Sonification of Complex Biomedical Data
Authors: Gerold Baier, Thomas Hermann, Ulrich Stephani
Journal Biol. Phys., submitted
We introduce data sonification as a new means to efficiently deal with biomedical data of high complexity. The approach is motivated by a reflection on the capabilities of the human auditory system to extract relevant information from acoustic clues. Complex physiologic rhythms appear to be particularly suited for an auditory representation. We describe event-based sonification of rhythms as a specific technique for the analysis of temporal relationships, particularly for multivariate data. We present example sonifications from human EEG and discuss normal and pathologic rhythms. For improved performance we suggest the use of virtual auditory environments for sonifications of biomedical data.
- Example S1: S1 (mp3, 396k)
Real-time EBS of an EEG segment with normal background activity in a subject at rest with eyes open. Recordings produced with 19 electrodes in a 10-20 system. Events are restricted to inter-maxima differences between 1 and 1/18 seconds (corresponding to the range between 1 and 18 Hz). Duration of sound: 25 s.
- Example S2: S2 (mp3, 492k)
EBS, as in Sound 1, of an EEG segment with epileptic activity in a patient with primarily generalized absence seizures. The recording corresponds to the time series displayed in Fig. 2 (top). Duration of sound: 30 s.
- Example S3: S3 (mp3, 472k)
EBS of the segment of EEG used in Sound 2, but with fixed minimum threshold for the amplitudes and with bell-shaped filtering (center frequency 3 Hz, bandwidth 2 Hz) of the sounds. Duration of sound: 30 s.
- Example S4: S4 (mp3, 544k)
EBS of a segment of EEG recorded during sleep stage 3 with delta wave activity in all electrodes. 27 electrodes in 10-10 setting, bipolar montage. Filter settings for the sonification: center frequency 2 Hz, bandwidth 2 Hz (c.f. caption of Sound 3). Duration of sound: 35 s.
- Example S5: S5 (mp3, 344k)
EBS of a segment of EEG recorded during sleep stage 2 with sleep spindle activity in some of the electrodes. Filter settings: center frequency 14 Hz, bandwidth 3 Hz. Duration of sound: 22 s.
modified 2007-11-14Thomas Hermann