The National Institute on Deafness and Other Communication Disorders (NIDCD), and the National Institutes of Health (NIH) held a workshop on November 20-21, 2019, on visualizing the human inner ear. Workshop participants from the auditory, vestibular, and imaging research fields discussed the current state, advancements, and limitations of visualizing the structures of both living and postmortem auditory and vestibular systems. One long-term goal is to image living human inner ear tissue at high resolution across both structural and functional levels. Such imaging will greatly advance the ability to diagnose and treat auditory and vestibular deficits.
Human temporal bone specimens, from normal and disease-state subjects, are particularly difficult to retrieve and prepare for postmortem study, as are smaller samples from living humans. Encased within the thick temporal bone, the auditory and vestibular systems are also difficult to view in living humans. Visualization of postmortem human temporal bone specimens is possible, but specimens are laborious to retrieve and prepare for study. Moreover, the number of laboratories experienced in the processing of postmortem human temporal bones is decreasing.
Through interactive and thought-provoking discussions, workshop participants identified gaps in knowledge, discussed the current state of imaging technology, and considered the technological developments necessary to attain the goal of high-resolution imaging of the structure and function of living human inner ear tissue. Participants shared their specific areas of expertise; discussed overall issues at a more global level; and considered, among other topics, the current state of the field, pieces of information that are helpful to clinicians, additional questions that could be answered with improved imaging techniques, the pros and cons of current techniques, missing technical components, techniques that might be translated from animals to humans, and other scientific fields that might provide useful answers or techniques.
Detailed images of the structure and functioning of the living human inner ear would greatly advance the NIDCD’s mission to help prevent, detect, diagnose, and treat deafness, balance, and other communication disorders. To date, such images have not been possible, but technology is rapidly evolving, putting a once lofty goal within reach.
Participants
The workshop was organized by Drs. Amy Poremba and Janet Cyr from the NIDCD’s Division of Scientific Programs, co-chaired by Dr. George Spirou from the University of South Florida and Dr. Konstantina Stankovic from Massachusetts Eye and Ear/Harvard Medical School.
Twelve additional extramural scientists from the U.S. and abroad were invited to bring their expertise to the discussion (see participant list). In addition, Dr. Bechara Kachar and Dr. Michael Hoa from the NIDCD’s intramural research laboratories participated. The workshop was open to all interested members of the scientific community.
Workshop Format
The workshop began with welcoming remarks from Dr. Poremba and introductions from the invited workshop participants, followed by introductory comments from Dr. Cyr and Dr. Spirou. The rest of day one consisted of three groups of presentations, each followed by panel discussions. Short summaries of the presentations appear below.
Day two of the workshop consisted of a presentation of clinical case studies and a discussion on the diagnosis and treatment of those cases with current technologies and tests. To identify gaps in the field, participants also raised questions that are not currently answered with ease. The discussion was facilitated by Dr. Stankovic.
After the case studies, the participants broke into four subgroups, each focused on a topic identified on day one of the workshops:
- Group A – Imaging Living Human Tissue
- Group B – Imaging Animal Tissue
- Group C – Multiuse of Human Temporal Bones and Molecular Profiling
- Group D – Data Sharing
Each subgroup then presented a brief synopsis of their breakout discussions with an emphasis on the current state of scientific knowledge and the identified challenges, obstacles, and unanswered questions.
Overall Summary
The delicate structures of the human inner ear are surrounded by the temporal bone, which serves as protective armor. This anatomical location hinders the ability of clinicians and scientists to visualize the cellular components of the auditory and vestibular systems at sufficient resolution to allow for detailed diagnosis and treatment of many causes of auditory and vestibular dysfunction. Presently, we do not have techniques with high enough resolution for structural imaging—let alone functional imaging—of the living human inner ear. Current clinical imaging techniques are generally limited to standard computerized tomography (CT) without contrast or magnetic resonance imaging (MRI) with contrast. Several deficiencies exist with these techniques, including:
- CT doesn’t differentiate soft tissue well.
- MRI doesn’t differentiate air from the bone.
- Neither CT nor MRI shows inner ear/cochlea vasculature well.
- CT and MRI from imaging centers are often only axial plus coronal, and given the placement of the auditory and vestibular end organs within the temporal bone, these imaging orientations are subpar.
- Cellular/subcellular resolution is not possible with these methods.
