This PDF file contains the front matter associated with SPIE Proceedings Volume 12354, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

Earwax or cerumen is a substance secreted by the ceruminous and sebaceous glands of the ear canal. The main function of this biofluid is as a physical barrier, but its buildup can lead to earwax impaction and result in hearing loss. Optical coherence tomography (OCT) is one potential method for assessing earwax. A catheter-based OCT system with a handheld probe and custom-made 3D-printed specula was designed and used to non-invasively acquire cross-sectional and volumetric images of the canal of adult human subjects. Features relating to quantity, structure, texture, and optical attenuation were extracted and correlated back to subjects’ ear health.

Otitis media (OM) is a prevalent disease among children worldwide. Antibiotic-resistant bacterial biofilms can develop in the middle ear during recurrent/chronic ear infections. OCT was used to compare microstructural texture features from primary bacterial biofilms in vitro. From 1200 ROI images of each biofilm class, 934 texture features were extracted. Principle component analysis and five-fold cross-validation were performed using Support vector machines (SVMs). Currently, the model has achieved 0.97 AUC (cubic kernel function) and an average classification accuracy of 89%. Texture analysis of bacterial biofilm OCT images with SVM may enable real-time visualization and differentiation of OM-causing bacterial biofilms in vivo.

Otitis media (OM) is a common disease of the middle ear, with 80% of children experiencing an infection before age three. Diagnostic methods rely on interpretation of symptoms from an otoscope, which help physicians visualize the eardrum. To provide precise structural and biochemical information, a prototype non-contact multimodal Raman spectroscopy (RS) and optical coherence tomography (OCT) system and handheld probe were created. Observation of in vitro physiologically-relevant ear models and comparison to in vivo scans from pediatric subjects presenting with OM detail application-specific development. Design challenges for clinical use, including maximum permissible exposure and physical size constraints, are presented.

Endoscopic optical coherence tomography (OCT) enables the assessment of the eardrum and the middle ear in vivo. However, revealing the ossicles is often limited due to shadowing effects of preceding structures and the 3D impression is difficult to interpret. To compare the identified middle ear structures, OCT and cone-beam CT of a patient were spatially aligned and showed a good agreement in locating malleus and the promontory wall. As CT imaging uses ionizing radiation and is thus limited in application, we furthermore provide a concept how radiology can be utilized as a priori knowledge for OCT imaging. Therefore, a statistical shape model derived from μCT data of temporal bone specimens was fitted to in vivo OCT measurements, potentially providing a real-time augmentation of endoscopic OCT for middle ear diagnostics in the future.

Anatomic optical coherence tomography (aOCT) endoscopy, combined with intraluminal pressure measurement, can be used to quantify upper airway wall mechanical properties for treatment planning and monitoring in upper airway obstructive disorders. Using an anatomic aOCT endoscope with integrated pressure catheter to measure airway deformation as a function of pressure, two different mechanical models are explored for airway wall elastography: a local compliance model and an elastic spring model. The models are compared over the circumference of the airway. Results illustrate differences between local compliance, which focuses on the radial deformation, and the elastic spring model, which treats expansion along the circumference.

Mucociliary clearance is vital for preventing any foreign substances from entering the upper airway that can later develop into acute and/or chronic respiratory diseases. Therefore, it is essential to further advance our understanding of the mucociliary functions. Our lab has been able to make key developments in imaging cilia, specifically measuring cilia beat frequency, with phase-resolved Doppler optical coherence tomography. In this system, we have further developed the system by incorporating phase-resolved spectrally encoded interferometric microscopy (SIEM) system with an FDML laser with MHz sweep rate to image cilia with higher accuracy and to minimize motion artifacts. In addition, we have designed a compact handheld probe system with a GRIN lens for easier in vivo imaging. The development of this system will allow us to further investigate cilia dynamics and ultimately utilize the system for clinical applications.

Creation of sub-epithelial voids within scarred vocal folds via ultrafast laser ablation may help in localization of injectable biomaterials towards an improved treatment for vocal fold scarring. Here, we investigated the feasibility of this technique by performing surgery in two adult canines using a miniaturized ultrafast laser surgery system. We identified ~2×2-mm2 voids ~110 μm beneath the epithelium surface using in situ endoscopic imaging and follow-up two-photon imaging and histology. We confirmed localization of an injectable model-biomaterial into a subset of these voids with fluorescence imaging. Future long-term studies on large animals will utilize the surgery system to further assess the clinical viability of such a vocal fold scarring therapy.

