José I. Alcántara

The use of psychophysical tuning curves to explore dead regions in the cochlea.

Objective “Dead regions” are regions in the cochlea with no functioning inner hair cells (IHCs) and/or neurons. Amplification (using a hearing aid) over a frequency range corresponding to a dead region may not be beneficial and may even impair speech intelligibility. The objective of this article is to illustrate the use of psychophysical tuning curves (PTCs) as a tool for investigating dead regions and to illustrate the variety of audiogram configurations that can be associated with dead regions. We explore the influence of signal level and signal frequency to test the hypothesis that the frequency at the tip of the tuning curve defines the boundary of the dead region. Design PTCs were measured for five subjects with sensorineural hearing loss who were suspected of having dead regions. One had a relatively “flat” loss, one had a mild mid-frequency loss and three had high-frequency losses, varying in severity from 70 dB to more than 120 dB. For each PTC, the level and frequency of the sinusoidal signal were fixed, and the level of a narrowband noise masker needed just to mask the signal was determined as a function of the masker frequency. When the signal falls in a frequency region that is not “dead,” the signal is detected via IHCs with characteristic frequencies (CFs) at or close to the signal frequency. In such a case, the tip of the PTC (the masker frequency at which the masker level is lowest) lies at or close to the signal frequency. When a dead region is present, the signal is detected via IHCs with CFs different from that of the signal frequency. In such a case, the tip of the PTC is shifted away from the signal frequency. Results PTCs with frequency-shifted tips (indicative of dead regions) were found for all subjects. The frequencies at the tips sometimes decreased slightly with increasing signal level. For the subject with a relatively flat loss, PTCs with tips close to 3000 Hz were obtained for signal frequencies of 400, 1000 and 1500 Hz. A PTC with a tip at 5000 Hz was found for a signal frequency of 6000 Hz. These results suggest that this subject had an “island” of surviving IHCs and neurons with CFs ranging from 3000 to 5000 Hz, with extensive dead regions on either side. For the subject with a mid-frequency loss, the pattern of results suggested a mid-frequency dead region. For the subjects with high-frequency loss, the results suggested the presence of high-frequency dead regions, in one case starting at a frequency where absolute thresholds were only slightly higher than normal. Conclusions PTCs can be used to detect and delimit dead regions. Often, the frequency at the tip of the PTC can be used to define approximately one boundary of the dead region. However, the detection of beats can affect the shape of the PTC around the tip when the signal frequency lies just inside the dead region. The level of the signal can also have some effect on the frequency at the tip of the PTC. Very low signal levels can lead to variable results. Dead regions can start at frequencies where absolute thresholds are near normal.

Comparison of different forms of compression using wearable digital hearing aids

Four different compression algorithms were implemented in wearable digital hearing aids: (1) The slow-acting dual-front-end automatic gain control (AGC) system [B. C. J. Moore, B. R. Glasberg, and M. A. Stone, Br. J. Audiol. 25, 171-182 (1991)], combined with appropriate frequency response equalization, with a compression threshold of 63 dB sound pressure level (SPL) and with a compression ratio of 30 (DUAL-HI); (2) The dual-front-end AGC system combined with appropriate frequency response equalization, with a compression threshold of 55 dB SPL and with a compression ratio of 3 (DUAL-LO). This was intended to give some impression of the levels of sounds in the environment; (3) Fast-acting full dynamic range compression in four channels (FULL-4). The compression was designed to minimize envelope distortion due to overshoots and undershoots; (4) A combination of (2) and (3) above, where each applied less compression than when used alone (DUAL-4). Initial fitting was partly based on the concept of giving a flat specific-loudness pattern for a 65-dB SPL speech-shaped noise input, and this was followed by fine tuning using an adaptive procedure with speech stimuli. Eight subjects with moderate to severe cochlear hearing loss were tested in a counter-balanced design. Subjects had at least 2 weeks experience with each system in everyday life before evaluation using the Abbreviated Profile of Hearing Aid Benefit (APHAB) test and measures of speech intelligibility in quiet (AB word lists at 50 and 80 dB SPL) and noise (adoptive sentence lists in speech-shaped noise, or that same noise amplitude modulated with the envelope of speech from a single talker). The APHAB scores did not indicate clear differences between the four systems. Scores for the AB words in quiet were high for all four systems at both 50 and 80 dB SPL. The speech-to-noise ratios required for 50% intelligibility were low (indicating good performance) and similar for all the systems, but there was a slight trend for better performance in modulated noise with the DUAL-4 system than with the other systems. A subsequent trial where three subjects directly compared each of the four systems in their everyday lives indicated a slight preference for the DUAL-LO system. Overall, the results suggest that it is not necessary to compress fast modulations of the input signal.

