Otitis media is a frequently occuring disease of the middle ear with an increasing prevalence. At the age of one, at least 60% have already suffered from acute otits media . In most cases, origin is a dysfunction of the eustachian tube, resulting in an accumulation of fluid in the tympanic cavity. In case of clear, serous fluid in the tympanic cavity, no antibiotic therapy is required. The progression of the disease leads to a mucous and purulent liquid filling of the tympanic cavity, requiring antibiotics for preventing severe complications and residual hearing loss. So far, the composition of the fluid in the tympanic cavity is difficult to determine non-invasively using the available diagnostics otoscopy and tympanometry. Thus, the selection of the appropriate therapy depends on the expertise of the physician, especially in case of inflammatory alterations.
In recent studies, optical coherence tomography (OCT) has been suggested as an innovative non-invasive diagnostic tool for the inspection of the middle ear. OCT is an interferometric imaging technique employing near infrared light for the depth-resolved visualization of superficial tissue with a spatial resolution of a few micrometers , .
OCT can be used for the three-dimensional visualization of the tympanic membrane and has been proposed, e. g., for the detection of scattering fluid located in the tympanic cavity ,  and for the detection of bacterial biofilms behind the tympanic membrane . In addition to the structural information, Doppler-OCT allows the measurement of the tympanic membrane oscillation , .
This ex vivo study investigates the capability of structural OCT imaging and Doppler-OCT imaging for the detection and differentiation of different fluid types with various filling levels in the human tympanic cavity.
2 Material and methods
2.1 Specimen preparation
For this ex vivo study, a freshly excised human temporal bone was used. Pathologic alterations were excluded via otoscopic examination by an ENT physician. The external ear channel was removed to provide access to the entire tympanic membrane. The sample was stored in a 0.9% saline solution before examination. For OCT imaging, the specimen was mounted using an articulated arm ensuring the orientation to be similar to a seating position of the patient.
Otitis media with effusion was simulated by filling the tympanic membrane with different fluids through the Eustachian tube. Therefore, a tube was inserted to the osseous part of the Eustachian tube. During the measurement, the filling level was kept constant by clamping this tube. The filling level was adjusted by means of otoscopy, ear microscopy and structural OCT imaging, if applicable. A serous and transparent fluid filling was simulated using 0.9% saline solution. The highly scattering purulent liquid filling was simulated using SMOFlipid (Fresenius Kabi Deutschland GmbH).
2.2 Optical coherence tomography
The swept source OCT-system utilizes a self-built laser source, which is a Fourier domain mode locked (FDML) laser  providing a line scan rate (A-scan rate) of 60 kHz. The acquisition time for volume scans and oscillation measurements is approximately 5.5 s. The center wavelength is 1300 nm and the bandwidth is typically 120 nm and can be modified for the adjustment of the maximum imaging depth. The emitted light is propagating to a scanner head containing a modified Michelson interferometer and two galvanometric scanners for three-dimensional image acquisition by means of telecentric scanning. The back scattered light propagates to a balanced detector, where the interference spectrum is measured in time multiplex. The A-scan is calculated thereof mainly by applying a Fourier transform. The axial and lateral resolution of the OCT-system is 14 μm and 11.5 μm, respectively. Details on the OCT system have been published in .
2.3 Measurement principle
For examining the middle ear, a three-dimensional image stack composed of 512 × 512 A-scans and a functional Doppler-OCT measurement were carried out for each simulated condition. The field of view was set to be 10 mm × 10 mm covering the entire tympanic membrane. For visualizing the position and orientation of the tympanic membrane, a depth-projection was calculated over the three-dimensional stack.
The oscillation of the tympanic membrane was acoustically stimulated using a loudspeaker. Simultaneously to the OCT-image acquisition, the applied sound pressure was measured using a probe tube microphone (ER-7C, Etymotic Research) positioned next to the tympanic membrane.
The membrane oscillation was measured on a 25 × 25 grid using Doppler-OCT. At each grid point, the tympanic membrane was stimulated using an acoustic chirp signal covering the frequency range from 500 Hz to 5 kHz, which is relevant for speech perception. The transfer functions were calculated via Fourier transform and via normalization with respect to the applied sound pressure. Thus, the transfer function provides the oscillation amplitude normalized to the sound pressure and resolved in frequency. Interpolated oscillation maps can be calculated in post-processing for frequencies within the excited frequency range. Details about the signal processing have been published in . For grid points being affected by measurement errors due to limited reflectivity, e. g. at the border of the field of view or near to the umbo, the oscillation amplitude was manually set to zero.
3 Results and discussion
The results of the OCT measurements are summarized in Figure 1. Structural OCT imaging allowed the three-dimensional visualization of the tympanic membrane, which had a thickness of about 140 μm in the pars tensa area. The middle column in Figure 1 shows depth projections of three-dimensional OCT-scans illustrating the tympanic membrane orientation. Exemplary cross-sections corresponding to the red marked positions are shown in the right column. In these cross-sections, the funnel-like shape of the tympanic membrane is visible.