We report a new approach for label-free metabolic imaging of oral lesions using a novel fluorescence lifetime imaging endoscopy (FLIMEnd) system. The FLIMEnd system is capable of simultaneous dual-wavelength excitation and fluorescence detection at four emission spectral bands. Preliminary results indicate that established metabolic autofluorescence biomarkers of oral pre-malignancy and malignancy can be clinically imaged with this novel FLIMEnd technology. Future work will focus on identifying both established and novel autofluorescence biomarkers of oral pre-cancer and cancer that can be used to develop machine-learning driven FLIMEnd systems for early detection of oral cancer.

Modern noninvasive imaging techniques are used in many areas of medical research, particularly where an additional modality is required to support close monitoring during pre- or post-care. In oral medicine, the combination of an experienced clinician and supporting noninvasive optical devices for oral imaging, such as optical coherence tomography (OCT), may ensure improved diagnosis. In this study, the pathological features of lichenoid oral mucosa were noninvasively imaged with our miniaturized pen-like OCT probe and compared to adjacent healthy areas. The non-homogeneous epithelial (EP) layer, increased cornification of the EP surface and unidentifiable basement membrane are the most important features recorded.

Current practices for detection of oral cancers require invasive biopsy and costly histopathologic assessment. We hypothesize that applying deep learning (DL) to volumetric images (n = 320 images, 141 healthy) acquired using optical coherence tomography (OCT) could facilitate image-guided biopsy by (1) providing accurate and objective assessments of the state of subsurface oral tissues where changes in morphology indicate disease, and (2) allowing for non-invasive identification of biomarkers associated with oral mucosa health. Preliminary DL model implementations have demonstrated Jaccard Similarity Index of 81.3% in automated identification of the stromal-epithelial boundary in the oral mucosa

Herein, we present an anatomy-specific classification model using FLIm to differentiate between benign tissue, dysplasia, and cancer within the oral cavity and oropharynx. A total of 54 features, comprising both time-resolved and spectral intensity features, were used to train and test the classification model. This anatomy-specific classifier improves on our previous classification approach, now yielding an overall ROC-AUC of 0.94 during binary benign vs. cancer classification, and 0.92 while discriminating between healthy, cancer, and dysplasia. The proposed classification model demonstrates that FLIm has the potential to be used as an adjunctive diagnostic tool to facilitate head and neck cancer surgical guidance.

Fluorescence imaging can result in poor tumor contrast due to non-specific probe accumulation of receptor targeted probes. Here we show using preclinical and clinical studies that fluorescence lifetime (FLT) imaging can significantly improve the sensitivity and specificity for tumor detection using epidermal growth factor receptor (EGFR) targeted near infrared probes. We also show that FLTs in tissue are highly correlated with receptor expression levels, thereby enabling quantification of receptor quantification in vivo. Ongoing efforts in our group towards the translation of FLT imaging for intraoperative image guidance during head and neck surgeries will also be discussed.

Near infrared autofluorescence (NIRAF) can aid in identifying parathyroid glands (PGs) during thyroidectomies. Additionally, surgeons use indocyanine green (ICG) to evaluate PG perfusion. However, there is spectral overlap between PGs and ICG making parathyroid identification challenging after ICG injection. ICG dose adjustment may be needed for quantitative tissue perfusion assessment without interfering with PG NIRAF detection. This work investigated the utility of probe-based detection to determine PG vascularity using ICG doses of 0.25 and 1.25 mg. A mathematical model was also developed to determine the ICG dosage threshold for concurrent vascularity assessment and NIRAF detection of PGs.

Parathyroid gland blood supply disruption causes post-surgical complications in nearly half of thyroid surgeries. Assessment of parathyroid gland vascularity is thus critical to preserving parathyroid function. The previous study used laser speckle contrast imaging (LSCI) for parathyroid vascularity difference detection and showed reliability and promises that the LSCI device can provide during surgery. The current study takes a step forward and developed a handheld laser speckle contrast imaging device that had shown the ability to detect flow and distinguish vascular of parathyroids in vivo.

