Auditory Training: From Speech Sounds to Heart Sounds

Summary

Auditory training is an intervention method used in rehabilitative audiology that aims to help individuals with hearing loss use their residual hearing maximally. It emphasizes the development of listening skills to improve the recognition and interpretation of speech sounds despite limited hearing ability. This chapter explains how auditory training techniques may be adapted to help medical students and physicians improve their listening skills for heart auscultation. First, research supporting auditory training efficacy for enhancing sound perception in people with and without hearing loss is reviewed, followed by a discussion of some of the auditory training strategies that are believed to promote auditory learning. The chapter then briefly describes how principles of auditory training have been applied to the design of a computer-assisted auditory training program that helps medical students and physicians develop a better mastery of the auditory skills necessary for differentiating between innocent and pathological heart murmurs.

In the field of audiology, auditory training refers to the process involved in improving the auditory skills of individuals with hearing loss through structured and repetitive listening exercises. In a nutshell, auditory training consists of exercises, also known as listening trials, where the person (1) listens to a large number of presentations of speech sounds or other kinds of sounds, (2) makes a judgement after listening to each presentation such as identifying the sound heard, and (3) receives feedback after each attempt about whether the judgment was correct or incorrect. A basic premise to this type of intervention is the notion that hearing is a sense but listening is a skill that can be improved with practice. Hearing loss is not only characterized by a reduction in the detection of auditory signals, but is often also accompanied by deficits in frequency and temporal resolution which can cause auditory signals to be perceived in a distorted fashion. Degraded auditory signals make speech recognition more difficult, particularly in the presence of background noise. While many people with hearing loss can be helped adequately with hearing aids or cochlear implants alone, others require more intensive hearing rehabilitation, including auditory training, for optimal speech perception with their hearing devices. Auditory training does not improve hearing levels; rather it helps individuals with hearing loss listen more effectively so that their ability to recognize speech sounds may be improved. Originally primarily used with children with hearing loss (and now integrated into auditory-verbal therapy), auditory training is also advocated for adults with auditory deficits acquired later in adulthood^1-4 and for normally hearing children with auditory processing disorders^5,6 or language learning difficulties^7,8 Although most research on this topic has addressed primarily the benefits of auditory training for the identification of speech sounds, some attention has also been given to the impact of such training on the perception of other kinds of sounds. Therefore, auditory training may also be a valuable approach in teaching listening skills for auscultation of heart sounds.

Auditory training efficacy

Support for auditory training efficacy is well documented in speech perception research and in the neuroscience literature^2,3,7,9-22. Hearing aid and cochlear implant users have been shown to make gains in the perception of speech sounds following intensive auditory training^10,11,15. For example, in the study by Woods and Yund¹⁰, older adults fitted with hearing aids received an auditory training program consisting of numerous repetitions of listening exercises involving 54 nonsense syllables recorded by two talkers. Research participants underwent approximately one-hour long training sessions, five days per week for a period of eight weeks. Post-training, they showed significant improvement in their ability to identify the syllables compared to the test sessions prior to training. Improvement was noted within one week of training and performance continued to increase steadily over the course of the eight weeks. Moreover, the enhanced ability to identify the syllables generalized to untrained voices. That is, improvement was observed not only for the two voices used for the training program, but also for two novel talkers who were recorded speaking the same syllables and used only for the test sessions.

Other researchers have examined the impact of auditory training on the perception of more complex speech materials such as words and sentences^2,3,12-14. After a 12-week training program that included repeated presentations of a set of 150 words, adults with hearing loss improved their performance on the trained words by about 40% and maintained their performance for at least three months post-training¹². In a separate study3 using a much larger set of different words (600 words) but a similar duration of training paradigm, thus resulting in less training time on each word, participants showed a smaller, albeit significant, improvement in scores. That is, scores improved only by 20% post-training. As pointed out by Humes et al³, these data suggest that listening to many repetitions of a smaller set of words may lead to greater gains in auditory perceptual skills than listening to fewer repetitions of a much larger number of words.