Improved imaging of the human inner ear is a critical unmet need that could directly impact patient care. Vestibular migraines and Ménière’s disease, for example, can present in very similar ways. It is currently difficult to distinguish between the two conditions, since there is no way to visualize the underlying inner ear pathology using existing imaging techniques, allowing a differential diagnosis. Treating severe Ménière’s disease could involve surgical ablation of parts of the ear, but ablation is not recommended for vestibular migraines and might exacerbate the condition. With improved imaging of the human ear, surgeons could provide more conservative treatment to those who would benefit. In addition, several inner ear gene therapies are under development, which may help to alleviate hearing and balance issues. Advanced imaging of the inner ear is needed to identify the correct patient cohort as candidates for such therapies in clinical trials. High-resolution imaging in real-time would also ensure the accuracy of delivery in gene therapy treatment.
The NIDCD convened a workshop of auditory, vestibular, and imaging experts to discuss imaging of the human ear, including discussions of the state of the science and unmet needs. Discussion throughout the workshop pointed to several areas where improvement in inner ear imaging would be desirable and could have an immediate impact on clinical practice. Examples included:
- For MRI, clinicians need a way to distinguish air from bone, a way to image the human ear in a living person, and techniques that can show a small volume of imaging that is surrounded by other material (such as bone). Some techniques are currently being explored, such as supplementing MRI with hyperpolarized helium to show air spaces (as used in sinuses).
- A type of cochlea/labyrinth micro-angiography is needed to visualize the loss of blood flow.
- For vestibular problems, methods are needed in living tissue to determine if calcium carbonate crystals (otoconia) have been displaced, and if so, in which direction the clinician should move the patient’s head to realign the otoconia.
- About Ménière’s disease, what is the underlying pathology of the inner ear? Although hydrops can be visualized, the presence of hydrops does not correlate 100% with symptoms. Current treatment for Ménière’s disease is based on symptoms; detailed images of the inner ear would provide critical information to determine which patients might benefit from treatment by surgical ablation.
- Higher resolution imaging and/or angiography would help to diagnose migraine-associated vertigo (which can be falsely diagnosed as Ménière’s disease), hydrops, inner ear angina, and clogged endolymphatic ducts.
- Cochlear and vestibular implantation would benefit from higher-resolution imaging to prevent the mispositioning of electrodes and/or devices.
Workshop participants also discussed the importance of postmortem human temporal bone studies and resources for translating hearing science studies to clinical treatment in humans. It is important to confirm that the hypotheses generated from animal models hold in humans before beginning to consider gene therapy or other treatments. Access to human auditory and vestibular tissue with known medical histories is an important tool to be able to make that translational step. Workshop participants also discussed the desire to standardize human temporal bone harvesting and processing for use by the scientific community. They emphasized the importance of improving imaging of harvested human and animal temporal bones before sectioning to facilitate prioritization for processing given the small number of laboratories processing bones and the extensive amount of time involved in doing so. They identified the need to expand the number of individuals across the country who are capable of performing the intricate task of temporal bone processing.
Day One Presentation Summaries
Konstantina Stankovic, Harvard University
Dr. Stankovic summarized what scientists can do and see within the human inner ear and the different technology available to explore the inner ear, the technologies’ capabilities and limitations, and the most promising translational technologies to assist scientists in examining, diagnosing, and treating the human ear. These technologies include micro-CT scans, synchrotron-based imaging, and optical coherence technology. She concluded with the importance of combining and further developing current technologies and techniques to vastly improve our ability to provide patients with effective personalized care to prevent disease and restore hearing and balance functions.
Akira Ishiyama, University of California, Los Angeles
Dr. Ishiyama presented research on methods for the imaging of endolymphatic hydrops (ELH), which is the primary pathological alteration in Ménière’s disease. The use of high-resolution 3D FLAIR MRI allows for consistent differentiation of endolymph from perilymph in the inner ear. Intratympanic injection of dilute gadolinium along with 3D FLAIR enhances the imaging contrast of the perilymph. Ménière’s disease is also associated with increased permeability in the blood-labyrinth barrier. He emphasized that delayed contrast MRI can demonstrate ELH in the cochlea and the vestibular system. The imaging can be used to assess treatment, confirm the diagnosis, and better understand the underlying pathophysiology of ELH.
