A5040601955- Advanced MRI Techniques and Applications
- MRI in cancer diagnosis
- Genetic and Kidney Cyst Diseases
- Radiomics and Machine Learning in Medical Imaging
- Brain Tumor Detection and Classification
- Advanced Neuroimaging Techniques and Applications
- Pediatric Urology and Nephrology Studies
- Fetal and Pediatric Neurological Disorders
- Liver Disease Diagnosis and Treatment
- Medical Imaging Techniques and Applications
- Bone and Joint Diseases
- Advanced X-ray and CT Imaging
- NMR spectroscopy and applications
- Cellular Mechanics and Interactions
- Multiple Sclerosis Research Studies
- 3D Printing in Biomedical Research
- Hepatocellular Carcinoma Treatment and Prognosis
- Diet and metabolism studies
- Atomic and Subatomic Physics Research
- Gene expression and cancer classification
- Voice and Speech Disorders
- Digital Imaging for Blood Diseases
- Brucella: diagnosis, epidemiology, treatment
- Cardiomyopathy and Myosin Studies
- Functional Brain Connectivity Studies
Cornell University
2022-2025
Weill Cornell Medicine
2022-2025
Cornell College
2023-2024
Hofstra University
2022
This study develops, validates, and deploys deep learning for automated total kidney volume (TKV) measurement (a marker of disease severity) on T2-weighted MRI studies autosomal dominant polycystic (ADPKD). The model was based the U-Net architecture with an EfficientNet encoder, developed using 213 abdominal in 129 patients ADPKD. Patients were randomly divided into 70% training, 15% validation, test sets development. Model performance assessed Dice similarity coefficient (DSC) Bland-Altman...
Background/Objectives: Accurate and reproducible spleen volume measurements are essential for assessing treatment response disease progression in myelofibrosis. This study evaluates techniques measuring on abdominal MRI. Methods: In 20 patients with bone marrow biopsy-proven myelofibrosis, 5 observers independently measured 3 MRI pulse sequences, 3D-spoiled gradient echo T1, axial single-shot fast spin (SSFSE) T2, coronal SSFSE using ellipsoidal approximation, manual contouring, 3D nnU-Net...
Organ volume measurements are a key metric for managing ADPKD (the most common inherited renal disease). However, measuring organ volumes is tedious and involves manually contouring outlines on multiple cross-sectional MRI or CT images. The automation of kidney using deep learning has been proposed, as it small errors compared to manual contouring. Here, deployed open-source segmentation pipeline extended also measure liver spleen volumes, which important. This 2D U-net approach was...
Total kidney volume (TKV) is an important biomarker for assessing function, especially autosomal dominant polycystic disease (ADPKD). However, TKV measurements from a single MRI pulse sequence have limited reproducibility, ± ~5%, similar to ADPKD annual growth rates.
To develop a tissue field-filtering algorithm, called maximum spherical mean value (mSMV), for reducing shadow artifacts in QSM of the brain without requiring brain-tissue erosion.
Abstract Background Autosomal dominant polycystic kidney disease (ADPKD) can lead to liver (PLD), characterized by cysts. Although majority of the patients are asymptomatic, massively enlarged secondary PLD cause discomfort, and compression on adjacent structures requiring cyst aspiration/fenestration, partial resection, or transplantation. Monitoring measuring volume fails track early stages when is too small affect volume. Purpose To improve assessment in automating detection segmentation...
Abdominal imaging of autosomal dominant polycystic kidney disease (ADPKD) has historically focused on detecting complications such as cyst rupture, infection, obstructing renal calculi, and pyelonephritis; discriminating complex cysts from cell carcinoma; identifying sources abdominal pain. Many features ADPKD are incompletely evaluated or not deemed to be clinically significant, because this, treatment options limited. However, total volume (TKV) measurement become important for assessing...
111
Multiple Sclerosis (MS) is an autoimmune disorder characterized by focal inflammatory demyelination. Chronic MS lesions can contain chronically activated, iron-laden microglia and macrophages. By comparing rim status on quantitative susceptibility mapping (QSM) phase imaging with histopathology that identifies iron, we demonstrate QSM a more reliable indicator of iron than phase. valuable clinical tool to identify positive smoldering not visible using conventional MRI techniques.
This study compared L1 and L2 regularized Quantitative Transport Mapping (QTM1-3) of dynamic contrast enhanced (DCE) MRI in breast neck tumor using image quality scoring. Improved consistent soft tissue lesion characterization was observed when the norm.