The ultimate goal of auditory training is to help individuals who use hearing aids or cochlear implants recognize messages spoken during everyday conversations. Adults with hearing loss have been shown to improve significantly on measures of self-reported hearing difficulties during everyday social interactions after completing the Listening and Communication Enhancement (LACETM) auditory training program². Such transfer of skills to daily situations relies on two critical issues in auditory training: first, the generalization of auditory skills to situations involving listening to novel materials, i.e., words or sentences not used during training, and second, the generalization of skills when listening to novel speakers or untrained voices. A speech sound can be somewhat acoustically different when it is articulated by different talkers who vary in voice pitch, voice level, and articulation patterns; therefore people with hearing loss must be able to transfer their auditory skills when listening to novel and unfamiliar talkers. Several studies have used multiple talkers during training to facilitate the transfer of auditory perceptual skills to talkers not used during training^{2,3,10-12,14,20}; and, in general, data show that the improved skills do generalize to novel talkers.

The effect of auditory training has also been investigated in normally hearing individuals, using non-speech stimuli. Research by Moore and Amitay⁷ showed that normally hearing adults can significantly improve their ability to perform frequency discrimination tasks after listening to 1500 to 2000 trials in less than two hours of training. In this study, participants were asked to listen to two or more pure tones varying in frequency and requested to identify the higher or lower pitch pure tone through matching exercises or picking the odd tone out of a set. The difficulty of the listening task was adaptive, such that the trial immediately following a correct response included pure tones that were closer in frequency whereas an incorrect response was followed by a trial with pure tones more disparate in frequency, thus keeping the task sufficiently challenging. Following training, participants demonstrated a decrease in the frequency difference that they required to discriminate between the pure tones.

There is also strong neurophysiologic evidence that suggests that auditory listening exercises can affect neural activity in the auditory system^17-22. Tremblay, Kraus, McGee, Ponton, and Otis¹⁸ trained young normally hearing adults to identify subtle differences between two acoustically similar syllables (“mba” and “ba”) and measured their brain activity using auditory cortical evoked potentials before and after training. Initially, the two syllables were both perceived as “ba” but with training, participants were able to distinguish between “mba” and “ba”. As the participants’ ability to distinguish between the two sounds progressed, there were accompanying changes in auditory cortical evoked potential waveform morphology. Neurophysiological changes post auditory training have been observed to occur rapidly, i.e., after 45 minutes of training,19 to precede improvements in auditory perceptual skills in some people¹⁹, and to be maintained at least 36 hours after training²². Moreover, neurophysiological changes have been shown to generalize to novel sounds not used during training²⁰.

Principles of auditory training

Many auditory training programs share a number of common training principles, such as allowing for multiple repetitions of the sounds used for training, providing listeners with immediate feedback on their performance following each listening trial, and progressively increasing the difficulty level of the listening tasks. The next sections of this chapter outline some of the auditory training parameters that are viewed as essential for promoting auditory learning.

Multiple repetitions of stimuli

It is well accepted that the optimal condition for auditory perceptual learning to occur incorporates intensive training that involves actively listening to many, many items during successive training sessions conducted over a relatively short period of time^2-4,10,23,24. However, it is less clear which specific training protocols are most effective. Some researchers have used fairly long training regimes, such as approximately one hour of training several days per week for three to four consecutive weeks^14,15,24 or even up to eight to 12 weeks^3,10,12; however, auditory perceptual changes were generally observed within the first week or two of training, with performance continuing to improve over the subsequent weeks of training^2,10. In contrast, others researchers have documented auditory perceptual learning after much shorter training paradigms, such as a total of four to six training sessions all concentrated in one week^16,18,21,25. Studies that have documented neurophysiological changes induced by auditory training typically observed such changes following a small number of auditory training sessions^18,21 or even following a single training session^19,22. Individual variability in auditory learning following training has also been noted on perceptual tasks, with some individuals learning at a faster rate than others^7,19,25, as well as in the maintenance of neurophysiological changes post-training²².