Development of a fast method for determining psychophysical tuning curves

Psychophysical tuning curves (PTCs) can be used to assess the frequency selectivity of the auditory system and to detect and delimit “dead regions” in the cochlea. However, the traditional method for determining PTCs takes too long for use in clinical practice. We evaluated a fast method for determining PTCs, using a band of noise that sweeps in centre frequency and a Békésy method to adjust the masker level required for threshold. The shapes of the PTCs were similar for the fast and traditional methods, for both normally hearing and hearing-impaired subjects. Rates of change of masker level of 2 dB/s or less gave the most reliable results. A relatively wide bandwidth (20 percent of the signal frequency or 320 Hz, whichever is the smaller) was needed to minimise the influence of beat detection. When the signal frequency fell within a dead region, the fast method gave PTCs with shifted tips.

Comparison of three procedures for initial fitting of compression hearing aids. III. Inexperienced versus experienced users

We assessed whether gain requirements differ for experienced users and new users when fitted with multi-band compression hearing aids. Three procedures for initial fitting were used: the Cambridge method for loudness equalization (CAMEQ), the Cambridge method for loudness restoration (CAMREST), and the desired sensation level input/output (DSL[i/o]) method. Twenty experienced hearing aid users and 20 new users with mild-to-severe sensorineural loss were fitted with Danalogic 163D digital hearing aids, using each procedure in turn in a counter-balanced order. The new users were given a pre-fitting with slightly reduced gains prior to the ‘formal’ fitting. Immediately after formal fitting with a given procedure, and 1 week after fitting, the gains were adjusted by the minimum amount necessary to achieve acceptable fittings. The amount of adjustment required provided the main measure of the adequacy of the initial fitting. On average, new users required decreases in gain for all procedures, the decreases being larger for DSL[i/o] than for CAMEQ or CAMREST. For experienced users, gain adjustments were small for CAMEQ and CAMREST, but were larger and mostly negative for DSL[i/o]. After these gain adjustments, users wore the aids for at least 3 weeks before filling out the Abbreviated Profile of Hearing Aid Benefit (APHAB) questionnaire and taking part in laboratory measurements of the speech reception threshold (SRT) for sentences in quiet and in steady and fluctuating background noise at levels of 60 and 75dBSPL. The scores on the APHAB test and the SRTs did not differ significantly for the three procedures. We conclude that the CAMEQ and CAMREST procedures provide more appropriate initial fittings than DSL[i/o]. For inexperienced users, gains typically need to be reduced by about 3 dB relative to those prescribed by CAMEQ or CAMREST, although the amount of reduction may depend on hearing loss. An analysis of gain adjustments as a function of order of testing provided some evidence for increased tolerance to high-frequency amplification with increasing experience during the 4-month course of the trial, but this effect did not differ for the experienced and new users. Sumario Evaluamos si existen diferencias en los requerimientos de ganancia entre usuarios inexpertos o experimentados, al adaptarles auxiliares auditivos (AA) de compresión multibanda. Se usaron tres procedimientos para adaptación inicial: los métodos Cambridge para ecualización (CAMEQ) y restauración (CAMREST) de la intensidad subjetiva y el de ingreso/egreso para nivel de sensación deseada (DSL[i/o] ). Se adaptaron AA digitales Danalogic 163D a veinte usuarios experimentados y a 20 nuevos, con pérdidas neurosensoriales medias a severas, usando balanceadamente los tres procedimientos. Los nuevos usuarios fueron pre-adaptados con ganancias ligeramente reducidas, antes de la adaptación “formal”. Inmediatamente después de ésta y una semana después, con cada uno de los procedimientos de adaptación, se ajustaron las ganancias con el mínimo necesario para lograr adaptaciones aceptables. El monto del ajuste fue la principal medición para basar la adaptación inicial. En promedio, los nuevos usuarios requirieron disminución de la ganancia con todos los métodos, siendo mayor con DSL[i/o] que con CAMEQ o CAMREST. Para los experimentados, los ajustes de ganancia fueron menores con CAMEQ y CAMREST, pero más amplios y en su mayoría negativos para DSL[i/o]. Después de estos ajustes de ganancia los AA fueron usados no menos de 3 semanas, antes de llenar el Perfil Abreviado de Beneficio de AA APHAB) y de tomar parte en mediciones de laboratorio del umbral de recepción del lenguaje (SRT) con palabras, sin y con ruido de fondo fijo o fluctuante, en niveles de 60 y 75 dB SPL. Las puntuaciones del APHAB y del SRT no variaron significativamente con los tres métodos. Concluimos que el CAMEQ y el CAMREST permiten una adaptación inicial más apropiada que el DSL[i/o]. En usuarios inexpertos, fue típicamente necesario reducir la ganancia en cerca de 3 dB cuando la prescripción se hizo con CAMEQ o CAMREST, aunque el monto de la reducción puede depender de la pérdida auditiva. Un análisis de los ajustes de ganancia como función del orden de las pruebas, mostró evidencias de mayor tolerancia a la amplificación en frecuencias agudas conforme aumentó la experiencia en los cuatro meses del protocolo, pero este efecto no varió entre usuarios experimentados o inexpertos.

