Faculty Bibliography

Oral medicine is the dental specialty dedicated to the oral health care of medically complex patients and the diagnosis and management of medically related diseases, disorders, and conditions affecting the oral and maxillofacial region. Like other dental and medical specialties, oral medicine patient care is often impacted by challenges such as limited manpower, time, or resources. Artificial intelligence (AI) tools seek to supersede these challenges by automating human tasks and ushering greater efficiency and productivity. For direct patient care in oral medicine, AI has applications in risk prediction modeling, diagnosis establishment, treatment decision-making, and prognosis and outcomes prediction modeling.

Oral medicine is the dental specialty dedicated to the oral health of medically complex patients and the diagnosis and management of medically-related diseases, disorders, and conditions affecting the oral and maxillofacial region. In addition to direct patient care, Oral Medicine specialists often engage in indirect patient care activities such as patient education and practice administration and/or academic activities such as student education and research. Artificial intelligence (AI) tools have been increasingly studied to facilitate these domains and contribute to more positive outcomes for practitioners, patients, and students alike. A review of the literature on these AI applications in Oral Medicine and related medical and dental fields provides an understanding of their current advantages and limitations.

OBJECTIVE:To define oncologic outcomes in Veterans in the modern era using a multi-institutional cohort designed to support development and validation of prognostic and predictive biomarkers for oropharyngeal squamous cell carcinoma (OPSCC). METHODS:A retrospective analysis was conducted including adult OPSCC patients treated at one of nine Veterans Affairs Medical Centers between 2000 and 2024; inclusive of 597 HPV-associated and 197 HPV-independent tumors. All patients were treated with curative intent external beam radiotherapy (100%) with (90%) or without concurrent chemotherapy. RESULTS:A total of 894 adult patients (mean age, 64 years; 881 (99.5%) male) were included in the study; 22% of patients self-identified as Black. The estimated 2- and 5-year OS rates for the entire cohort were 71% and 54%, respectively and lagged substantially behind locoregional control (LRC) and distant metastatic control (DMC). For Veterans with HPV-associated OPSCC, LRC and DMC at 5 years were 87% and 87% respectively. The strongest drivers of OS and LRC were T-classification and chemotherapy choice on univariate and multivariate analysis. CONCLUSIONS:Although LRC and DMC rates among Veterans track well with recently completed clinical trial outcomes, OS rates lag substantially suggestive of higher rates of non-cancer-specific mortality. Together, these data suggest that predictive biomarker strategies focused on treatment effectiveness should be predicated on LRC and DMC rather than OS. This multicenter study is the first step in providing a robust dataset capable of developing and optimizing artificial intelligence (AI)-informed prognostic and predictive strategies essential to a precision oncology approach to OPSCC.

Spinal tumor surgery requires rapid tissue diagnosis to guide surgical decisions and further treatment strategies, yet current intraoperative methods are time-intensive and require specialized expertise. No AI systems exist for real-time spinal tumor classification during surgery. We developed SpineXtract, the first AI-powered system for rapid intraoperative spinal tumor diagnosis using stimulated Raman histology (SRH) - a label-free Raman spectromics imaging technique without tissue processing available during surgery. We created a transformer-based classifier optimized for spinal tissue characteristics to identify common tumor types: meningioma, schwannoma, ependymoma, and metastasis. The system was tested in an international, multicenter, simulated, single-arm study using existing SRH datasets (44 patients, 142 slide-images) from three international institutions, with final pathological diagnosis as reference standard. SpineXtract achieved a 92.9% macro-average balanced accuracy (95% CI: 85.5-98.2) within 5 minutes (tumor-specific accuracy range, 84.2-98.6%), while providing quantitative microscopic feedback for granular tissue analysis. Performance remained consistent across institutions (macro balanced accuracy 91.4-92.0%) and outperformed existing brain tumor classifiers by 15.6%. Our results demonstrate clinical applicability, enabling rapid intraoperative diagnosis with performance exceeding current methods, potentially transforming intraoperative diagnostic workflows in spinal tumor surgery.