In appears that whether one conducts auditory training sessions daily versus weekly may have less influence on improvements in performance than the actual amount of training or total number of training sessions. Nogaki et al²⁵ compared perceptual skills of normally hearing listeners after completing five auditory training sessions that were delivered either within one week, three times per week, or once per week for five weeks. Results showed that training rate did not have an impact on performance. On the other hand, the specific auditory training task may be more likely to have an effect on the number of repetitions needed to yield an improvement in performance. Wright and Sabin²⁶ investigated the number of trials needed for normally hearing young adults to show an improvement on a pure tone frequency discrimination task and a temporal-interval auditory discrimination task. Participants listened to either 360 trials or 900 trials daily for six days. For the temporal-interval task, improvement was shown with 360 listening trials per day, and subjecting the listeners to additional practice trials did not lead to greater gains. In contrast, for the frequency discrimination task, 360 listening trials were insufficient to produce auditory learning but improvements occurred with 900 trials. Also using a frequency discrimination task with normally hearing adults, Moore and Amitay⁷ noted auditory perceptual improvements after 500 trials, which continued to increase with additional trials until a plateau was reached after 1500 to 2000 trials.

Feedback and task difficulty level

In the vast majority of the studies on auditory training, feedback on performance was provided immediately after each listening trial^{2,3,10,14,16,18,21}. Indeed, the importance of providing the listener with immediate feedback regarding progress or lack of progress during training is strongly advocated as a way to promote optimal auditory learning^4,5,9,13,27. When a group of young normally hearing listeners who received feedback during auditory training was compared to another group who did not, lower gains in word recognition ability were observed for the group who did not receive feedback¹³.

Computer-assisted auditory training programs are well suited for easy implementation of immediate feedback after each listening trial^{2,4,10,14,24,28,29}. Moreover, as listening exercises must be challenging enough to maintain motivation and interest but yet not so difficult that they create frustration^2,4-6,23,24, computer-assisted auditory training programs enable difficulty levels to be automatically adapted based on the performance of each individual. For example, the LACETM program² uses an adaptive difficulty level where a correctly identified sentence is followed immediately by a more difficult sentence and an incorrectly identified sentence is followed by an easier sentence. For listening trials using a multiple-choice format, the task is increased in difficulty following three consecutive correct responses.

Designing auditory training steps of progressive difficulty

Listening exercises that progressively increase in difficulty are an integral part of auditory training programs. Erber^30,31 proposed a hierarchy of listening skills outlining four levels of auditory skills: (1) sound awareness, (2) sound discrimination, (3) sound identification, and (4) sound comprehension. This auditory skills hierarchy is often used as a framework for the organization of auditory training curricula^{2, 6, 23, 24}. The first level, sound awareness, is the most basic auditory skill and simply refers to being able to detect the absence or presence of a sound. The second auditory skill level is sound discrimination which refers to the ability to judge whether two or more sounds are the same or different from one another, regardless of whether one is able to associate meaning with the sounds or name them. Obviously, the more disparate two sounds are acoustically, the easier it is to judge whether they are the same or different. Therefore, when progressing from easy to more difficult listening exercises, acoustically dissimilar pairs of sounds that involve broad or gross discriminations are used before gradually moving to similar pairs of sounds that require fine discriminations^23,24,31. For example, a listening trial involving gross sound discriminations may require a person to listen to short versus long sounds, such as judging whether the monosyllabic word “bat” sounds the same as, or different from, the multisyllabic word “banana”. This listening exercise merely requires the discrimination of speech duration cues and does not require the recognition of the words “bat” and “banana”. In comparison, asking a person to discriminate between two consonants that share some common acoustic properties, such as the consonants /g/ and /k/, targets finer discrimination skills and therefore represents a more challenging auditory task. Auditory exercises to improve discrimination of these consonants may involve listening to pairs of words such as “goat-coat”, “coat-goat”, “goat-goat”, or “coat-coat”, and judging whether the two words in each pair are the same or different. Other types of auditory discrimination exercises may involve picking the odd sound out of a set (e.g., “coat, coat, goat, coat”).

The third level in Erber’s³⁰ hierarchy of auditory skills is sound identification which involves the ability to correctly label or name a sound or word that is heard. Two categories of sound identification tasks may be used for planning the difficulty level of auditory training exercises: closed-sets and open-sets. In a closed-set format, a restricted choice of alternative responses is provided in a written or picture format. For example, individuals may be provided visually with the following closed-set of words: “bead, seed, mean, wreath”; they then hear the target word “seed” and are asked to pick the correct word among the set of four, similar to a multiple-choice test. In contrast, in an open-set the target word “seed” would be presented in isolation through audition only, and the person is simply asked to repeat the word that was heard.