Trigeminal neuralgia (TN) is considered one of the most severe and most common facial pain. TN is characterized by a pain similar to an electric shock or sharp pain, restricted to one or more branches of the trigeminal nerve. TN can develop from idiopathic causes, and/or may be related to other conditions such as neoplasms, multiple sclerosis, nerve impairments. The most used TN treatments are drug or surgical therapies. However, both can cause side effects. A potential alternative or complementary TN treatment is low-level laser therapy (LLLT), which is known to control different types of pain. LLLT use in the treatment of trigeminal neuralgia as an adjunct to drug therapy has shown considerably positive results in previous studies. Our work evaluated the results achieved in the drug treatment associated with LLLT at red (660 nm) and infrared (808 nm) wavelengths based on a clinical case report of a TN patient. The Recover® and Vacumlaser® (MMOptics, São Carlos, SP, Brazil) devices were used, with applications in the region of the trigeminal nerve branches, three times a week, according to the protocol developed by researchers at the Physics Institute of São Carlos (IFSC-USP). The treatment outcome was assessed at every session regarding the patient’s progression. The clinical outcome observed was outstanding, resulting in a reduction of between 50-100% of the dosage of every medicine prescribed. We hypothesize that such result occurred due to photobiomodulation effects providing neuroprotection through anti-inflammatory, analgesic and antioxidant pathways, which can promote the metabolic homeostasis necessary for the reestablishment of tissue function.

Currently in the realm of medical diagnostics there is an abundance of data – more than can be analyzed by humans in a finite amount of time. Specifically in otolaryngology, there are numerous biophotonics imaging modalities used in a wide spectrum of disease processes from benign to malignant. Within the past decade, artificial intelligence (AI) has been an increasingly studied field as there is a need for accurate, expeditious, and automated data processing. However, application of AI in this field is still expanding. As AI can provide robust and automated processing, it can be integrated into the physicians’ toolbox.

Past studies used the VCSEL OCTs to image neonatal upper airways. However, due to manufacturing difficulties, the optimal focal length of the probe for VCSEL OCTs has yet to be determined. To determine this, both animal tracheas (duck and rabbit) and in vivo intubated neonates' upper airways were imaged with different probe focal lengths. In this study, the optimal VCSEL OCT probe focal lengths were provided for better image quality.

Acute Respiratory Distress Syndrome (ARDS) is a heterogenic clinical condition that affects critically-ill patients and is associated with high mortality rates and treatment costs. It is characterized by severe acute hypoxemia and alveolar lung injuries. We previously designed an optical coherence tomography (OCT) system to evaluate the changes in mucosa thickness (MT) and proximal airway volume in a swine model after a smoke inhalation injury. However, the analysis relied on manual segmentation of OCT images. Since the manual segmentation of large amounts of OCT data is time-consuming, tedious, and prone to error, this study aims to assess proximal airway volume (PAV) using an automated method based on deep learning. We use convolutional neural networks (CNN) to calculate PAV in a swine model affected by ARDS. We compare the PAV of the swine affected by ARDS with non-ARDS swine. We evaluate OCT images obtained at baseline (BL), post-injury (PI), 24 hours, 48 hours, and 72 hours after smoke inhalation injury. The neural network is modeled utilizing the U-net architecture. The accuracy is evaluated by computing the Sørensen-Dice similarity coefficient. We also demonstrate the correlation between PAV and MT, PFR values obtained from our previous study.

Mucociliary clearance is an important physiological mechanism for clearing the upper airways. Previously, it has been shown that different disease processes and drugs affect ciliary beat frequency (CBF). Namely, epinephrine has been shown to accelerate CBF in various animal models. Additionally, phase contrast microscopy (PCM) and spectrally encoded interferometric microscopy (SEIM) have been used to image dynamic tissue of the upper airway. Herein, we explore the effects of epinephrine on human sinonasal mucosa through PCM and SEIM. Sinonasal mucosa was harvested from patients undergoing endoscopic sinus surgery (ESS). Tissue was imaged using PCM and SEIM, maintaining physiological temperature through the use of warmed HBSS and a heating plate. Videos were taken before addition of any drugs as baseline. Epinephrine was diluted to 1 mg/mL (1:1000) and 1mL of solution was introduced to the sinonasal mucosa. PCM and SEIM was performed after to determine effects of epinephrine on CBF. Data analysis was performed using MATLAB (Mathworks, Natick, Massachusetts). Human sinonasal mucosa, taken from various anatomic locations, showed CBF values on PCM and SEIM consistent with what has been shown in previous literature. Upon addition of epinephrine to sinonasal mucosa, a marked increase in CBF was observed in both PCM and SEIM. In conclusion, the addition of epinephrine to sinonasal mucosa increased ciliary beat frequency. This validates the use of SEIM for determining CBF in sinonasal tissues. Further studies include adding to our sample size to determine a more accurate magnitude of increase of CBF.

In this study, Optical Coherence Tomography (OCT) was used to image the large upper airway in a rabbit model. U-net convolutional neural network (CNN) was used to automate the segmentation of large airway edema and tissue changes. Peak edema volume was reached at 30-minutes post-chlorine gas exposure, then down trended until the 6-hour timepoint. Herein, we show the streamlining of OCT imaging analysis status-post chlorine inhalation injury using CNNs.