Mark Ellisman, University of California, San Diego
Dr. Ellisman described how 3-dimensional electron microscopy can be used to image macromolecules, cells, and tissues. For example, he used single tilt series electron tomography to map the 3D architecture of ribbon synapses in hair cells. Now, extreme, multi-tilt, multi-angle TEM tomography can produce a 3D structure of cells by reconstructing images of the cells taken from multiple angles. Research is enabling the use of new probes for light and electron microscopy to study neurons in mapping synaptic connectivity. Serial block face imaging can also generate 3D images of tissue by using an in-situ ultramicrotome inside a scanning electron microscope. The ultramicrotome slices through the sample, and the block face is imaged by the scanning electron microscope. This can be repeated throughout the entire tissue sample to create a 3D image.
Bechara Kachar, National Institute on Deafness and Other Communication Disorders
Dr. Kachar explained that hearing and vestibular sensory organs are distinguished by their exquisite sensitivity, dynamic range, and frequency selectivity. Their extraordinary performance depends on intricate molecular and structural interactions at a broad range of lengths and time scales. There is a need to match the imaging method with the different physical properties of the tissue. No single microscopy technique gives a 360-degree view of the tissue structure of the inner ear, so different types of microscopic imaging technology must be employed. It is beneficial to use complementary imaging across different electron microscopy methods to view the same cells. To avoid fixation and/or dehydration artifacts, specimens can undergo complementary cryogenic electron microscopy methods, such as freeze-substitution or freeze-etching. Confocal microscopy can be used in conjunction with immunofluorescence to view protein localization and expression in cells.
Hanif Ladak, Western University
Dr. Ladak gave an overview of synchrotron-based imaging, which can be used to create more clearly defined imaging of the cochlea and cellular structures within it, such as hair cells. This imaging technique uses extremely bright light that is focused into collimated (parallel) beams of light. The wavelengths are manipulated and used for imaging objects. Synchrotron-radiation phase contrast imaging (SR-PCI) is used to greatly improve the clarity of X-ray and micro-CT imaging in a non-living specimen. Scientists and clinicians can use this method to view the cochlea for postmortem study. There is no need to prepare the temporal bone, which saves time and effort. The next steps in applying this new technology are to improve its ability to show where objects are about each other and to move to live animal testing. The overall goal is to improve and enhance imaging of the living human ear, but accomplishing this with the synchrotron is difficult due to the possible risks involved.
Brian Applegate, University of Southern California
Dr. Applegate described the use of optical coherence tomography (OCT) and a vibrometry system for functional and morphological imaging in the inner ear. OCT is analogous to ultrasound imaging but uses light instead of sound. OCT is already used to image the retina, coronary arteries, and esophagus. He described plans for developing a hand-held OCT endoscope that can view the microstructures of the cochlea through the round window membrane. OCT and vibrometry are powerful new tools that can be used for imaging the morphology and function of the middle and inner ear and provide physical evidence for diagnosis and guidance of therapy.
Peter Santi, University of Minnesota
Dr. Santi and his team use thin-sheet laser imaging microscopy (TSLIM), which is a technique that scans images from a light sheet fluorescence microscope that can be used on a larger specimen, such as the human cochlea. The samples are specially prepared through a process that involves bleaching and injecting fluorescence into the object being examined. The microscopic images are converted to 3D images. This technique has enabled Dr. Santi and his team to create 3D images of the cochlea and the organs within, and view hair cells and neurons. Detailed discussions at the end of his presentation included how postmortem tissues need to be prepared for different types of imaging.
Bryan Ward, Johns Hopkins University
Dr. Ward is applying his expertise in the vestibular system to research the causes of vertigo by using high-strength MRI machines to improve inner ear imaging. Currently, a 7-Tesla (7T) MRI is being used for research and diagnostic tools for some medical conditions. The issue is that the stronger the magnetic field is in the MRI, the clearer and more detailed the image is, but the vertigo effect on patients is also stronger. Currently, people are not able to use MRIs with more than 7T magnetic fields, though fields up to 11.7T can be produced to image specimens. Dr. Ward has researched the causes of vertigo and would like to find ways to overcome this and improve what doctors can see in the labyrinth and the vestibular system. If the vertigo problem initiated by the strong magnetic fields could be ameliorated, these high spatial resolution images could help diagnose the original patient complaints of unexplained dizziness and may lead to more conservative treatments.