Comparison of two adaptive procedures for fitting a multi-channel compression hearing aid

We compared two adaptive procedures for fitting a multi-channel compression hearing aid. “Camadapt” uses judgements of the loudness of speech stimuli and the tonal quality of music stimuli. “Eartuner” uses judgements of the loudness and clarity of speech stimuli with differing spectral characteristics. Sixteen new users of hearing aids were fitted unilaterally, using each procedure. The fittings were assigned to Programs 1 and 2 in the aid, in a counter-balanced order. Subjects kept a diary of their experiences with each program in everyday life. Following 2-4 weeks of experience, they filled in the APHAB and other questionnaires and were re-fitted using both procedures. Camadapt generally led to higher low-level gains and lower high-level gains than Eartuner. Gains recommended by the procedures did not change following experience. Eight subjects preferred the Camadapt fitting and eight preferred the Eartuner fitting. Most subjects gave high overall satisfaction ratings for both procedures. Test-retest reliability was better for Eartuner than for Camadapt. Preference for the Camadapt fitting was associated with slightly better speech communication with Camadapt, while preference for the Eartuner fitting was associated with fewer problems with aversion for that procedure.

The effect on speech intelligibility of varying compression time constants in a digital hearing aid.

The identification of nonsense syllables in quiet and in three types of background (babble, cafeteria and single female speaker) was measured using four hearing aid compression algorithms differing in attack and release time constants, and using linear amplification. The speech level was always 65 dB SPL. The compression algorithms, which were implemented in a Phonak Claro ITE hearing aid, were: (1) ‘very fast’—the attack time was 8 ms and the release time was 32 ms, for all 20 channels; (2) ‘slow–fast’—the attack and release times decreased from 500 ms for low frequencies to about 100 ms for high frequencies; (3) ‘fast–slow’—the attack and release times increased from about 50 ms for low frequencies to 500 ms for high frequencies; and (4) ‘slow+fast’—a very slow-acting gain control signal was combined with a fast-acting gain control signal, for each channel in a 10-channel system. Acoustical stimuli were presented monaurally via a circumaural headphone mounted over the hearing aid. The linear condition did not use the Claro aid; instead, the signal was digitally filtered to implement the Cambridge formula prior to delivery via the earphone. Five subjects with moderate sensorineural hearing loss were tested in a counter-balanced order across conditions. In quiet, performance was best for linear amplification and worst for the slow+fast algorithm. In the presence of background sounds, the highest scores were obtained with the linear-gain Cambridge formula implemented via headphones; a supplementary experiment suggested that this was due to the greater high-frequency gain resulting from the use of this formula. No significant differences were found between scores for the different compression algorithms. We conclude that the intelligibility of speech at a fixed level, presented in background sounds, is not markedly affected by rather substantial variations of the time constants in a multichannel compression system. Sumario Se midió la identificación de sílabas sin sentido realizada