The human middle ear in motion: 3D visualization and quantification using dynamic synchrotron-based X-ray imaging

Abstract

The characterization of the vibrations of the middle ear ossicles during sound transmission is a focal point in clinical research. However, the small size of the structures, their micrometer-scale movement, and the deep-seated position of the middle ear within the temporal bone make these types of measurements extremely challenging. In this work, dynamic synchrotron-based X-ray phase-contrast microtomography is used on acoustically stimulated intact human ears, allowing for the three-dimensional visualization of entire human eardrums and ossicular chains in motion. A post-gating algorithm is used to temporally resolve the fast micromotions at 128 Hz, coupled with a high-throughput pipeline to process the large tomographic datasets. Seven ex-vivo fresh-frozen human temporal bones in healthy conditions are studied, and the rigid body motions of the ossicles are quantitatively delineated. Clinically relevant regions of the ossicular chain are tracked in 3D, and the amplitudes of their displacement are computed for two acoustic stimuli.

Fig. 1. 2D visualization of the tympanic membrane (TM) in a fresh-frozen human specimen (Fresh1) stimulated at 127 Hz and 2 different sound pressure levels (SPLs).

a A 2D slice is taken from the 3D volume reconstructed at phase p₀ of the movement. b The intensity profile along the yellow line is displayed over time (over the 10 phases of movement), for the 2 stimulation levels, 125 dB SPL and 140 dB SPL. c The intensity profiles along the dashed lines in b (at two specific times t₁ and t₂ corresponding to the extreme positions of the TM) are plotted in the corresponding colors.

Fig. 2. 2D visualization of the ossicular motions in a fresh-frozen human specimen (Fresh1) stimulated at 127 Hz and 2 different sound pressure levels (SPLs).

a The 3D volume containing the ossicular chain reconstructed at phase p₀ is projected in 2D using a maximum intensity projection. b The 3D volume shown in a is vertically resliced along the 3 dashed lines shown in a, along the malleus (green), the incus (orange) and the stapes and umbo (blue). c The reslice shown in b is done in each 3D volume for all phases of movement. The ten resliced images of the same region are then projected in 2D with a standard deviation projection, that highlights the moving features. The yellow arrows point to the location of the temporal bone (static during the sound stimulation).

Fig. 3. Extraction of the ossicular motions in a fresh-frozen human specimen (B-Fresh3) stimulated at 128 Hz and 120 dB sound pressure level.

a Maximum intensity projection and b 3D visualization with segmentation of the static tomographic reconstruction at phase p₀ of the movement, showing the tympanic membrane and the ossicular chain. The tympanic membrane, malleus, incus, and stapes are shown in yellow, green, red, and blue, respectively. c Rotation angles and d magnitudes of the translation vectors of the malleus as a function of time (one blue square per phase of movement). The raw data is shown in blue, with stars for individual data points, squares for the mean values and error bars for the standard deviations of the distribution across the n = 7 independent sub-volumes. e The rotation axes of the three ossicles are displayed for each phase of movement in rainbow colors. All the rotation axes are plotted from the same origin and the magnitude is scaled according to the rotation angle and normalized.

Fig. 4. 3D visualization of the malleus and incus motions in a fresh-frozen human specimen (B-Fresh3) stimulated at 128 Hz and 120 dB sound pressure level.

a Cranial and b lateral view for the malleus, and c medial and d lateral view for the incus positions at 4 different phases of the movement: p₀, p₂, p₅, and p₈, shown in gray, blue, green, and yellow, respectively. The vibration amplitude is amplified x100 for visualization purpose.

Fig. 5. 3D visualization of the stapes motions in a fresh-frozen human specimen (B-Fresh3) stimulated at 128 Hz and 120 dB sound pressure level.

a Cranial, b lateral, c vestibular, and d posterior view of the stapes positions at 4 different phases of the movement: p₀, p₂, p₅, and p₈, displayed in gray, blue, green and yellow, respectively. The vibration amplitude is amplified x100 for visualization purpose.

Fig. 6. Displacement of different regions of interest within the ossicular chain in a fresh-frozen human specimen (B-Fresh3) stimulated at 128 Hz and 120 dB sound pressure level.

a 3D visualization of the segmented ossicular chain at phase p₀ of the movement. The six selected ROIs are marked with a golden dot and labeled: the stapes footplate, the stapes head, the lenticular process, the umbo and two points belonging to the incus and the malleus, respectively, near their common joint. The colored lines correspond to the principal axes of rotation of the malleus (green), incus (red), and stapes (blue). b–g Visualization of the 3D displacement of the ROIs, a the umbo, b the lenticular process, c the stapes footplate, d a point in the malleus near the incudo-malleolar joint, e a point in the incus near the incudo-malleolar joint and f the stapes head. Each color of the 3D rainbow plots corresponds to a phase of movement (from phase p₀ - used as the origin (0,0,0) - to phase p₉), and the 2D projections of the displacements are shown on the XY (blue), YZ (red), and ZX (green) planes.

Fig. 7. Amplitude of displacement of the regions of interest (ROIs) within the ossicular chain.

a–f The magnitudes of the displacement vectors are shown over time for a fresh-frozen human specimen (B-Fresh3) stimulated at a–c 110 dB SPL and at d–f 120 dB SPL, for a, d the umbo, b, e the lenticular process and c, f the stapes footplate. The raw data (dashed blue line) are fitted with a sine (solid green line). The distribution across the n = (11, 5, 6) independently selected points (respectively for the umbo, lenticular process and footplate) is shown: blue squares for mean values and pink crosses for individual data points. g The amplitudes of displacement of B-Fresh3 stimulated at 120 dB SPL are plotted as a function of the ROIs. Each individual point correspond to a manual selection of a point within the ROI. h The amplitudes of displacement averaged over the 6 samples are shown as a function of the ROIs, for the two different stimulation sound pressure levels, 110 dB SPL (orange) and 120 dB SPL (brown). Individual data points are shown as stars, mean values as squares and the error bars represent the standard deviations of the distributions across the n = 6 independent samples.

Fig. 8. Image acquisition and post-gating.

a Experimental setup implemented at the TOMCAT beamline. The sample in its holder a1 is placed on the sample manipulator a2 and illuminated with the X-ray beam coming from the left (yellow arrow). A shutter and slits system a3 protect the sample between acquisitions and adjust the beam size to the corresponding field of view of the low-resolution (LR) setup or the high-resolution (HR) setup. The microscope tower of the HR setup can be translated to let the X-ray beam reach the LR setup. The sound wave is produced by a signal generator a4 connected to a subwoofer a5. The sinusoidal signal generated is systematically recorded, as shown in c for a sound stimulation at 120 dB and 128 Hz. b Zoom-in on the sample holder: the black cylinder is used to plug the silicon tube carrying sound stimulation, and the red-headed tube is used to plug a probe microphone for sound calibration. The full peripheral auditory system of the sample is intact. c The post-gating processing retrospectively associates each image with its corresponding phase of motion, among the 10 phases used to decompose one period of the recorded signal (gray vertical lines, each period starting at the pink line). As examples, the yellow, blue, green, and red points show the images post-gated into the motion phases p₁, p₃, p₅ and p₈, respectively.