Facial nerve palsy is a debilitating condition with substantial physical and psychosocial impacts. Facial reanimation encompasses surgical reconstructive procedures aimed at restoring the functions of the facial nerve to improve function and quality of life in patients with facial palsy. This educational review outlines the essential principles for interpreting imaging studies for facial reanimation including fundamental anatomy, technical descriptions and imaging appearances of common reconstructive procedures, and key findings that should be included when reporting studies for patients being considered for facial reanimation. The information provided in this review equips radiologists to contribute effectively to a multidisciplinary team necessary for the treatment of patients with facial nerve palsy.

OBJECTIVE:Nasolacrimal duct obstruction (NLDO) in children typically resolves without surgery. Endoscopic dacryocystorhinostomy (En-DCR) is considered in cases refractory to irrigation, probing, and/or stent placement. The incidence of revision after pediatric En-DCR ranges from 0% to 22%. The objective of this review is to determine the incidence of revision and failure after pediatric En-DCR. DATA SOURCES/METHODS:In this systematic review, Medline, Embase, and Cochrane databases were searched on 11/21/2025. REVIEW METHODS/METHODS:Studies investigating primary, pediatric En-DCR outcomes were included. Case reports and articles that published no primary data or reported results aggregated with data from adult, revision, or external DCRs were excluded. Two reviewers (M.H. and R.W.) selected studies using these criteria and assessed quality with the Newcastle-Ottawa Scale and Cochrane Risk of Bias Tool. R V4.1.1 and GraphPad Prism 10.3.1 were used in statistical analysis and creation of Forest plots. The study protocol was pre-registered with Prospero. RESULTS:Thirty-one studies were included, involving 1470 ducts in 1230 patients. The mean age of these patients was 5.3 years old, and the study population was 49.5% male. Revision was performed in 9.1% of cases, and surgical failure occurred in 11.7% of cases. The mean follow-up time was 17.4 months. CONCLUSION/CONCLUSIONS:En-DCR is an effective treatment for NLDO. The incidence of revision was found to be lower than that of surgical failure, potentially due to short follow-up times of some studies or reluctance to undergo revision. The calculated incidence of complications and revision may underestimate true values due to significant heterogeneity among studies.

OBJECTIVES/HYPOTHESIS/OBJECTIVE:This study sought to characterize the attitudes of performers regarding factors commonly believed to influence the voice. The primary goal was to describe the vocal health habits of singers to provide a framework for clinical counseling and a potential schema for future investigation. Secondary objectives included examining changes in these attitudes over time and identifying differences between professional and amateur vocalists. STUDY DESIGN/METHODS:Prospective, longitudinal analysis. METHODS:A longitudinal assessment was performed across ten years at two time points (2013 and 2023) to investigate potential changes in attitudes and practices over time. Participants were recruited by word of mouth and social media. RESULTS:One hundred nine participants completed the 2013 survey, approximately half of whom were professional vocalists (52.3% professionals vs 47.7% amateurs), defined by those who either considered themselves a professional performer or had at some point obtained their primary source of income from singing or performing. The 2023 cohort included 155 participants who completed the survey, 61.3% of whom were classified as professional vocalists, with the remaining 38.7% classified as amateurs. In both the 2013 and 2023 cohorts, professional vocalists were more likely than amateurs to report factors that affected their voice. Using 2013 survey results, five vocal habits showed statistically significant differences in how amateurs and professionals judged their effects: warming up for 20-60 minutes prior to singing, decongestants, holding breath, cough drops (menthol), and one glass of wine. Using 2023 survey results, eight vocal habits showed statistically significant differences in how amateurs and professionals judged their effects: belting, cough drops (menthol), decongestants, holding breath, menthol spray, spicy foods, steroids, and Throat Coat. CONCLUSION/CONCLUSIONS:This study provides a rich characterization of factors vocalists report as influencing vocal health to varying degrees. Singers' beliefs about vocal health practices at times diverged from available scientific evidence.