en silencio y en tres tipos de ruidos de fondo (balbuceo, cafetería, hablante femenino individual) utilizando cuatro algoritmos de compresión para auxiliar auditivo, con diferencias en las constantes de los tiempos de ataque y liberación, y usando amplificación lineal. El nivel del lenguaje fue siempre de 65 dB SPL. Los algoritmos de compresión, que fueron implementados en un auxiliar auditivo intra-auricular Phonak Claro fueron: (1) “muy rápido”—el tiempo de ataque fue de 8 ms y el tiempo de liberación de 32 ms, para los 20 canales; (2) “lento-rápido”—los tiempos de ataque y liberación disminuyeron de 500 milisegundos para las frecuencias graves hasta alrededor de 100 ms para las agudas; (3) “rápido-lento”—Los tiempos de ataque y liberación se incrementaron alrededor de 50 ms para las frecuencias graves hasta 500 ms para las agudas: y (4) “lenta+rápido—se combinó una señal de control de ganancia de actuación muy lenta con una señal de control de ganancia de actuación muy rápida, para cada canal, en un sistema de 10 canales. Se presentaron estímulos acústicos monoauralmente por medio de auriculares circumaurales montados sobre el auxiliar auditivo. La condición lineal no utilizó el audífono Claro; en su lugar, la señal fue filtrada digitalmente para implementar la fórmula Cambridge antes de presentarla a través del auricular. Cinco sujetos con una hipoacusia sensorineural moderada se evaluaron en un orden contra-balanceado en todas las condiciones. En silencio, el rendimiento fue mejor con la amplificación lineal y fue peor con el algoritmo “lento+rápido”. En presencia de sonidos de fondo, los puntajes más altos fueron obtenidos con la fórmula de Cambridge de ganancia lineal, implementada a través de auriculares; un experimento suplementario sugirió que esto se debía a la mayor ganancia en frecuencias agudas producto del uso de esta fórmula. No se encontraron diferencias significativas entre los puntajes para los diferentes algoritmos de compresión. Concluímos que la inteligibilidad para el lenguaje a un nivel fijo, presentado ante ruido de fondo, no se afecta marcadamente por variaciones sustanciales de las constantes de tiempo en un sistema de compresión multicanal.

Auditory temporal-envelope processing in high-functioning children with Autism Spectrum Disorder.

Individuals with Autism Spectrum Disorder (ASD) perform worse than controls when listening to speech in a temporally modulated noise (Alcántara, Weisblatt, Moore, & Bolton, 2004; Groen et al., 2009). The current study examined whether this is due to poor auditory temporal-envelope processing. Temporal modulation transfer functions were measured in 6 high-functioning children with ASD and 6 control listeners, using sinusoidal amplitude modulation of a broadband noise. Modulation-depth thresholds at low modulation rates were significantly higher for the ASD group than for the Control group, and generally higher at all modulation rates tested. Low-pass filter model estimates of temporal-envelope resolution and temporal-processing efficiency showed significant differences between the groups for modulation-depth threshold values at low modulation rates. Intensity increment-detection thresholds, measured on a subset of individuals in the ASD and Control groups, were not significantly different. The results are consistent with ASD individuals having reduced processing efficiency of temporal modulations. Possible neural mechanisms that might underlie these findings are discussed.

