain” for

tinnitus management. The training here also targeted selective attention and localisation of

sound. THI scores were reduced following participants playing the Terrain Game by 6.25

points, an improvement in tinnitus handicap considered clinically-relevant (6-7 points)

(Zeman et al., 2011). The Terrain Game automated scoring of participant performance. The

game made it possible to customise stimuli to account for hearing loss and any day-to-day

changes in tinnitus perception. The gaming interface also provided a reward for playing.

Many participants were actively curious about how their gaming scores were “stacking up”

against other players and some emailed daily scores that they felt were especially good;

Page 22 of 27

Accep

ted

Man

uscr

ipt

22

progress was also entered daily into a log of use. The computer-based approach made it

possible to gain an objective index of compliance (all scores, timing of play, pause sequences

and warnings were populated to the flash drive as a text file).

This study was not designed as a clinical trial, nor did it seek to determine the mechanisms of

any benefit; future studies could measure attention and obtain objective measures (e.g. EEG

or MRI) before and after training. This study identified several improvements that should be

made to the game and research methodology before such controlled trials. From comments

made in participants’ daily log of use, potential issues and changes to consider for future use

were identified. All of the comments made were pooled and analysed into themes so that the

software could be de-bugged and re-programmed to improve end-user experience. Some

participants commented on the difficulty of the game, indicating it was challenging to

distinguish the target sound from the distracters especially for the higher levels. It was

decided for future versions the target sound’s duration should be increased. Because of the

potential for high frequency hearing loss to affect distractor audibility, it was decided that in

future versions the distractors should only be below tinnitus pitch. The bandwidth of the

distractors should also be altered and the inter-stimulus interval, which was initially

randomised should be set to a standard 250 ms. The initial version of the game had a penalty

for aggressive play (quick, successive navigation key-pressing) reducing the time of play by 5

seconds. For the future version it is recommended that a “Listen Carefully” pop-up warning

is used, to ensure all players receive the recommended 30 minutes of play. Despite some

individuals conceivably receiving less than full 30 minutes in this pilot study, there was still a

significant improvement in tinnitus achieved. Some participants, curiously investigated the

limitations of the DTC, and admitted trying slider adjustments well outside of their tinnitus

sound match, to “...see what it would do” or “...see what it would sound like” and were

Page 23 of 27

Accep

ted

Man

uscr

ipt

23

surprised by either no sound (if bandwidth adjustment was out of range) or distortion (peak-

clipping). To emphasise those DTC slider adjustments that would result in unfavourable or

inadequate output from the programme, a “clip” indicator was proposed to be added next to

the slider showing a red warning light for less-than optimal settings. For the feasibility

version of the software, there were some initial installation issues with some models of

laptop. One participant reported some difficulty with determining how to make the laptop’s

arrow (navigation) keys active for game play. This will be resolved for future versions of the

software. As several participants commented that displaying play-time in minutes rather than

seconds would be more meaningful, the next version will incorporate both minutes and

seconds. There was also a desire to be able to pause the game. Play will be able to be paused

for a brief breaks in future versions. Although the game corrected for hearing loss, to reduce

interaural loudness differences, the game may be best suited to symmetrical, or mildly

asymmetric, hearing one participant had a large asymmetry in hearing that he reported

affected his ability to localise sounds in the game. A future RCT might exclude asymmetric

hearing losses. A future study should use an appropriate control and may benefit from

additional measures of effect. The THI may not be the most responsive measure of tinnitus, a

future study might benefit from use of a measure sensitive to outcome of treatment, such as

the TFI (Meikle et al., 2012).

The treatment game could be used use largely independent from clinicians, undertaking

measures of the individual’s tinnitus in a calibration phase prior to each auditory training

session. The assessment of tinnitus on a daily basis is consistent with the suggested need to

incorporate greater ecological vailidity into our evaluations of tinnitus (Searchfield, 2014).

There was considerable variation between participants in their psychoacoustical

characterisation of tinnitus, and although some participants had fairly consistent pitch and

Page 24 of 27

Accep

ted

Man

uscr

ipt

24

loudness, others varied a great deal. Whether the variation is due to fluctuation in tinnitus or

measurement error has not been ascertained. This variation does highlight the potential error

in taking a “snapshot” of tinnitus on a single day through laboratory based imaging and EEG

studies and then attributing a neurophysiological mechanism on single measures. New

methods are needed to address the heterogenous nature of tinnitus (Searchfield, 2014).

All participants involved in the study were familiar with computers, some set the training up

on their work computer as well as home PC, applying the approach to best-suit their lifestyle.

