A5101966233- Radiomics and Machine Learning in Medical Imaging
- Head and Neck Cancer Studies
- Medical Imaging Techniques and Applications
- Single-cell and spatial transcriptomics
- Liver Disease Diagnosis and Treatment
- Dysphagia Assessment and Management
- Cell Image Analysis Techniques
- AI in cancer detection
- Liver Diseases and Immunity
- Bioinformatics and Genomic Networks
- Oral health in cancer treatment
- MRI in cancer diagnosis
- Pancreatic and Hepatic Oncology Research
- Genomics and Phylogenetic Studies
- Immunotherapy and Immune Responses
- Computational Drug Discovery Methods
- Sparse and Compressive Sensing Techniques
- Cutaneous Melanoma Detection and Management
- Sarcoma Diagnosis and Treatment
- Statistical Methods and Inference
- Advanced Radiotherapy Techniques
- Microbial Natural Products and Biosynthesis
- Radiopharmaceutical Chemistry and Applications
- Gene expression and cancer classification
- COVID-19 Impact on Reproduction
University of Michigan
2006-2024
Rice University
2018-2023
The University of Texas MD Anderson Cancer Center
2016-2019
First-order radiomic features, such as metabolic tumor volume (MTV) and total lesion glycolysis (TLG), are associated with disease progression in early-stage classical Hodgkin lymphoma (HL). We hypothesized that a model incorporating first- second-order features would more accurately predict outcome than MTV or TLG alone. assessed whether extracted from baseline PET scans predicted relapsed refractory status cohort of 251 patients stage I-II HL who were managed at tertiary cancer center....
Accurate clinical target volume (CTV) delineation is essential to ensure proper tumor coverage in radiation therapy. This a particularly difficult task for head-and-neck cancer patients where detailed knowledge of the pathways microscopic spread necessary. paper proposes solution auto-segment these volumes oropharyngeal using two-channel 3D U-Net architecture. The first channel feeds network with patient's CT image providing anatomical context, whereas second provides location and...
Spatial transcriptomics enables the measurement of mRNA counts at spatial locations within a tissue but faces challenges such as high experimental costs, technical expertise requirements, and low RNA detection efficiency resolution. We present STING (Spatial Transcriptomics Inference using Graph neural networks), computational approach that infers gene expression patterns with efficiency. integrates convolutional networks (CNNs) pre-trained on histological images graph (GNNs) to model...
Abstract Purpose: Merkel cell carcinoma (MCC) is an aggressive cutaneous malignancy whose pathogenesis and prognosis are related to the integrity of host immune system. Despite promising clinical responses immune-checkpoint blockade, response resistance remain unpredictable, underscoring a critical need delineate novel prognostic biomarkers and/or therapeutic targets for this disease. Experimental Design: Expression immune-regulatory markers (PD-L2, B7-H3, B7-H4, IDO-1, ICOS, TIM3, LAG3,...
Osteoradionecrosis (ORN) is a major side-effect of radiation therapy in oropharyngeal cancer (OPC) patients. In this study, we demonstrate that early prediction ORN possible by analyzing the temporal evolution mandibular subvolumes receiving radiation. For our analysis, use computed tomography (CT) scans from 21 OPC patients treated with Intensity Modulated Radiation Therapy (IMRT) subsequent radiographically-proven ≥ grade II ORN, at three different time points: pre-IMRT, 2-months, and...
The systematic inference of biologically relevant influence networks remains a challenging problem in computational biology. Even though the availability high-throughput data has enabled us to use probabilistic models infer plausible structure such networks, their true interpretation biology process is questionable. In this work, we propose network methodology, based on Directed information criterion, which incorporates transcription within framework, so as enable experimentally verifiable...
The development of osteosarcoma therapeutics has been challenging, in part because the lack appropriate criteria to evaluate responses. We developed a novel clinical trial radium-223 dichloride (223RaCl2) for response assessment osteosarcoma, NAFCIST (Na18F PET Criteria Solid Tumors).Patients received one six cycles 223RaCl2, and cumulative doses varied from 6.84 MBq 57.81 MBq. Molecular imaging with technetium-99m phosphonate scintigraphy, fluorine-18-fluorodeoxyglucose (18FDG) positron...
Background & AimsClinical trials for reducing fibrosis in steatotic liver disease (SLD) have targeted macrophages with variable results. We evaluated intrahepatic patients SLD to determine if activity scores or stages influenced phenotypes and expression of druggable targets, like CCR2 galectin-3.Methods ResultsLiver biopsies from controls minimal advanced (n=30) were analyzed differences macrophage-related genes. Several genes, including those known be pro-fibrotic (e.g., CD206, TREM2,...
Introduction: The acquisition of high-resolution digital pathology imaging data has sparked the development methods to extract context-specific features from such complex data. In context cancer, this led increased exploration tumor microenvironment with respect presence and spatial composition immune cells. Spatial statistical modeling may yield insights into role played by system in natural cancer as well downstream therapeutic interventions. Methods: paper, we present SPatial Analysis...
