9th Speech in Noise Workshop, 5-6 January 2017, Oldenburg

Common Audiological Functional Parameters (CAFPAs): an abstract representation of audiological expert data

Mareike Buhl(a)
Medizinische Physik and Cluster of Excellence Hearing4all, CvO Universität Oldenburg, Germany

Marc René Schädler(b)
Medical Physics and Medizinische Physik and Cluster of Excellence Hearing4all, CvO Universität Oldenburg, Germany

Anna Warzybok(b), Birger Kollmeier
Medical Physics and Medizinische Physik and Cluster of Excellence Hearing4all, CvO Universität Oldenburg, Germany

(a) Presenting
(b) Attending

Correct diagnosis of hearing impairment and indication for individual treatment is a complex task influenced by many factors like the choice of audiological measurements, the most frequent cases at a clinic's location or the experience of the ENT doctor. The goal of the Common Audiological
Functional Parameters (CAFPAs) approach is to build up a supporting tool uniting expert knowledge and providing it to the community of ENT doctors.

The CAFPAs are designed as an abstract representation of audiological measurements, thereby generalizing over different measurements and representing functional principles of the auditory system. First data for CAFPAs and measurement results, for different, predefined diagnostical cases and indications for treatment, was collected by the use of a survey among experts from Hanover and Oldenburg. With this data, it is possible to distinguish different diagnostical cases on the basis of audiological measurements as well as on the basis of the newly-introduced CAFPAs. Thus, the CAFPAs serve as an abstract and compact, but interpretable representation of the measurements that do not take into account every detail. Furthermore, the survey results show that the degree of agreement of the experts on the measurement and CAFPA outcomes differed for different diagnostical cases. This property supports the idea of using statistical methods, e.g. machine learning methods, for modeling the audiological diagnostical path.

Last modified 2017-01-04 23:51:47