OBJECTIVE/UNASSIGNED:Cochlear implant (CI) speech processors have undergone technological advancements. Therefore, patients upgrade speech processors when new features are available or when their previous device becomes broken and is no longer serviceable. This study aimed to identify factors that impact the time to a speech processor upgrade and to evaluate patient experiences with upgrading and following upgrade. METHODS/UNASSIGNED:In this retrospective cohort study, 46 CI surgeries at a single tertiary care center in 2017 and that subsequently received a speech processor upgrade were included. Data on patient demographics, hearing loss history, CI manufacturer, insurance type and status, and configuration were collected. Time to first upgrade, reasons for upgrade, patient-reported satisfaction, and speech perception scores were analyzed. RESULTS/UNASSIGNED:The mean time to a speech processor upgrade was 5.13 years after implantation. The most common reason for an upgrade was the device being over five years old, followed by the device being out of warranty. 45.7% of patients expressed satisfaction with speech processor upgrade, while 8.7% were not satisfied. There were no statistically significant associations between the time to upgrade and demographic factors such as age, sex, insurance type, or CI manufacturer. Following the upgrade, there were no significant changes in speech perception scores. CONCLUSION/UNASSIGNED:Speech processor upgrades at this center align with when insurance companies typically deem upgrades medically necessary. Demographic factors, insurance, and device manufacturer did not significantly influence time to upgrade. While objective speech perception measures did not significantly improve, many patients reported subjective satisfaction with the upgrade.

OBJECTIVE:We sought to apply a previously developed transimpedance (TIM) heatmap pattern classification scheme in patients with no known risk factors for cochlear anomalies, in addition to patients implanted in the revision setting, to better understand the incidence of pattern variants, and potential clinical implications. STUDY DESIGN/METHODS:Single-center retrospective review. SETTING/METHODS:Tertiary referral centre. PATIENTS/METHODS:Patients older than 6 months of age who underwent cochlear implantation between June 2020 and June 2024 with normal gross cochlear anatomy and no concern for fibrosis that had intraoperative TIM testing completed. Patients undergoing revision implantation were also included as a separate cohort. INTERVENTION/METHODS:None. MAIN OUTCOME MEASURES/METHODS:The number of patients with normal and variant TIM patterns was evaluated for each cohort. TIM patterns were subsequently compared with the electrode position found on intraoperative x-ray. RESULTS:There were 321 ears that underwent implantation and subsequent intraoperative TIM assessment meeting inclusion criteria. Of these, 310 (96.6%) were in the primary surgery setting, and 11 (3.4%) were in the revision surgical setting. In the primary surgical setting, 86.4% (n=268) of the implants demonstrated a normal TIM heatmap. Compared with the primary surgical setting, where only 45.5% (n=5) of revision surgery TIM heatmaps were interpreted as normal. One patient in the revision setting had a newly identified "double X" pattern corresponding to a normal electrode position on x-ray. CONCLUSIONS:There is a decreased incidence of previously developed TIM heatmap pattern variants in CI recipients with normal gross cochlear anatomy.

Neuromodulatory subcortical systems (NSSs) are uniquely susceptible to dementia-related pathology, leading to frequent molecular and behavioral impairments associated with altered function of these nuclei. Some of these systems display clear sex-specific cytoarchitecture and signaling leading to distinct physiology and behavioral outputs in males and females, while other regions display nominal sex differences. However, the relevance of sex differences in modulating dysfunction of NSSs in Alzheimer's disease (AD) and related dementias is not well understood. This review is a joint effort by the Neuromodulatory Subcortical Systems and Sex and Gender Differences in Alzheimer's Disease Professional Interest Areas of the Alzheimer's Association. We review sex differences in NSSs, both in non-disease states and in AD models and patients. We highlight the possible role of NSSs in driving sex-specific AD susceptibility and potential footholds for sex-based interventions targeting these systems. We conclude by outlining immediate and long-term actions to address the intersection of NSSs, sex, and AD.