International technical standards: Whose problem is it? Response to M. C. Martin

Dear Sir, Mike Martin’s first criticism is that we used the ‘wrong’ number when referring to the IEC standard method that was used for testing the aids. It is our understanding that a standard does not cease to exist if it is subsequently renumbered. In referring to IEC 118, 1983, we were merely using the same reference number as specified in the reference guide for the hearing aid ‘test box’ that we employed (a Rastronics PortaRem 2000). The instruction manuals of several other test boxes also refer to the ‘old’ standard numbers, rather than the recently changed ones. Most readers will recognize the ‘old’ numbers and will be unfamiliar with the ‘new’ numbers. In any case, what is important is that the reference referred to is traceable, and IEC 11 8, 1983 certainly is traceable. Mike Martin’s second point is that the standard method for measuring equivalent input noise is ’a simple method and only allows for measurement on a linear portion of the input-output characteristic’. This is quite true. Yet, most manufacturers apply the standard method to compression aids, often using a ‘conditioning’ tone with a level above the compression threshold. This gives a misleading measure of the equivalent input noise, leading to a noise value higher than the ‘true’ value. In our paper, we emphasize the need to use a conditioning tone with a level below the compression threshold. Mike Martin criticizes our statement that ‘Many of the traditional methods of measuring the electroacoustic characteristics of hearing aids do not work properly, or give misleading results, when used with modern aids, especially digital aids.’ We stand by this statement, and believe that it is correct. It was intended as a general statement, not aimed specifically at the methods specified in international standards. For example, measurements of frequency response and gain obtained with sinusoidal signals can sometimes be completely unrepresentative of frequency response and gain for speech or music signals. We entirely agree that users of the methods specified in the international standards should be aware of the limitations of the standards. That is one of the important conclusions to be drawn from our paper. Mike Martin points out that the object of IEC 601 18-0 includes the statement: ‘The methods are chosen first of all to be practical and reproducible, and consequently they are based on fixed parameters chosen, to a certain extent, arbitrarily. This should be taken into consideration when comparisons are being made between test results for hearing aids of different models and manufacture.’ It was an awareness of this that led us to conduct the ‘extended measurements’ described in our paper, which do not conform to any standard, but which may nevertheless give a better indication of the relative performance of these technically diverse hearing aids in ‘real-life’ situations. We do not wish to enter into the debate as to whether 2-cc couplers used in test boxes meet the requirements of the IEC standards. A different coupler (Briiel and Kjller 41 53) was used for ‘extended’ measurements in our paper, because it provides a more accurate simulation of the characteristics of the human ear, and because its output was more readily available than that of the 2-cc coupler built into the test box. We are pleased to hear that the IEC standard is currently being revised to include correction curves that will allow simulated insertion gains to be established without the use of a head and torso simulator. No doubt the standard will also emphasize that there may be considerable errors associated with this method. Such curves are no substitute for individual (probe tube) measures of real-ear aided gain. It was not our intention to ‘complain’ about the deficiencies of the standards relating to audiology. We wanted rather to point out that limitations do exist, and also to suggest novel measurement methods that would have a closer relationship to the ‘true’ performance of the aids being tested. It is indeed unfortunate that in the UK there is little input to the work of standards committees from academic institutions. University researchers today come under increasing pressure to demonstrate their ‘effectiveness’ by publishing their work in refereed journals and by obtaining research grants to support their work. In the eyes of the various evaluation bodies (Teaching Quality Assessment, Research Assessment Exercise), and grant-giving bodies, work on standards committees does not ‘count’. While this situation continues, there is little prospect that university researchers in the UK will have the time or the resources to become members of standards committees. However, we believe that university researchers can contribute effectively to the development of novel methods for characterizing the performance of hearing aids, and we hope that our paper represents a step in this direction.