Auditory training exercises typically gradually progress from closed-sets to open-sets as the open-set response format is a much more challenging task than the closed-set response format^22,31,32. When constructing closed-sets, the acoustic characteristics of the items within the set are taken into account and related to hearing loss characteristics as a way to vary listening difficulty. For example, people with hearing loss generally hear low frequencies better than high frequencies; therefore vowels, having greater low frequency energy, are usually easier to perceive than consonants which tend to be weaker with many characterized by high frequency energy (for example, /s/, /f/, /th/, /h/, /sh/). Thus, an easier sound identification task may involve a closed-set of items that vary in both consonants and vowels, while a more advanced difficulty level may comprise items that differ only by high frequency consonants that are similar in their acoustic properties, such as “sigh, shy, hi, thigh”.

The final level in Erber’s³⁰ hierarchy of auditory skills is sound comprehension. This refers to the ability to interpret the sounds that have been identified and to understand the meaning of spoken messages. Facilitating the generalization of improved auditory skills to a variety of speech materials, talkers, and acoustic environments, is an important aspect of auditory training. To help achieve the transfer of skills to novel or unfamiliar talkers, it is advocated that multiple speakers be used during training^23,24, and that, post-training, individuals’ auditory skills be assessed using untrained stimuli⁴. The use of multiple talkers can be easily implemented with computer-assisted listening exercises with recorded speech. The auditory training program developed by Stacey et al¹⁴ for example, employs twenty different talkers representing male and female adults’ as well as children’s voices.

An auditory training program for the recognition of innocent and pathological heart murmurs

Caissie and Finley adapted auditory training techniques employed in rehabilitative audiology to develop auditory training exercises to assist medical students in recognizing innocent and pathological heart murmurs (Chapter 6). The main intent was to develop a computer-assisted teaching tool enabling the distinction of those murmurs that are pathological and that require further medical investigation from the far more common innocent murmurs. Key elements of the program include listening to multiple repetitions of heart murmurs recorded from a variety of patients, provision of immediate feedback after each listening trial, and progressive adjustment of task difficulty levels tailored to individual learners’ performance. Listening exercises were designed following principles of auditory training described earlier in this chapter. The program consists of several closed-set listening exercises that progress in difficulty level and that are based to some extent on Erber’s³⁰ hierarchy of auditory skills. The program does not address the detection of murmurs (sound awareness level) in comparison to normal heart sounds with absent murmurs; rather it emphasizes first the discrimination, and later the identification, of innocent versus pathological murmurs.

The computer-assisted program consists of three levels of training. Level I of the training program employs a same-different sound discrimination task; Level II uses a sound identification task with a closed-set response format; and Level III involves the identification of heart murmurs presented in isolation. Multiple training steps are embedded within each of the three levels. As they progress through the auditory training program, students are required to listen to hundreds of repetitions of innocent and pathological heart murmurs. Similar to computer games, they must achieve a certain correct performance criterion to be able to advance to the next step. More details on the construction of the listening exercises are presented in Chapter 6.

As for any auditory training programs, an important aspect is the generalization of training to heart sounds from novel or unfamiliar patients, that is, when listening to heart murmurs that were not utilized during training. Consistent with auditory training protocols that use speech materials recorded by several talkers as a way to facilitate transfer of auditory skills to novel or untrained voices^2,3,10-14, the heart sounds training program was designed using heart murmurs recorded from a large number of patients to help the generalization of skills to novel patients. Support for generalization of auditory skills has been documented when listening to speech sounds produced by novel voices^10,11 as well as when listening to non-speech stimuli⁷. Improved proficiency in cardiac auscultation has also been found to generalize to novel patients following auditory training that involved listening to numerous repetitions of six pathological heart sounds³⁴.

The efficacy of repetitive and structured listening exercises for improving speech perception skills is well documented for individuals with normal hearing and with hearing loss. Similar auditory training procedures also appear to be highly promising for helping medical students and physicians gain a better mastery of the auditory skills necessary for the correct identification of innocent and pathological heart murmurs. The following chapter provides preliminary data to examine the efficacy of auditory training as an innovative teaching approach for heart auscultation.