Publications

2024

1. IFAR

   Comparison of patient-reported outcomes and clinical characteristics among patients with pituitary macroadenomas and giant adenomas

   Jessa E. Miller, Hye Rhyn Chung, Benjamin R. Uy, Nikitha Kosaraju, and 16 more authors

   International Forum of Allergy & Rhinology, 2024

   Abs

   Key Points Patients with giant adenomas are more likely to have tumor extension into the paranasal sinuses. Compared to macroadenomas, giant adenomas are not associated with worse preoperative SNOT-22 scores.

2. AJO

   Impact of facility volume on survival in primary endoscopic surgery for sinonasal squamous cell carcinoma

   Benjamin F. Bitner, Nolan A. Huck, Pooya Khosravi, Sina J. Torabi, and 3 more authors

   American Journal of Otolaryngology, 2024

   Abs

   Objectives To evaluate the impact of facility volume on outcomes following primary endoscopic surgical management of sinonasal squamous cell carcinoma (SNSCC). Methods The 2010–2016 National Cancer DataBase (NCDB) was queried for patients diagnosed with T1-T4a SNSCC surgically treated endoscopically as the primary treatment modality. Factors associated with overall survival (OS) were evaluated, including facility volume. Results A total of 330 patients who underwent endoscopic surgical management of SNSCC were treated at 356 unique facilities designated as either low-volume (LVC; treating 1–2 cases; 0-75th percentile), intermediate-volume centers (IVC; 3–4 cases total; 75th–90th percentile), or 144 high-volume (HVC; treating 5+ cases total; >90th percentile) centers. HVC treated patients with higher T staging (42.1 % vs. 29.8 %) and tumors in the maxillary sinus (26.9 % vs. 13.2 %) and ethmoid sinus (10.3 % vs. ≤8.3 %), while LVCs treated lower T stage tumors (70.2 % vs. 57.9 %) and tumors that were located in the nasal cavity (70.2–78.5 % vs. 62.8 %). On multivariable analysis, factors associated with decreased OS included higher T stage (T3/T4a vs. T1/T2; OR 1.92, 95 % CI 1.06–3.47) and older age (>65 vs. <65; OR 2.69, 95 % CI 1.62–4.49). Cases treated at high-volume centers were not associated with a higher likelihood of OS when compared to low-volume centers (OR 0.70, 95 % CI 0.36–1.35). Conclusions HVC are treating more primary tumors of the maxillary and ethmoid sinuses and tumors with higher T stages with endoscopic approaches, although this does not appear to be associated with increased OS. Short summary Sinonasal squamous cell carcinoma (SNSCC) presents late in disease process with poor prognosis. We investigated the impact of facility volume on outcomes following endoscopic treatment of SNSCC. High-volume centers treat more advanced and complex disease with comparable OS.

3. OHNS

   How Variable are Patient Comorbidity Profiles Among Practicing Otolaryngologists?

   Sina J. Torabi, Ellen M. Hong, Rahul A. Patel, Theodore V. Nguyen, and 4 more authors

   Otolaryngology–Head and Neck Surgery, 2024

   Abs

   Abstract Objective To determine whether certain groups of otolaryngologists (ORLs) are treating cohorts of patients with more comorbidities. Study Design Cross-sectional population-based analysis. Setting 2019 Medicare Provider Utilization and Payment Dataset. Methods Each ORL’s average Medicare hierarchical condition category (HCC) risk score, a comorbidity index calculated from a patient’s comorbidities, was collected. These were stratified and compared by various physician characteristics, including practice region and rurality, years in practice, gender, subspecialty, and setting (academic vs community). Results Among 8959 ORLs, the mean HCC risk score for Medicare patients was 1.35 ± 0.35. On univariate analysis, ORLs practicing in urban (compared to rural), ORLs in academic settings (compared to community), and early career ORLs all had a patient population with a higher HCC risk score (P < .001 for all). On multivariate analysis controlling for gender, rurality, graduation year, and region, rural setting was associated with decreased odds of having a high-risk patient population (odds ratio: 0.58 [95% confidence interval, CI: 0.48-0.71]; P < .001), while those more recently graduated has an increased risk (2000-2009: 1.41 [1.01-1.96], P = .046; 2010-2015: 2.30 [1.63-3.25], P < .001). In a separate subgroup analysis, subspecialty differences were seen and community setting was associated with decreased odds of having a high-risk patient population (0.36 [0.23-0.55]; P < .001). Conclusion There is variability in patient comorbidity profiles among ORLs, with those in urban settings, those more recently graduated, and those in academic settings treating a group with more comorbidities. As the comorbidity burden may increase the cost of practice and complications, these findings may have important implications for health inequity.

