Grant Recipient Search
	Keyword:
	Type:   Grants Only Awards Only Grants & Awards    Funding:   Currently Funded Previously Funded All Funding Periods    Sort By:   Name Start Date

	Total Recipients Returned: 74 (Recipients sorted by NAME)	Items per page: 5 | 10 | 25 | 50
	Current Keyword Filter: No Filter Applied
Page 1 of 8 (1 - 10)

1   Ulas Bagci, MSc Northwestern University Emerging Issues Grant - Long-term Effects of COVID-19 (2023 - 2024) PASC Pulmonary Fibrosis Prediction with Deep Learning and Multimodal Data   Abstract:   The significant portion of COVID-19 patients are at risk for developing long-term Post-Acute Sequelae of SARS-CoV-2 (PASC) with pulmonary fibrosis. Identifying patients who are most at risk for PASC pulmonary fibrosis could allow early intervention with anti-fibrotic drugs and other potential treatment strategies. The overall goal of this proposal is to address this unmet clinical need of predicting PASC pulmonary fibrosis at an early stage by developing novel explainable deep learning (DL) algorithms on multimodal data (combined imaging (initial CT scan) and electronic health record (EHR) data) from multiple centers. In Aim 1, we will curate multi-center data by gathering CT imaging and associated non-imaging (EHR) data from Northwestern, Columbia, NIH, and UPenn, and perform advanced radiologic image analysis. For each patient, we will collect EHR data consisting of demographic, clinical and laboratory information. In total, we expect to retrospectively collect a balanced dataset of 1150 initial CT scans (from distinct patients) and associated EHR data to be analyzed throughout the project. In Aim 2, based on our previously established capsule networks, we will develop explainable capsules (X-Caps) for prediction of PASC pulmonary fibrosis formation. We will integrate our newly established visual explanation algorithm (called IBA) into the machine generated results in addition to radiographical explanations, PA, and BCA, captured by the X-Caps. In Aim 3, we will employ our established optimal biomarker method (OBM) to determine the top features (potential biomarkers) from EHR and imaging data to predict PASC pulmonary fibrosis at the highest accuracy possible. PASC is a massive emergency and very little is known about it. Once accomplished, our proposed framework will provide early prediction of PASC pulmonary fibrosis and determine biomarkers to understand PASC better.   More Activities by Ulas Bagci, MSc 2   Sharath Bhagavatula, MD Brigham and Women's Hospital GE Healthcare/RSNA Research Scholar Grant (2023 - 2025) Development of Superparamagnetic Iron-Oxide Nanoparticle (SPIO-NP)-Enhanced Microwave Ablation   Abstract:   Percutaneous microwave ablation (MWA) is used to focally treat cancers. However, it generates indiscriminate killing of tumor and non-tumor tissues, which can result in severe complications or incomplete treatment. Here we propose using superparamagnetic iron-oxide nanoparticles (SPIO-NPs) to ‘prime’ tumors for more selective and effective ablation. SPIO-NPs are approved as MRI contrast agents (e.g., Feraheme) and have been shown to enhance microwave energy absorption/heating. They can be injected percutaneously or delivered intravenously with prolonged blood-pool phase, during which high tumor-to-background tissue uptake can be achieved. They have been shown to promote immunogenic cell death, and could enhance MWA anti-tumoral immune effects. Further work is needed to investigate SPIO-NP/MWA thermal and immune synergy, including optimizing intra-tumoral SPIO-NP delivery and conducting animal studies. Our proposal will address these next steps in murine tumor models: In Aim 1, we will identify an optimal SPIO-NP delivery strategy that maximizes tumor-to-background uptake in-vivo. In Aim 2, we will systematically test whether optimally delivered SPIO-NPs enhance MWA heating/killing of tumors relative to common background tissues, using MR thermometry and histopathology. In Aim 3, we will use a powerful, recently-developed spatial analysis method to comprehensively evaluate local and distant (abscopal) MWA-induced immune effects with and without SPIO-NPs. In this study, we expect to demonstrate feasibility of SPIO-NPs to improve MWA by providing thermal and immune synergy. We also expect to generate preliminary data for more ambitious long-term proposals. For example, 1-as SPIO-NPs and MWA are FDA-approved, successful results would support a clinical trial; 2-more tailored SPIO-NP formulations and MWA systems can be developed; and 3-Aim 3 could generate new mechanistic insights into how MWA can be better utilized to enhance immunotherapies. Our team has multidisciplinary expertise in MWA, tumor biology, immuno-oncology, nanomedicine, and MR physics, with the infrastructure to realize the full potential of this treatment paradigm.   