Hideko Nakajima, Harvard University
Dr. Nakajima and her team are studying the mechanics of the cochlea and how it moves. The anatomy of the cochlear partition is similar among non-human animals, but in humans, there is a notable difference. The human ear contains a bridge of soft tissue between the osseous spiral membrane and the basilar membrane in the cochlea. In addition to basilar membrane movement seen in animals, in humans, the osseous spiral lamina and the bridge also move. Dr. Nakajima’s team has mathematically modeled the movement and is using OCT (optical coherence tomography) measurements to make images and take measurements. This will increase their understanding of the differences in how sound travels through the human cochlea and will update the current model of the cochlear partition and how it moves. The goal is to learn how these differences affect human hearing and illustrate the need for fresh human tissue.
Matthew Shew, Washington University School of Medicine
Dr. Shew discussed how machine learning can be applied to studying disorders of the inner ear. Machine learning enables computers to analyze and learn from large amounts of data and to perform specific tasks without explicit instructions, relying instead on patterns and inferences. Dr. Shew presented research on using machine learning to predict hearing loss based on perilymph microRNA expression. He noted that sensorineural hearing loss, conductive hearing loss, and Ménière’s disease exhibit disease-specific microRNA profiles within the inner ear. Machine learning to analyze microRNA profiles may be a possible diagnostic tool for inner ear disease.
Alexandra Pacureanu, University College of London
Dr. Pacureanu discussed how synchrotron X-ray microscopy could serve hearing research. The benefits of using X-ray microscopy to view tissue are numerous: high contrast; fast data collection; the ability to image large samples with relatively high resolution (e.g., the spatial resolution of tens of nm and 3D isotropic spatial resolution); and the preservation of tissue samples with no need to decalcify, viewing bone as well as soft tissue. Dr. Pacureanu also discussed improvement in phase contrast X-ray microscopy and the generation of phase maps that use tomographic reconstruction to develop high-resolution images of tissue. Correlative X-ray holography and X-ray fluorescence are additional techniques to image tissue. Dr. Pacureanu also shared the data storage and sharing practices of those in Europe.
Michael Hoa, National Institute on Deafness and Other Communication Disorders
Dr. Hoa and his lab are using single-cell RNA sequencing (RNA-seq) to identify different cells in the cochlea, with a focus on the stria vascularis, which generates energy for the ear and allows the hair cells to work. Their overall goal is to apply this technique to human temporal bones to study human inner ear diseases that currently do not have strong mouse models. Gene expression profiling is being used to make connections between gene expression and function and has become the starting point for a better understanding of diseases. The starting material is whole tissue or cell populations. Dr. Hoa’s lab is researching the possibility of using bulk and single-cell RNA-seq to identify dysfunctional cells that may be causing disease, and they emphasized the need to expand human temporal bone resources for this purpose.
Anna Lysakowski, the University of Illinois at Chicago
Dr. Lysakowski presented her research visualizing the vestibular periphery, specifically Type I and Type II hair cells. She described the advantage of using transmission electron microscopy (TEM) to view synaptic ribbons and afferent nerve endings in Type I and II vestibular hair cells and observed regional (central vs. peripheral) differences in the synaptic innervation in vestibular hair cells. She noted that her TEM studies showed central/striolar Type II hair cell ribbons are larger, longer, occur in clusters, and can be hollow when compared to peripheral Type II ribbons or any Type I ribbons.
Bradley Schulte, Medical University of South Carolina
Dr. Schulte presented on the importance of collecting, processing, and using temporal bones for pathological studies on presbycusis (age-related hearing loss). His initial research focused on pathophysiological and histopathological studies on gerbils and then shifted to morphologic and molecular studies of human temporal bones. The collection of animal and human temporal bones is a time-sensitive process but is crucial to pathological studies. For example, postmortem examination of temporal bone tissue can reveal possible causes of presbycusis, such as stria vascularis degeneration, loss of auditory nerve fibers and spiral ganglion neurons, damage and loss of auditory hair cells, and inflammation.
Charles Della Santina, Johns Hopkins University
Dr. Della Santina represented the viewpoint of ear surgeons and clinicians as end-users of research and diagnostic technology. He discussed the current state of diagnosing and treating patients and a wish list of goals for where clinicians would like to be and the capabilities needed to get there. Most of the 3D imaging and sequencing techniques discussed are not available for the day-to-day diagnosis and treatment of patients. Some current deficiencies are that CT scans do not differentiate soft tissue well, MRI doesn’t differentiate air from bone well, and no imaging allows the living inner ear and cochlea vasculature to be visualized with detail. Optimal imaging views are also not well covered by current technology and practice.
