A cochlear implant is applied in cases of profound and total deafness, when a patient has no benefit from conventional hearing aids. The cochlear implant provides direct electrical stimulation of the auditory nerve. Thus it is possible to induce hearing sensations by-passing the damaged inner ear and enable perception of environmental sounds using the remaining efficient part of the auditory pathway. In order to realise this task, the implant system must convert acoustic information into a suitable electrical signal. The conversion is performed within an integral sub-unit of the system, the speech processor. The results that the patient can achieve in the post-operative rehabilitation period depend to a great extent on appropriate fitting of the processor, and its parameters must be carefully adapted to the individual needs of a patient.

The speech processor provides continuous electrical stimulation of the auditory nerve. The elicited hearing sensations should be similar to those which would result from the perceived original acoustic signal. In modern cochlear implant systems, several techniques are employed to meet this objective. Generally, the hardware of a processor can be of two different types, consisting of analogue or digital circuits.

As far as the signal processing method is concerned, there are several different algorithms, called speech coding strategies, that can be used.

Fitting of a speech processor includes selection of an appropriate coding strategy, and loading specific parameters determining the form of the output signal most suitable for the particular patient. In digital processors, the parameters include stimulus amplitudes referring to hearing sensation threshold THR_µA¹ and comfort hearing level MCL_µA, stimulus pulse duration, the compression coefficient of acoustic-to-electrical signal conversion, stimulation polarity mode (bipolar or monopolar), etc.

In analogue systems, the transmitting circuit´s frequency characteristics are varied.

Parameters such as THR_µA and MCL_µA, are determined based on psychoacoustic tests. Accurate measurement of the parameter values is very important, because this determines the range of the stimulating current and the dynamics of the auditory nerve reaction to electrical stimulation (Lorens 1998). The auditory nerve reaction dynamic is typically not much more than 10dB, while the speech signal is over 30dB. Therefore, an appropriate signal compression must be applied along with signal processing. Relatively small errors in determining the values of parameters THR_µA and MCL_µA may lead to a situation in which the patient will not perceive hearing sensations on an appropriate level, assessing the subjective hearing level as too high or too low, or will not able to distinguish changes in acoustic signal magnitude. The above parameters determine the compression characteristic of the speech processor. Therefore, precise measurement of THR_µA and MCL_µA parameters is critical for speech understanding (Lehnhardt, 1992).

In determining the principles of speech processor fitting for a child rehabilitation programme, the authors have analysed data collected from 63 children implanted since 1992. The group consists of both single-channel analogue implant users (Med-El „Comfort” type), and multi-channel digital system users (Med-El Combi 40 and Combi 40+, Cochlear Nucleus MSP and Nucleus Spectra, Philips Hearing Implants types Laura III and Laura Flex).

The measurement technique applied in psychoacoustic tests for speech processor fitting must be compatible with the hearing and language competence of the child, related to his/her age. The consecutive steps, and the requirements presented, constitute the speech processor fitting procedure developed at the Institute of Physiology and Pathology of Hearing in Warsaw. The following are the stages of the procedure;

Stage I concerns training prior to the first speech processor fitting, including conditioning the child to respond to a stimulus, and teaching the child to understand the concepts „not enough”, and „too much”. During this training, a simple visual signal like a light or vibrations can be used as a stimulus (Beiter 1991). One can also apply an electrostimulation test signal as the conditioning stimulus, preferably using extra-tympanic electrodes (Lorens, 1997). To avoid discomfort to the child due to overstimulation during the first fittings, it is crucial for the child to be familiar with the concepts of „not enough” and „too much”. We have developed several techniques to train the child how to react to a stimulus producing a sensation which is too strong for them.

Stage II includes conditional training with the use of an implant. It is not uncommon for deaf children to react spontaneously and stressfully to an electrical stimulus. They usually have no hearing experience, and therefore even soft sounds due to electrostimulation can produce fear. When raising the level of stimulation, the child´s behaviour should be observed very carefully, to specify the level of electric stimulus causing any change in the child´s behaviour such as touching of the ear or the head, sudden quietness, or turning the head searching for the source of stimulus. A conditioning exercise should be performed for a determined level of stimulus. The stimulation must be terminated immediately if any sign of discomfort or distress appears, or if any changes in the child´s reactions are noticed.

Stage III refers to measurement of THR. When the child is conditioned, the threshold level of electrical stimulation should be found. Children aged from 3 to 5 usually start to react to a stimulus level, which is above the threshold. This is why the authors have decided to download 50% of the measured value of THR as a parameter for the processor.

Stage IV consists of measurements of the MCL parameters. It is very difficult for prelingually deafened children to decide which level of stimulus is most comfortable for them. Therefore the authors recommend finding a level which is closest to the uncomfortable level. Seventy-five percent of the measured value is downloaded as an MCL parameter.

Each session of fitting of the processor should be approached on an individual basis. Duration and frequency of the fitting sessions, as well as the methods of collaboration with the child should be adapted to the age level of the child´s development and speech competence. It is necessary to organise a multidisciplinary team of engineers, speech therapists, psychologists and teachers to be involved in speech processor fitting, as one of the most important elements in the long process of rehabilitation after cochlear implantation.