We have developed a methodology of preparing explant livers for perfusion and vessel size imaging. Off-label use ferumoxytol is known to provide high quality maps in vivo; however, adverse reactions may hinder adoption into the clinical workflow. In this case, gadolinium be more attractive. VSI experiments on perfused liver proof concept that Gadolinium used as an alternative contrast
Myelin Volume Fraction (MVF) is an important biomarker of demyelination various diseases Multiple Sclerosis. In this study, we propose a method that provides quantitative MVF maps from routine multiple echo gradient acquisitions and dictionary matching. The generated using the Hollow Cylindrical Fiber Model (HCFM) employs both magnitude signal decay QSM obtained mGRE phase. show qualitative superiority over standard multiexponential fitting-based myelin water fraction (MWF) maps.
Organ volume measurements on MRI are typically performed a single pulse sequence because of the tedious process manual contouring. Here we use deep learning to automate kidney segmentations so that can be measured five abdominal sequences. In 17 subjects scanned twice within 3 weeks (when no change in was expected), power averaging 5 improved reproducibility, achieving 2.5% absolute percent difference compared 5.9% with contouring (p<0.05). Absolute error reduced further 2.1%, p<0.05...
In chronic liver disease, fibrosis develops as excessive deposition of extracellular matrix macromolecules, predominantly collagens, progressively form fibrous scars that disrupt the hepatic architecture, and fibrosis, iron, fat are interrelated. Fibrosis is best predictor morbidity mortality in disease but biopsy, reference method for diagnosis staging, invasive limited by sampling interobserver variability risks complications. The overall objective this study was to develop a new...
Myelin quantification is used in the study of demyelination neurodegenerative diseases. A novel noninvasive MRI method, Microstructure-Informed Mapping (MIMM), proposed to quantify myelin volume fraction (MVF) from a routine multi-gradient echo sequence (mGRE) using multiscale biophysical signal model effects microstructural and iron. In MIMM, are modeled based on Hollow Cylinder Fiber Model accounting for anisotropy, while iron considered as an isotropic paramagnetic point source. This...
Motivation: To use fluid-mechanics based deep learning method to predict perfusion parameters from dynamic imagesGoal(s): We propose explore the possibility neural network trained on simulated data fluid mechanics simulation analyze medical images. Approach: quantitative transport mapping (QTMnet), which is concentration propagation profile generated constrained constructive optimization (CCO) and equation-based tracer simulation, including flow rate, permeability, vasculature volume, DCE...
Motivation: To assess the ability of quantitative transport mapping (QTM) to estimate blood flow in stroke from DSC MRI through a deep learning model. Goal(s): use an automated based method measure using MRI. Approach: A network (QTMnet) is trained on synthetic MR data generated realistic vascular models learn between and underlying tissue flow. Results: QTMnet demonstrates decreased perfusion ischemic lesion compared contralateral healthy (p=0.0006), similar results traditional modeling....
Motivation: In diffusion-weighted arterial spin labeling (DW-ASL) images, quantification of the water exchange rate $$$k_{w}$$$ uses a single-pass approximation (SPA) which introduces systematic error while fitting non-linear model is difficult. Goal(s): Our goal was to reduce blood-brain-barrier (BBB) ($$$k_{w}$$$) errors in DW-ASL images. Approach: We introduced biophysical-modeling-based deep learning method (QTMNet) and tested both simulated vivo data. Results: On data, QTMNet has 90%...
Parkinson's disease patients with motor complications are often considered for deep brain stimulation (DBS) surgery. DBS candidate selection involves an assessment known as the levodopa challenge test (LCT). The LCT aims to predict outcomes by measuring symptom improvement accompanying changes in dosage. While used patient process, inconsistent predictions have been widely documented, verified here Pearson's correlation r=0.12. Estimating separate responders and non-responders remains unmet...
Motivation: To improve outcome prediction for deep brain stimulation (DBS) surgery using radiomic features on quantitative susceptibility maps (QSMs). Goal(s): address the inconsistent levodopa challenge test (LCT) DBS outcomes by describing target variable, motor symptom improvement, as a weighted sum of QSM features. Approach: A least absolute shrinkage and selection operator (LASSO) model is implemented, trained, tested patient data known outcomes. Results: Model predictions outperform...
Motivation: To validate deep learning based Quantitative Transport Mapping (QTMnet) on a perfused tissue phantom. Goal(s): Evaluate the accuracy of QTMnet derived flow and compare to traditional tracer-kinetic estimation. Approach: We developed workflow prepare porcine liver as perfusion phantom1. n=8 livers with controllable pump acquired DCE-MRI. then estimated tracer-kinetics. Results: accurately estimates our phantom (mean error: -2.82%, mean absolute 10.0%). Furthermore, estimation was...
Motivation: Abdominal organ volumes are critical MRI biomarkers in many diseases including autosomal dominant polycystic kidney disease. Goal(s): We aim to develop a segmentation model with an enhanced ability generalize across various abdominal organs and MR pulse sequences. Approach: construct multi-modality foundation expanding upon our existing ADPKD which adapts diverse tissues minimal new training data. Results: The was trained using model-in-loop methodology evaluated against...