Tinnitus impact varies greatly for individuals and therefore, any treatments that can be

individualised to address these diverse characteristics may be useful. The value of

customisation of the game will need to be evaluated, as will the mechanism of effect, it is

possible that attention modification is the mechanism of benefit. But as the game provided

reward for correct responses, an alternative mechanism might be classical conditioning. The

software recorded tinnitus-related psychoacoustic changes over the treatment period and

compliance to the treatment. The daily measurement of tinnitus pitch, loudness and especially

location, in the participants own “real” environment provides a degree of ecological validity

usually missing from psychoacoustic measures of tinnitus (Searchfield, 2014).

Our concepts have been developed to the point where we have a new, working prototype

software that is user-friendly and self-monitoring. Investigations of the mechanisms

underlying training and trials are the next step in this interactive process. The game was

implemented on PCs; a logical extension would be to transfer to mobile computing platforms

such as tablet computers and smart phones.

Page 25 of 27

Accep

ted

Man

uscr

ipt

25

5.0 Conclusions

A computer training program appears to be a feasible means to provide attention-based

perceptual training at home. The difference in THI score from the initial questionnaire

administration as compared to post-training was 6.25 points which is consistent with a

clinically-significant improvement (Zeman et al., 2011). The game enabled daily logging and

customisation of training to the individual’s tinnitus percept, although it is not known

whether such customisation is an important consideration for tinnitus training. The

computerised, take-home approach was determined to be a practical approach to tinnitus

treatment. Training could be an effective and cost efficient alternative to the clinician-

intensive interventions currently used. Trials are required to ascertain the clinical usefulness

and mechanisms of effect of the game.

6.0 Conflict of interest statement

The authors have intellectual property protection of the game concept described and intend to

make the game available commercially if future trials confirm its usefulness as an assessment

and treatment method.

7.0 References

Eggermont, J. J. Tinnitus and neural plasticity (Tonndorf lecture at XIth International Tinnitus Seminar, Berlin, 2014). Hearing Research, 2015.319C, 1-11.Eggermont JJ, Roberts LE. The neuroscience of tinnitus. Trends Neurosci, 2004; 27: 676-82.Flor H, Hoffmann D, Struve M, Diesch E. Auditory discrimination training for the treatment of tinnitus. Appl Psychophysiol Biofeedback, 2004; 29: 113-20.Folmer RL, Martin WH, Shi Y, Edlefsen LL. Tinnitus Sound Therapies. In Tyler RS, editor. Tinnitus Treatment Clinical Protocols. Thieme: New York, 2006: 176-86.Henry JA, Fausti SA, Flick CL, Helt WJ, Ellingson RM. Computer-automated clinical technique for tinnitus quantification. Am J Audiol, 2000; 9: 36-49.