Day Two Breakout Group Summaries
Breakout Group A – Imaging Living Human Tissue
Studies of the human inner ear present scientists and clinicians with an opportunity to impact all of medicine, because many of the subcellular, cellular, and system-level components function at the extreme. Increased resolution and functional imaging are needed at every level—from the ear canal to the middle ear to the inner ear, and from the least to the most invasive. Foci could include examining the organ of Corti; visualizing otoconia to improve diagnoses of BPPV (benign paroxysmal positional vertigo), especially when repositioning maneuvers fail; and visualizing blood vessels, which may be as informative about inner ear disease conditions as it has been for heart disease, stroke, retinal disease, etc. Current technologies with low and high resolution could be combined. If MRIs and CT scans were to improve five to tenfold, scientists and clinicians could see many structures—including endocochlear or endovestibular structures—in sufficient detail to begin to impact clinical decision-making. Optical tools providing higher resolution, including optical coherence tomography (OCT) and micro OCT, might be coupled with functional tests such as electrophysiology or vibration patterns, the results of which would be better than a single modality alone and would allow for functional evaluation.
Breakout Group A also discussed possible ways to clear the temporal bone to help visualization, and other ways to increase imaging capabilities to visually penetrate through the bone in living human ears. Developing a synergy with other tools, such as molecular profiling, would greatly increase the usefulness of live images and might provide advantages to the clinic for diagnosis and treatment. Lastly, informing other areas of science and technology of the issues faced by clinicians in hearing and balance may speed up needed advancements.
Breakout Group B – Imaging Animal Tissue
There is a continued need for animal research of the auditory and vestibular systems at both the structural and functional levels, as most of the research questions cannot be answered in humans at this time due to the lack of technology. The group discussed the use of higher resolution imaging technologies, such as synchrotron imaging, for postmortem animal tissue and even perhaps living animal tissue. Functional studies are also of great interest in the living animal and could perhaps be translated for human use. Comparative studies of the inner ear (between animal models and humans) will also inform our clinical understanding. Breakout Group B also expressed the need to improve the sharing of tissue, genetic models, and data, as well as expand the number of animal and genetic models available to researchers.
Breakout Group C – Multiuse of Human Temporal Bones and Molecular Profiling
Postmortem human temporal bones and surgical samples from patients are precious resources needed for several research questions, including molecular profiling. Breakout Group C discussed how best to coordinate the use of postmortem human temporal bones across the country. Suggestions included discussion among researchers and interested parties to create a list of cooperative research activities and to try to standardize temporal bone and inner ear tissue processing protocols for specific types of cellular and molecular studies. The need for flexibility was also discussed, as some research protocols require specialty processing. The group discussed the benefits of obtaining high-quality imaging of samples before any tissue was cut for histological purposes, and they considered the types of additional data that would be useful to collect about the samples. It was suggested that existing human temporal bone resources, such as the NIDCD National Temporal Bone Registry and the NIDCD Temporal Bone Laboratory, could facilitate the collection of additional information or dissemination of standardized protocols. Discussants emphasized the continued need for postmortem human temporal bone tissue, especially because imaging of the living human inner ear is not yet possible at the needed resolution.
Breakout Group D – Data Sharing
Given limited resources and complex techniques and questions, science must become more collaborative. Since temporal bone and human surgical tissue data and collection can be both scarce and expensive, researchers want to use these resources to the fullest. Data sharing and archiving are important tools to make data accessible. Many practices from other fields can be adapted to make the data-sharing plans for auditory and vestibular tissues a reliable endeavor. The use of FAIR (findable, accessible, interoperable, and reusable) practices for data sharing and storage were discussed as important principles to uphold when thinking about how to approach the sharing of temporal bone data, auditory and vestibular processing patient test results, molecular profiling, and sensitive patient information. Education and dissemination of knowledge through a virtual temporal bone repository of images in a publicly accessible database would also improve knowledge and research in this area and would expand the scientific impact of this rare resource. Standardized data from different modalities need to be shared while maintaining flexibility.
Breakout Group D acknowledged that data stored in a centralized, accessible portal can be problematic in that imaging studies generate very large file sizes, and ideally one would also want to store the molecular data corresponding to that sample. Implementation of these data sites is likely to be expensive, would not be easily supported by existing funding sources, and requires cooperation with publication regulations to protect publishing rights. Despite these obstacles, the participants were uniform in their estimation that the effort would contribute to auditory and vestibular research in meaningful ways and positively impact clinical decision-making processes.