More Activities by Sharath Bhagavatula, MD 3   Pratheek Bobba, BS Yale University RSNA Research Medical Student Grant (2023 - 2024) The Impact of Postnatal Growth on Brain White Matter Microstructure in Very Preterm Infants   Abstract:   Background: Infants born very preterm (<32 postmenstrual weeks gestation) represent approximately 1.6% of live births in the United States. Up to half of these very preterm infants (VPI) suffer from long-term cognitive delay. While birth weight and post-natal growth have been reported to correlate with neurodevelopmental outcome, the neurobiology and underlying mechanism of these associations remain elusive. Diffusion tensor imaging (DTI) can help delineate the microstructural development of neonatal brains and fill in this knowledge gap. In this project, we aim to characterize how birth weight and post-natal growth are associated with brain microstructural development in VPIs. Methods: Utilizing a cohort of n=246 neonates who had DTI as part of near-term pre discharge brain MRI, we will include neonates with (1) structurally normal discharge MRI, (2) normal neurological exam, and (3) acceptable DTI scan quality. We aim to (a) determine the relationship of birth weight with white matter (WM) microstructural development pattern, and (b) delineate the impact of postnatal growth on WM microstructure independent of VPI’s weight at time of birth. We will apply both voxel-wise tract-base statistical analysis and tract-based multivariate model to test our hypotheses, adjusting for gestational age at birth as well as at the time of scan. Clinical Significance: The information gained from this study will further our knowledge of the neurobiological mechanisms linking birth weight and postnatal growth to early microstructural development in VPIs. Notably, postnatal growth (and nutrition) are modifiable factors which can serve as treatment targets for improving long-term neurodevelopmental outcomes in VPIs. Our study will characterize imaging markers of brain microstructural development in response to post-natal growth, which can provide early feedback regarding the efficacy of nutritional interventions for improving long-term outcomes in this vulnerable population.   More Activities by Pratheek Bobba, BS 4   Maxime Bouthillier, MD University of Montreal RSNA Presidents Circle Research Resident Grant (2023 - 2024) MRI Biomarkers in Spinal Cord Injury: End-to-end Analysis Pipeline for Traumatic Spinal Cord Injury Research   Abstract:   Context - Traumatic spinal cord injury (tSCI) can lead to devastating neurological impairment, and long term reduced quality of life. Currently, no studies in neuroprotective treatment in the context of tSCI have identified a successful therapy to significantly and reliably improve neurological outcomes. However, improving the initial classification of the disease is an area of active research because it could lead to better identifying patients most likely to benefit from investigational therapies in the next decade. Long-term goals - 1) To develop an end-to-end and fully automated spinal cord magnetic resonance imaging (MRI) analysis pipeline able to improve the initial severity classification of tSCI, and accordingly predict the long-term outcomes of tSCI patients; 2) To identify novel spinal cord MRI quantitative biomarkers and develop accurate prediction models of long-term tSCI patients outcomes; 3) To assess the generalizability in the clinical setting of such image-based biomarkers and prediction models using multicenter cohorts. Methods - We will use a recent pan-Canadian multi-site MRI database of tSCI patients in collaboration with the Praxis Institute. We will develop an image processing pipeline to extract quantitative MRI imaging biomarkers of tSCI. These biomarkers will be incorporated into a prediction model that, together with other non-imaging clinical scores, can i) assess the initial severity of tSCI in comparison to the International Standards for Neurological Classification for Spinal Cord Injury (ISNCSCI) and ii) predict American Spinal Injury Association Impairment Scale (AIS) grade at hospital discharge. Impact - This project will fill a gap in the current management of tSCI patients by developing an automated quantitative image-based prediction model that will be widely available. Our approach mitigates overfitting issues that could hamper clinical implementation and acceptance. Knowledge derived from our work will benefit similar conditions characterized by spinal cord injuries such as spondylotic myelopathy and multiple sclerosis.   