DDAP is a tool for predicting the biosynthetic pathways of products type I modular polyketide synthase (PKS) with focus on providing more accurate prediction ordering proteins and substrates in pathway. In this study, module docking domain (DD) affinity performance hold-out testing dataset reached 0.88 as measured by area under receiver operating characteristic (ROC) curve (AUC); Mean Reciprocal Ranking (MRR) pathway 0.67. has advantages compared to previous informatics tools several...
Purpose Head and neck cancers (HNC) radiotherapy (RT) is associated with inevitable injury to parotid glands subsequent xerostomia. We investigated the utility of standardized uptake values (SUV) derived from routinely performed 18-fluorodeoxygluocose positron-emission tomography (18FDG-PET) develop metabolic imaging biomarkers (MIBs) RT-related injury. Methods Data for oropharyngeal cancer (OPC) patients treated RT at our institution between 2005-2015 available planning computed (CT), dose...
In recent years, protein kinases have become some of the most significant drug targets in cancer patients. Kinases are known to regulate activity many human proteins, and consequently their inhibition has been used control proliferation. A challenge discovery is rapid efficient identification new small molecules. this study, we propose a novel silico approach identify kinase that impinge on nuclear receptor signaling with data generated using high-content analysis (HCA). high-throughput...
Early detection of Pancreatic Ductal Adenocarcinoma (PDAC), one the most aggressive malignancies pancreas, is crucial to avoid metastatic spread other body regions. Detection pancreatic cancer typically carried out by assessing distribution and arrangement tumor immune cells in histology images. This further complicated due morphological similarities with chronic pancreatitis (CP), co-occurrence precursor lesions same tissue. Most current automated methods for grading cancers rely on...
Abstract Pancreatic Ductal Adenocarcinoma (PDAC) is one of the most lethal cancers, with a 5-year survival rate only 13%. PDAC develops from precursor lesions, commonly Intraepithelial Neoplasia (PanIN). Studying PanIN in human patients difficult due to its microscopic nature and lack indication for sampling pancreas absence disease. In collaboration Gift Life - Michigan, local organ tissue donor organization, we obtained over 80 pancreata research. Surprisingly, found vast majority organs,...
Recent advances in multiplex imaging have enabled researchers to locate different types of cells within a tissue sample. This is especially relevant for tumor immunology, as clinical regimes corresponding stages disease or responses treatment may manifest spatial arrangements and immune cells. Spatial point pattern modeling can be used partition images according these regimes. To this end, we propose two-stage approach: first, local intensities pair correlation functions are estimated from...
Discrimination of pseudoprogression and true progression is one challenge to the treatment malignant gliomas. Although some techniques such as circulating tumor DNA (ctDNA) perfusion-weighted imaging (PWI) demonstrate promise in distinguishing PsP from TP, we investigate robust replicable alternatives distinguish two entities based on more widely-available media. In this study, use low-parametric supervised learning geographically-weighted regression (GWR) utility both conventional MRI...
Head and neck cancers radiotherapy (RT) is associated with inevitable injury to parotid glands subsequent xerostomia. We investigated the utility of SUV derived from 18FDG-PET develop metabolic imaging biomarkers (MIBs) RT-related injury.Data for oropharyngeal cancer (OPC) patients treated RT at our institution between 2005 2015 available planning computed tomography (CT), dose grid, pre- & first post-RT 18FDG-PET-CT scans, physician-reported xerostomia assessment 3-6 months (Xero ms) per...
We consider the problem of performing matrix completion with side information on row-by-row and column-by-column similarities. build upon recent proposals for estimation smoothness constraints respect to row column graphs. present a novel iterative procedure directly minimizing an criterion in order select appropriate amount smoothing, namely perform model selection. also discuss how exploit special structure scale up selection via Hutchinson estimator. simulation results application...
Abstract Osteoradionecrosis (ORN) is a major side-effect of radiation therapy in oropharyngeal cancer (OPC) patients. In this study, we demonstrate that early prediction ORN possible by analyzing the temporal evolution mandibular subvolumes receiving radiation. For our analysis, use computed tomography (CT) scans from 21 OPC patients treated with Intensity Modulated Radiation Therapy (IMRT) subsequent radiographically-proven ≥ grade II ORN, at three different time points: pre-IMRT, 2-months,...
ABSTRACT Intrahepatic macrophages influence the composition of microenvironment, host immune response to liver injury, and development fibrosis. Compared stellate cells, role intrahepatic in fibrosis remains ill defined. Multispectral imaging allows detection multiple markers situ human formalin-fixed, paraffin-embedded tissue. This cutting-edge technology is ideal for analyzing tissues since it spectral unmixing fluorophore signals, subtraction auto-fluorescence, preserves architecture vivo...