4. AJO

   Retinal Nerve Fiber Layer Optical Texture Analysis and 10-2 Visual Field Assessment in Glaucoma

   Alireza Kamalipour, Sasan Moghimi, Pooya Khosravi, Natchada Tansuebchueasai, and 8 more authors

   American Journal of Ophthalmology, 2024

   Abs

   Purpose To apply retinal nerve fiber layer (RNFL) optical texture analysis (ROTA) to 1) investigate the association between papillomacular and papillofoveal bundle defects with 10-2 visual field (VF) sensitivity abnormalities, and 2) integrate the information from RNFL bundle defect and 24-2 VF central test locations to determine the likelihood of 10-2 VF sensitivity abnormalities. Design Cross-sectional Methods A total of 841 eyes (144 healthy, 317 glaucoma suspect, and 380 glaucoma) of 442 participants were included. Eyes underwent 24-2, and 10-2 VF testing and OCT for ROTA. The borders of RNFL defects were delineated from ROTA, and the involvement of the arcuate, papillomacular, and papillofoveal bundles was determined for each eye. Multilevel logistic regression analysis was applied to evaluate the structure–function association. Results Papillomacular (92.1%) and papillofoveal (37.9%) RNFL bundle defects were prevalent in eyes with glaucoma. A 10-2 VF location that was projected onto a papillomacular or a papillofoveal RNFL bundle defect had a significantly increased likelihood of reduced sensitivity (ORs of 18.61 at PDP < 5%, and 20.17 at TDP < 5%, respectively, P < 0.001 for both). When predicting the likelihood of VF abnormality in a 10-2 test location, noticeably higher odds ratios were observed when overlapping with an RNFL bundle defect, compared to when an abnormal corresponding 24-2 central point was present. Conclusions Papillomacular and papillofoveal RNFL bundle defects are present in a considerable proportion of eyes with glaucoma. When detected, they significantly increase the likelihood of abnormality in the corresponding central VF test locations assessed by the 10-2 test.

5. Medical image registration via neural fields

   Shanlin Sun, Kun Han, Chenyu You, Hao Tang, and 7 more authors

   Medical Image Analysis, 2024

   Abs

   Image registration is an essential step in many medical image analysis tasks. Traditional methods for image registration are primarily optimization-driven, finding the optimal deformations that maximize the similarity between two images. Recent learning-based methods, trained to directly predict transformations between two images, run much faster, but suffer from performance deficiencies due to domain shift. Here we present a new neural network based image registration framework, called NIR (Neural Image Registration), which is based on optimization but utilizes deep neural networks to model deformations between image pairs. NIR represents the transformation between two images with a continuous function implemented via neural fields, receiving a 3D coordinate as input and outputting the corresponding deformation vector. NIR provides two ways of generating deformation field: directly output a displacement vector field for general deformable registration, or output a velocity vector field and integrate the velocity field to derive the deformation field for diffeomorphic image registration. The optimal registration is discovered by updating the parameters of the neural field via stochastic mini-batch gradient descent. We describe several design choices that facilitate model optimization, including coordinate encoding, sinusoidal activation, coordinate sampling, and intensity sampling. NIR is evaluated on two 3D MR brain scan datasets, demonstrating highly competitive performance in terms of both registration accuracy and regularity. Compared to traditional optimization-based methods, our approach achieves better results in shorter computation times. In addition, our methods exhibit performance on a cross-dataset registration task, compared to the pre-trained learning-based methods.

2023

1. IJMS

   Deep Learning Approach for Differentiating Etiologies of Pediatric Retinal Hemorrhages: A Multicenter Study

   Pooya Khosravi, Nolan A. Huck, Kourosh Shahraki, Stephen C. Hunter, and 8 more authors

   International Journal of Molecular Sciences, 2023

   Abs

   Retinal hemorrhages in pediatric patients can be a diagnostic challenge for ophthalmologists. These hemorrhages can occur due to various underlying etiologies, including abusive head trauma, accidental trauma, and medical conditions. Accurate identification of the etiology is crucial for appropriate management and legal considerations. In recent years, deep learning techniques have shown promise in assisting healthcare professionals in making more accurate and timely diagnosis of a variety of disorders. We explore the potential of deep learning approaches for differentiating etiologies of pediatric retinal hemorrhages. Our study, which spanned multiple centers, analyzed 898 images, resulting in a final dataset of 597 retinal hemorrhage fundus photos categorized into medical (49.9%) and trauma (50.1%) etiologies. Deep learning models, specifically those based on ResNet and transformer architectures, were applied; FastViT-SA12, a hybrid transformer model, achieved the highest accuracy (90.55%) and area under the receiver operating characteristic curve (AUC) of 90.55%, while ResNet18 secured the highest sensitivity value (96.77%) on an independent test dataset. The study highlighted areas for optimization in artificial intelligence (AI) models specifically for pediatric retinal hemorrhages. While AI proves valuable in diagnosing these hemorrhages, the expertise of medical professionals remains irreplaceable. Collaborative efforts between AI specialists and pediatric ophthalmologists are crucial to fully harness AI’s potential in diagnosing etiologies of pediatric retinal hemorrhages.