More Activities by Maxime Bouthillier, MD 5   Jason Chiang, MD, PhD University of California, Los Angeles RSNA Research Scholar Grant (2023 - 2025) Augmenting Microwave Ablation With Supercharged NK Cell Therapy in an Oncopig Liver Tumor Model   Abstract:   Locoregional therapy, using image-guided ablation or transarterial therapy, is the treatment of choice for early to intermediate stage hepatocellular carcinoma (HCC). However, these treatments can sometimes be associated with higher rates of recurrence or local tumor progression, especially for HCCs larger than 3 cm in diameter. Immunotherapy has attracted significant attention recently as it stimulates the patient’s own immune system to recognize the tumor cells as foreign and target them for destruction. Immunotherapy has recently been adopted into standard-of-care guidelines for advanced HCC. However, the efficacy of immunotherapy for early to intermediate stage HCC remains unknown, with mixed results in pre-clinical and clinical investigations. Natural killer (NK) cells are a type of immune cell that uniquely do not rely on antigen presentation and can be leveraged to kill tumor cells upon contact. Our lab uses a new technique that allows for rapid isolation and expansion of NK cells. These “supercharged” NK cells are able to kill tumor cells at higher levels compared to normal NK cells. The goal of this project is to evaluate the feasibility of using adjuvant transcatheter-directed supercharged NK cells to augment the cytotoxic profile of microwave ablation in a novel Oncopig liver tumor model. We will be taking a step-wise approach to evaluate the feasibility of using supercharged NK cells in combination with neoadjuvant locoregional therapy to treat an in-vivo Oncopig liver tumor model. The first aim will focus on isolating NK cells from porcine PBMCs and creating sNK cells. These porcine-derived sNK cells will be tested on an Oncopig liver tumor cell line. The second aim will look at using neo-adjuvant locoregional therapy to augment the cytotoxicity of sNK cells in an immunocompetent Oncopig liver tumor model.   More Activities by Jason Chiang, MD, PhD 6   Abigail Corkum University of Maryland RSNA Research Medical Student Grant (2023 - 2024) CT-Based Computer Aided Detection (CAD) Tools for Traumatic Torso Hemorrhage: Assessing Performance Bias and Hidden Stratification   Abstract:   Background: Computed Tomography (CT) has emerged as a routine screening modality in treating trauma patients. Algorithms are emerging that detect, grade, and quantify hemorrhage. Objectives: Our study is designed to test for hidden stratification and bias in CT-based hemorrhage CAD algorithms, but this first requires determining baseline incidences of torso-related hemorrhage injuries across patient demographics. We will then evaluate the performance of computer-aided detection (CAD) tools used to visualize torso hemorrhage. Methods: In this retrospective study, 15 years of data from the National Trauma Data Bank and the American College of Surgeons tqip database will be used to calculate overall incidence rates of injury type and mechanism across age, race, and geographic location (urban versus rural). Data for our second objective will include 5,000 trauma patients from an existing dataset of which 1,500 have had varying degrees of torso hemorrhage burden. Outcome variables of imaging performance will include area under the curve (AUC), sensitivity, specificity, negative and positive predictive values (PPV, NPV), dice similarity coefficients (DSC), and volume similarity index (VSI). Statistical Analysis: Chi-square and Fisher exact tests will be used to compare baseline incidences of traumatic injury across subgroups as well as differences in algorithm sensitivity, specificity, PPV and NPV. Differences in AUC will be evaluated using the DeLong method. Differences in mean DSC and VSI compared to ground truth annotations will be assessed using the Shapiro-Wilk test. Differences in means will be evaluated using the Mann-Whitney U test (non-normal distributions) and a two-tailed t-test (normal distributions). Results will be reported using a significance threshold of p=0.05. Significance: Results from this study will provide invaluable baseline descriptive statistics for evaluating traumatic hemorrhage-related CAD robustness and generalizability.   