Liquid filling in the tympanic cavity is only directly visible by means of OCT if the fluid contains scattering particles. Thus, the SMOFlipid emulsion can be clearly identified as a substance with high backscattering intensity directly behind the tympanic membrane as well as areas with intensity enhancement in the depth projections. In case of 0.3 ml SMOFlipid, the fluid is located mainly behind the inferior quadrants of the tympanic membrane and the filling level is approximately reaching the umbo, the deepest point of the tympanic membrane. In case of the transparent NaCl-solution, the fluid itself cannot be visualized using structural OCT imaging, since it doesn’t contain any scattering particles. Only at some places the fluid-air interface causes additional reflexes.
In addition, Doppler OCT was used for the measurement of the tympanic membrane oscillation. In Figure 1, the normalized oscillation amplitude is shown in the left column for an oscillation frequency of 1168 Hz. Without pathological alternations, this frequency corresponds almost to an in-phase oscillation of the entire pars tensa. The normalized oscillation amplitude of the pars tensa is approximately 0.2 μm/Pa and has a maximum of 0.4 μm/Pa in the posterosuperior quadrant. At the manubrium of malleus, the oscillation amplitude is reduced.
An effusion in the tympanic cavity results in a decreased oscillation amplitude. Since the oscillation is widely damped at those areas of the tympanic membrane being in contact with the fluid, the estimation of the filling level is possible. This effect on the oscillation amplitude can be measured in case of both, scattering fluids (SMOFlipid) and transparent fluids (NaCl). In case of 0.3 ml SMOFlipid filling, the area of reduced oscillation amplitude is almost congruent to the area of liquid filling being visible in the corresponding depth-projection. The oscillation was examined at 1168 Hz, because nodal lines of the oscillation are not expected at this frequency. This facilitates the straightforward comparison and interpretation of the amplitude maps. In case of fluid filling, the measured oscillation amplitude was in the order of the detection limit. Thus, the oscillation amplitude was reduced at least to approximately 5%.
The extension to Doppler-OCT imaging permits the indirect detection of an effusion in the tympanic cavity for transparent and scattering fluids. In comparison, structural OCT imaging would not be able to visualize transparent effusions, which are free of scattering particles. Thus, Doppler OCT can improve the classification of effusions in the tympanic cavity. Additionally, the indirect detection of fluids could be necessary in case of tympanic membranes being less translucent due to inflammatory or degenerative alterations. These advantages might be essential using Doppler-OCT as a reliable tool in middle ear diagnostics.
This proof-of-principle study has investigated the ability of OCT to detect an effusion in the tympanic cavity. Standard OCT imaging measures the back scattering intensity and can be used for the visualization of scattering fluids. Structural OCT imaging alone is not suited for the detection of transparent fluid, where scattering particles are missing in the fluid.
The extension to Doppler-OCT allows the measurement of the tympanic membrane oscillation. The oscillation amplitude is reduced in areas of the tympanic membrane being in contact with fluid in the tympanic cavity. Thus, Doppler-OCT is suited for the indirect detection of scattering fluids and transparent fluids.
Research funding: The author state no funding involved. Conflict of interest: Authors state no conflict of interest. Material and Methods: Informed consent: Informed consent has been obtained from all individuals included in this study. Ethical approval: The research related to human use complies with all the relevant national regulations, institutional policies and was performed in accordance with the tenets of the Helsinki Declaration, and has been approved by the authors’ institutional review board or equivalent committee.
Leibovitz E, Greenberg D. Acute otitis media in children: current epidemiology, microbiology, clinical manifestations, and treatment. Chang Gung Medical Journal J. 2004;27:475–88.
Huang D, Swanson EA, Lin CP, Schuman JS, Stinson WG, Chang W, et al. Optical coherence tomography. Science 1991;254:1178–81.
Walther J, Gaertner M, Cimalla P, Burkhardt A, Kirsten L, Meissner S, et al. Optical coherence tomography in biomedical research. Analytical and Bioanalytical Chemistry Chem. 2011;400:2721–43.
Burkhardt A, Walther J, Cimalla P, Bornitz M, Koch E. Endoscopic optical coherence tomography for imaging the tympanic membrane. Proc. SPIE 2011;8091:80910Y.
Monroy GL, Pande P, Shelton RL, Nolan RM, Spillman DR, Porter RG, et al. Non-invasive optical assessment of viscosity of middle ear effusions in otitis media. J Biophotonics. 2016; DOI: .
Nguyen CT, Robinson SR, Jung W, Novak MA, Boppart SA, Allen JB. Investigation of bacterial biofilm in the human middle ear using optical coherence tomography and acoustic measurements. Hear Res. 2013;301:193–200.
Burkhardt A, Kirsten L, Bornitz M, Zahnert T, Koch E. Investigation of the human tympanic membrane oscillation ex vivo by Doppler optical coherence tomography. J Biophotonics. 2014;7:434–41.
Subhash HM, Nguyen-Huynh A, Wang RK, Jacques SL, Choudhury N, Nuttall AL. Feasibility of spectral-domain phase-sensitive optical coherence tomography for middle ear vibrometry. J Biomed Opt. 2012;17:060505.
Huber R, Wojtkowski M, Fujimoto JG. Fourier Domain Mode Locking (FDML): a new laser operating regime and applications for optical coherence tomography. Opt Express. 2006;14:3225–37.