Medical Applications

Acoustic Tube Endoscopy

Sonotubometry allows for the detection of the dynamic Eustachian tube function under physiological conditions, whether the eardrum is intact or perforated. In cooperation with the ENT Clinic of the UK Aachen and the ENT Clinic of the DIAKO Bremen an innovative concept for sonotubometric assessment of the Eustachian tube function ‐ in particular its dynamic behavior ‐ has been developed [Antweiler 2006a], [Antweiler 2006b], [Telle 2012]. Quality and reliability of state‐of‐the‐art sonotubometry have been substantially improved by the new real‐time system. With digital signal processing algorithms, the generation of a new feature relevant for medical diagnostics is made accessible. The method is based on techniques known from speech processing: the acoustic tube model, the normalized least‐mean‐square (NLMS) algorithm with so‐called perfect sequence (PSEQ) excitation and the Levinson Durbin algorithm.

In the real‐time acoustic measurement prototype the Eustachian tube is treated as a linear transmission system (Fig. 1).

We apply a PSEQ in the nasal cavity and record simultaneously the reaction of the nose/ear system using a microphone located in the ear (Fig. 2). Its impulse response is obtained by a subsequent NLMS‐type system identification.

With digital signal processing algorithms we extract two different features according to Fig. 1:

  • The fluctuations of the sound level intensity in the outer ear indicate activity of the Eustachian tube provoked by, e.g., yawning or swallowing. They are mapped with the quadratic norm of the impulse response.
  • Based on techniques known from speech processing such as the acoustic tube model and the Levinson‐Durbin algorithm, a novel virtual model of the nose/ear transmission link can be built. The dynamic opening and closing process of the tube is visualized by an animation of the virtual tube model over time. By means of this model, a virtual acoustic tube endoscopy can be generated.

An example of the results is shown in Fig. 3 where three openings of the Eustachian tubes were provoked by yawning of the person under test. The quadratic norm of the measured impulse responses clearly indicates not only the principal opening and closing of the tube but provides even information about the dynamic process itself. Thus, rather than providing only one single amplitude value as does the conventional sonotubometry applying a sinus signal, here a complete impulse response is obtained. As a result, additional information is obtained at the same measurement expenditure. This gain of information is utilized on one hand to map the dynamic behavior of the tube function in much better quality by means of the quadratic norm of the impulse response. Due to the fact that the norm of the impulse response exploits the information of all frequency components up to 16 kHz, a significant improvement in the visualization of the Eustachian tube activity can be achieved.

On the other hand the information gain can be exploited by the extraction of a completely new feature which takes the form of the frequency response into account. Experiments with the Eustachian tube of test persons reflect a reasonable correlation between the virtual model and the expected anatomy of the nose/ear system. With an animation of the virtual tube model over time the dynamic open and closing process of the Eustachian tube can be visualized.

It is of special interest that the norm and the virtual tube model are based on orthogonal information such that the features relevant for diagnostics complement each other.

As a result, the acoustic measurement system allows real‐time monitoring of the Eustachian tube activity at the perforated, as well as at the intact eardrum in physiological conditions. New insights into the dynamics of the Eustachian tube function are currently investigated within numerous studies. Some examples can be found in the References on this page.

Future work will aim at a dual‐channel measurement prototype to visualize the transmission links between the two nostrils and the ear as a function of time.

References

[antweiler06b]
Christiane Antweiler, Aulis Telle, Peter Vary, and Ercole Di Martino
A New Otological Diagnostic System Providing a Virtual Tube Model
Proceedings of Biomedical Circuits and Systems Conference (BIOCAS), November 2006

[antweiler06a]
Christiane Antweiler, Peter Vary, and Ercole Di Martino
Virtual Time-Variant Model of the Eustachian Tube
Proceedings of IEEE International Symposium on Circuits and Systems (ISCAS), May 2006

[asenov2010]
Dejan Asenov, Viorel Emanoil Nath, Aulis Telle, Christiane Antweiler, Leif E. Walther, Peter Vary, and Ercole Di Martino
Sonotubometry with Perfect Sequences: First Results in Pathological Ears
Acta Oto-Laryngologica, November 2010

[dimartino09a]
Ercole Di Martino, Viorel Emanoil Nath, Aulis Telle, Christiane Antweiler, Leif E. Walther, and Peter Vary
Evaluation of Eustachian Tube function with Perfect Sequences: Technical Realization and First Clinical Results
European Archives of Oto-Rhino-Laryngology, March 2010

[dimartino08]
Ercole Di Martino, Viorel Emanoil Nath, Aulis Telle, Leif E. Walther, Martin Westhofen, Christiane Antweiler, and Peter Vary
Akustische Tubenfunktionsuntersuchung mit Perfekten Sequenzen - Eine neue Methode zur Beurteilung der Eustachischen Röhre
Laryngo-Rhino-Otologie, June 2008

[dimartino08c]
Ercole Di Martino, Viorel Emanoil Nath, Dejan Asenov, Christiane Antweiler, Leif E. Walther, and Peter Vary
Evaluation of eustachian tube function in sonotubometry with perfect sequences in children
8th Conference of European Society of Pediatric ORL (ESPO), June 2008

[dimartino06a]
Ercole Di Martino, Christiane Antweiler, Aulis Telle, Peter Vary, and Jens Beckschebe
Assessment of Eustachian Tube with Perfect Sequences
Otolaryngology - Head and Neck Surgery, August 2006

[telle12]
Aulis Telle
Sonotubometrie mit Methoden der digitalen Signalverarbeitung
Dissertation, January 2012