2. MICCAI

   MedGen3D: A Deep Generative Framework for Paired 3D Image and Mask Generation

   Kun Han, Yifeng Xiong, Chenyu You, Pooya Khosravi, and 4 more authors

   In Medical Image Computing and Computer Assisted Intervention – MICCAI 2023, 2023

   Abs

   Acquiring and annotating sufficient labeled data is crucial in developing accurate and robust learning-based models, but obtaining such data can be challenging in many medical image segmentation tasks. One promising solution is to synthesize realistic data with ground-truth mask annotations. However, no prior studies have explored generating complete 3D volumetric images with masks. In this paper, we present MedGen3D, a deep generative framework that can generate paired 3D medical images and masks. First, we represent the 3D medical data as 2D sequences and propose the Multi-Condition Diffusion Probabilistic Model (MC-DPM) to generate multi-label mask sequences adhering to anatomical geometry. Then, we use an image sequence generator and semantic diffusion refiner conditioned on the generated mask sequences to produce realistic 3D medical images that align with the generated masks. Our proposed framework guarantees accurate alignment between synthetic images and segmentation maps. Experiments on 3D thoracic CT and brain MRI datasets show that our synthetic data is both diverse and faithful to the original data, and demonstrate the benefits for downstream segmentation tasks. We anticipate that MedGen3D’s ability to synthesize paired 3D medical images and masks will prove valuable in training deep learning models for medical imaging tasks.

3. AJO

   Deep learning estimation of 10-2 visual field map based on circumpapillary retinal nerve fiber layer thickness measurements

   Alireza Kamalipour, Sasan Moghimi, Pooya Khosravi, Mohammad Sadegh Jazayeri, and 7 more authors

   American Journal of Ophthalmology, 2023
4. AJO

   Combining Optical Coherence Tomography and Optical Coherence Tomography Angiography Longitudinal Data for the Detection of Visual Field Progression in Glaucoma

   Alireza Kamalipour, Sasan Moghimi, Pooya Khosravi, Vahid Mohammadzadeh, and 7 more authors

   American Journal of Ophthalmology, 2023
5. Artificial neural network prediction of post-thyroidectomy outcome

   Kotaro Tsutsumi, Khodayar Goshtasbi, Khwaja H Ahmed, Pooya Khosravi, and 5 more authors

   Clinical Otolaryngology, 2023

2022

1. OHBM

   Machine Learning in the Management of Lateral Skull Base Tumors: A Systematic Review

   Kotaro Tsutsumi, Sina Soltanzadeh-Zarandi, Pooya Khosravi, Khodayar Goshtasbi, and 2 more authors

   Journal of Otorhinolaryngology, Hearing and Balance Medicine, 2022

2021

1. Adapting personal therapies using a mobile application for tinnitus rehabilitation: a preliminary study

   Mehdi Abouzari, Khodayar Goshtasbi, Brooke Sarna, Yaser Ghavami, and 6 more authors

   Annals of Otology, Rhinology & Laryngology, 2021
2. A web-based deep learning model for automated diagnosis of otoscopic images

   Kotaro Tsutsumi, Khodayar Goshtasbi, Adwight Risbud, Pooya Khosravi, and 4 more authors

   Otology & Neurotology, 2021

2020

1. Prediction of vestibular schwannoma recurrence using artificial neural network

   Mehdi Abouzari, Khodayar Goshtasbi, Brooke Sarna, Pooya Khosravi, and 4 more authors

   Laryngoscope Investigative Otolaryngology, 2020

2019

1. Medical malpractice of vestibular schwannoma: a 40-year review of the United States legal databases

   Jack Birkenbeuel, Kimberly Vu, Brandon M Lehrich, Mehdi Abouzari, and 7 more authors

   Otology & neurotology: official publication of the American Otological Society, American Neurotology Society [and] European Academy of Otology and Neurotology, 2019

2018

1. IJMS

   Fetal bovine serum-derived extracellular vesicles persist within vesicle-depleted culture media

   Brandon M Lehrich, Yaxuan Liang, Pooya Khosravi, Howard J Federoff, and 1 more author

   International journal of molecular sciences, 2018