Page 26 of 27

Accep

ted

Man

uscr

ipt

26

Henry JA, Schechter MA, Zaugg TL, Griest S, Jastreboff PJ, Vernon JA, Kaelin C, Meikle MB, Lyons KS, Stewart BJ. Outcomes of clinical trial: tinnitus masking versus tinnitus retraining therapy. J Am Acad Audiol, 2006; 17: 104-32.Herraiz C, Hernandez FJ, Plaza G, de los Santos G. Long-term clinical trial of tinnitus retraining therapy. Otolaryngol Head Neck Surg, 2005; 133: 774-9.Hoare D, Stacey P, Hall D. The Efficacy of Auditory Perceptual Training for Tinnitus: A Systematic Review. Ann Behav Med., 2010; 40: 313- 24.Hoare DJ, Kowalkowski VL, Hall DA. Effects of frequency discrimination training on tinnitus: results from two randomised controlled trials. Journal of the Association for Research in Otolaryngology : JARO, 2012; 13: 543-59.Hoare DJ, Van Labeke N, McCormack A, Sereda M, Smith S, Al Taher H, Kowalkowski VL, Sharples M, Hall DA. Gameplay as a source of intrinsic motivation in a randomized controlled trial of auditory training for tinnitus. PloS one, 2014; 9: e107430.Jastreboff PJ, Jastreboff MM. Tinnitus Retraining Therapy (TRT) as a method for treatment of tinnitus and hyperacusis patients. Journal of the American Academy of Audiology, 2000; 11: 162-77.Jepsen K, Sanders M, Searchfield G, Kobayashi K. Perceptual training of tinnitus. Proceedings of ‘Tinnitus Discovery’: Asia-Pacific Tinnitus Symposium, 11-12 Sept 2009. The New Zealand Medical Journal, 2010; 123 141 - 53.Langers DR, de Kleine E, van Dijk P. Tinnitus does not require macroscopic tonotopic map reorganization. Frontiers in systems neuroscience, 2012; 6: 2.Meikle MB, Henry JA, Griest SE, Stewart BJ, Abrams HB, McArdle R, Myers PJ, Newman CW, Sandridge S, Turk DC, Folmer RL, Frederick EJ, House JW, Jacobson GP, Kinney SE, Martin WH, Nagler SM, Reich GE, Searchfield G, Sweetow R, Vernon JA. The tinnitus functional index: development of a new clinical measure for chronic, intrusive tinnitus. Ear and hearing, 2012; 33: 153-76.Roberts LE, Eggermont JJ, Caspary DM, Shore SE, Melcher JR, Kaltenbach JA. Ringing Ears: The neuroscience of tinnitus. Journal of Neuroscience, 2010; 30: 14972-9.Searchfield G. Tinnitus what and where: an ecological framework. Front. Neurol., 2014; 5.Searchfield GD, Baguley DM. The Role of the Audiologist in Tinnitus Practice. In Moller AR, Langguth B, De Ridder D, Kleinjung T, editors. Textbook of Tinnitus. Springer: New York, 2010.Searchfield GD, Kobayashi K, Wise K. Interactive gaming system. New Zealand, 2011.Searchfield GD, Morrison-Low J, Wise K. Object identification and attention training for treating tinnitus. In B. Langguth GHTKAC, Møller AR, editors. Progress in Brain Research. Elsevier, 2007: 441-60.Searchfield GD, Morrison-Low, J, & Wise, K. Object identification and attention training for treating tinnitus. Progress in Brain Research, 2007; 166: 441-60.Sindhusake D, Golding M, Newall P, Rubin G, Jakobsen K, Mitchell P. Risk factors for tinnitus in a population of older adults: the blue mountains hearing study. Ear and Hearing, 2003; 24: 501-7.Vernon JA, Meikle MB. Tinnitus masking. In Tyler RS, editor. Tinnitus Handbook. Singular Publishing: San Diego, 2000: 313-56.Zeman F, Koller M, Figeiredo R, Aazevedo A, Rates M, Coelho C, Kleinjung T, de Ridder D, Langguth B, Landgrege M. Tinnitus Handicap Inventory for evaluating treatment effects: Which changes are clinically relevant? Otolaryngology-Head and Neck Surgery, 2011; 145: 282-7.

Page 27 of 27

Accep

ted

Man

uscr

ipt

27

Table 1. Participant identified issues, interpretation and potential solutions.

Theme Interpretation Technical solution

Game difficulty• Difficult to identify

target from distracters

• Game concentration requirement high

• Audibility (hearing asymmetry)

Level one is already difficult. Most participants couldn’t reach level 4. Many mistakes were made.30 min is demanding on concentration for some participants.Target/distractor confusion, need greater contrast.Asymmetric steep hearing loss with and tinnitus pitch in region of severe hearing loss.

Increase target intensity ±25dB as maximum.Change target duration from others.Separate target and distractors by 1 octave. Increase stimulus bandwidth to 1/2 octave.Change number of trials in each level to 5, except last level (5) which should be continued until time over.Add "highest score" on the game monitor as motivation.Stop random Inter-stimulus intervals, fix at 250 ms.Ensure level of tinnitus match distractor is always greater than 0dB from the comfortable level even if player moves far away. Distractors are not changed in location unless the player finds the target treasure.Stimuli presentation order is not changed unless the player finds the target treasure.Hearing asymmetry should be compensated. Changed all stimuli to lower frequencies than tinnitus pitch (except tinnitus distractor match).Comfortable loudness calibration - adjust all comfortable loudness levels for all stimuli. Comfortable loudness calibration - provide larger gain range for volume slider.

Keyboard Difficult playing with laptop keyboard, numeric keys on right in laptop keyboard don't work.

Program so number (Arrow) keys on the right of the laptop computer are available.

Incorrect play Searching entire grid, one square at a time, rather than listening and moving towards target.Random key pressing will reduce game time so can’t ensure 30 minutes playing.

Change to visual alert "Listen to target carefully!“, and remove time penalty.

Headphones Some players were not frequent headphone users and had a little difficulty with placement and rarely L/R confusion.

Ensure instructions are adequate and headphones are appropriate for the task.

Miscellaneous comments• Time display• Scoring signal• Pause button

Display minutes rather than seconds.Currently score is flashing twice.

Need pause for unexpected incidents (ringing phones, knock

doors, etc.).

Change display to minutes (0-99): seconds (0-59) display.Leave score display as adequate.Add pause button.

Change "stop" to "halt" button (to avoid confusion).