More Activities by Abigail Corkum 7   Farouk Dako, MD, MPH University of Pennsylvania Emerging Issues Grant - Healthcare Disparities (2023 - 2024) Community Support Program for Lung Cancer Screening   Abstract:   Lung cancer screening (LCS) using low-dose CT (LDCT) for high-risk individuals is a critical public health tool for reducing lung cancer mortality. Despite the recent expansion of the inclusion criteria to be more equitable, uptake and adherence remains low and is lowest among low-income and racial/ethnic minority populations and in individuals with a negative baseline screening study (Lung-RADS 1&2). Community-based programs promoting cancer screening uptake have demonstrated increased screening engagement compared to clinic-based programs in reaching disadvantaged individuals but suffer limited data demonstrating impact on a population level. Community-based participatory research (CBPR) utilizes a partnership approach, bringing together researchers and community stakeholders as equal partners throughout the research process to contribute expertise and share in decision making. CBPR has demonstrated the ability to reduce cancer health disparities and reduce mistrust between academia and surrounding communities through reciprocal learning. The overall objective of this project is to demonstrate the impact of a community support program (CSP) on adherence to LCS follow-up in an urban environment. It utilizes novel population level data that includes geospatial analysis of Philadelphia neighborhood-level data of patients who have received a LDCT for LCS. This allows for targeted community level interventions and the ability to evaluate impact on a population level. We will target individuals in our healthcare system residing in Philadelphia with a negative baseline screening CT and an upcoming or missed follow-up screening CT. Participants will be divide, based on residential location, into receiving clinic + community support vs routine clinic only (control). Primary analysis will measure the 12-month follow-up LDCT adherence rate difference between clinic + CSP vs clinic only group and its association with CSP enrollment. Planning and evaluation of our program will be performed using RE-AIM framework with emphasis on measuring reach, effectiveness, and sustainability of our approach.   More Activities by Farouk Dako, MD, MPH 8   Laurent Dercle, MD Columbia University Canon Medical Systems, USA/RSNA Research Resident Grant (2023 - 2024) Prospective Deployment of Artificial Intelligence Software Using Standard-of-Care CT Scans To Predict the Efficacy of Cancer Treatments   Abstract:   Artificial intelligence algorithms trained on standard-of-care CT scan data from multicenter international clinical trials can enhance the role of imaging in guiding therapy for patients with a diagnosis of cancer. Our previous work showed that OmniSig, a radiomics signature developed using clinical trial data from thousands of patients with a diagnosis of advanced cancer, could predict overall survival (OS) on a range of therapies better than the standard method based on tumor diameter, Response Evaluation Criteria in Solid Tumors (RECIST 1.1). The long-term goal of the proposed work is to determine whether OmniSig can be generalized and translated to clinical practice. Three specific aims will guide our work: 1) Externally validate OmniSig in a new dataset of 11,384 patients with measurable metastatic disease with a range of disease types in clinical trials testing multiple therapy types. We expect to replicate prior findings by showing that OmniSig score is significantly associated with OS. 2) Establish the robustness of OmniSig to variation from imaging reconstruction settings. We expect to show that the association with OS is not compromised by the heterogeneity of the multicenter setting in factors known to influence radiomics by comparing groups with varying slice thicknesses (<2 mm, 2-5 mm, and >5 mm) and reconstruction algorithms (sharp, smooth). 3) Prospectively validate OmniSig. We will test the ability of OmniSig to identify cancer patients who will derive long-term benefits in a cohort of 100 patients included in Phase 1 trials at two external world-renowned cancer centers. Successful completion of these aims will demonstrate the potential clinical significance of OmniSig for all patients with advanced cancers treated with systemic therapies and will overcome a key barrier to moving our innovative radiomic signature out of the laboratory into routine clinical use by radiologists to improve clinical decision-making.   More Activities by Laurent Dercle, MD 9   Mike Dohopolski, MD The University of Texas Southwestern Medical Center RSNA Research Seed Grant (2023 - 2024) Identifying Residual or Recurrent Local Disease on Post-Treatment PET Imaging in Patients With Head and Neck Cancer Using Reliable and Explainable Deep Learning Models   Abstract:   There are approximately 54,000 new cases and 11,000 deaths attributed to head and neck cancer (HNC) annually. Many of these patients are treated with either surgical resection +/- chemoradiation (CRT) or definitive CRT. Unfortunately, as many as 15-50% of patients will experience a locoregional recurrence following treatment—the most significant contributor to mortality. Salvage treatment after a clinical recurrence has been shown to improve oncologic outcomes; however, even after salvage therapy, these patients still do worse compared to those who responded well to the initial treatment. Early intervention prior to an overt clinical recurrence would presumably improve survival, but current diagnostic tools are inadequate to successfully identify this population. PET/CT imaging is the preferred surveillance modality for patients with HNC following CRT. However, post-treatment PET imaging is difficult to interpret in the setting of significant post-treatment inflammation. Diagnostic uncertainty regarding the presence of residual disease leads to increased patient stress and surgical interventions seeking pathologic confirmation, which may lead to complications. A non-invasive technique using image-based biomarkers is desperately needed to complement the evaluation of these difficult cases. Imaged-based machine learning (ML) algorithms have been reported to "diagnose" cancers with astounding accuracy, comparable to respective experts. Yet, these models are not commonly used in clinical practice. This issue is multifold, but critical factors include prediction reliability, model explainability, and general real-time integration into clinical workflows. Our project will attempt to aid physician decision-making following the completion of CRT for HNC. We hypothesize that: (1) the use of ML will improve primary site residual disease identification on surveillance imaging following initial definitive CRT; (2) uncertainty estimation techniques will be able to identify reliable predictions; and (3) class activation maps and feature responsibility/importance values will identify appropriate features on imaging to highlight a model's “reasoning.”   More Activities by Mike Dohopolski, MD 10   Elizabeth George, MD University of California San Francisco RSNA Research Scholar Grant (2023 - 2025) Quantitative Susceptibility Mapping of Neonatal Cerebral Oxygenation to Predict Neurodevelopmental Outcomes in Congenital Heart Disease   Abstract:   Advances in the care of congenital heart disease (CHD) have improved survival and neurodevelopmental delay (ND) has emerged as a major morbidity. CHD causes altered oxygenation, contributing to abnormal brain development, greatly impacting quality of life. Quantifying cerebral oxygenation is critical for identifying high-risk children for early intervention and for devising neuroprotective strategies. This proposal evaluates quantitative susceptibility mapping (QSM) of the neonatal brain for assessing cerebral oxygenation due to its ability to measure paramagnetic deoxyhemoglobin vs. diamagnetic oxyhemoglobin. QSM has distinct advantages over existing techniques such as near infrared spectroscopy (NIRS), T2* mapping and MR susceptometry in providing regional tissue oxygenation. However, limited availability of QSM has hindered its validation for this important clinical application. Our team is uniquely positioned to perform this work given access to data from an ongoing longitudinal study of children with critical CHD (n=100) and institutional expertise in QSM. We hypothesize: 1) Susceptibility (?) is a quantitative and sensitive metric of tissue oxygenation. We will test this by correlating bifrontal ? with NIRS-derived oxygenation pre-surgery (SA1A); and by comparing the change in regional ? pre- and post-surgery among those with transposition of great arteries (who undergo definitive surgery with normalized oxygenation) vs. single ventricle physiology (who undergo palliative surgery with persistent hypoxia, SA1B). 2) Susceptibility is associated with macrostructure and ND. We will test this by comparing mean white matter ? among regions with and without white matter injury and by correlating regional ? with regional brain volume on preoperative MRI (SA2A); and by establishing the association of regional ? with short term neurodevelopmental outcomes at 12-18 and/or 30 months (SA2B). Our long-term goal is to validate imaging-based metrics of the pathophysiology of brain oxygenation in CHD which are critical in early identification of high-risk children and for systematic assessment of strategies for neuroprotection.   More Activities by Elizabeth George, MD              Page 1 of 8 (1 - 10)