A5028041549- Liver Disease Diagnosis and Treatment
- Hepatocellular Carcinoma Treatment and Prognosis
- Liver Disease and Transplantation
- Ultrasound and Hyperthermia Applications
- Domain Adaptation and Few-Shot Learning
- Liver Diseases and Immunity
- Cardiovascular Disease and Adiposity
- Radiation Dose and Imaging
- Radiative Heat Transfer Studies
- Machine Learning in Materials Science
- Spectroscopy Techniques in Biomedical and Chemical Research
- Thyroid Cancer Diagnosis and Treatment
- Radiation Therapy and Dosimetry
- Machine Learning and Algorithms
- Advanced Neural Network Applications
- Brain Tumor Detection and Classification
University Hospital of Geneva
2022-2023
University of California, San Diego
2019-2022
University of Cincinnati Medical Center
1983
Background Advanced confounder-corrected chemical shift–encoded MRI-derived proton density fat fraction (PDFF) is a leading parameter for quantification in nonalcoholic fatty liver disease (NAFLD). Because of the limited availability this MRI technique, there need to develop and validate alternative parameters assess fat. Purpose To relationship quantitative US PDFF multivariable models detect hepatic steatosis quantify Materials Methods Adults with known NAFLD or who were suspected having...
A low-iodine diet was developed for used in decreasing iodine intake and excretion patients undergoing evaluation with radioactive I-131 ablation of thyroid remnants as treatment cancer. It has been demonstrated to effectively lower less than 25% basal values. Preliminary calculations suggest that such depletion may be potentially useful increasing the radiation dose per mCi administered activity ablative therapy.
Objectives To develop and evaluate deep learning models devised for liver fat assessment based on ultrasound (US) images acquired from four different views: transverse plane (hepatic veins at the confluence with inferior vena cava, right portal vein, posterior vein) sagittal (liver/kidney). Methods US (four separate views) were 135 participants known or suspected nonalcoholic fatty disease. Proton density fraction (PDFF) values derived chemical shift‐encoded magnetic resonance imaging served...
Abstract Objectives To compare the diagnostic accuracy of US shear wave elastography (SWE) and magnetic resonance (MRE) for classifying fibrosis stage in patients with nonalcoholic fatty liver disease (NAFLD). Methods Patients from a prospective single-center cohort clinical biopsy known or suspected NAFLD underwent contemporaneous SWE MRE. AUCs biopsy-determined stages ≥ 1, 2, 3, = 4, their respective performance parameters at cutoffs providing 90% sensitivity specificity were compared...
Background MRI-derived proton density fat fraction (PDFF) is an accurate, reliable, and safe biologic marker for use in the noninvasive diagnosis of hepatic steatosis patients with nonalcoholic fatty liver disease (NAFLD). Because cost limited availability MRI, it necessary to develop accurate method diagnose NAFLD potential point-of-care access. Purpose To compare diagnostic accuracy quantitative US (QUS) (FF) estimator that controlled attenuation parameter (CAP) using contemporaneous PDFF...
Aramchol, an oral stearoyl‐coenzyme‐A‐desaturase‐1 inhibitor, has been shown to reduce hepatic fat content in patients with primary nonalcoholic fatty liver disease (NAFLD); however, its effect human immunodeficiency virus (HIV)–associated NAFLD is unknown. The aramchol for HIV‐associated and lipodystrophy (ARRIVE) trial was a double‐blind, randomized, investigator‐initiated, placebo‐controlled test the efficacy of 12 weeks treatment versus placebo NAFLD. Fifty NAFLD, defined by magnetic...
This study focuses on assessing the performance of active learning techniques to train a brain MRI glioma segmentation model.The publicly available training dataset provided for 2021 RSNA-ASNR-MICCAI Brain Tumor Segmentation (BraTS) Challenge was used in this study, consisting 1251 multi-institutional, multi-parametric MR images. Post-contrast T1, T2, and T2 FLAIR images as well ground truth manual were input model. The data split into set 1151 cases testing 100 cases, with remaining...
Nonalcoholic fatty liver disease (NAFLD) affects ~25% of the world population. Confounder-corrected chemical-shift-encoded MRI-derived proton density fat fraction (MRI-PDFF) is an established quantitative noninvasive biomarker hepatic steatosis but has limited availability. There a clinical need for more practical and accessible methods to noninvasively assess steatosis. Previous work shown that two ultrasound (QUS) biomarkers - attenuation coefficient (AC) backscatter (BSC) are correlated...
We have developed quantitative ultrasound (QUS) and deep learning algorithms to estimate hepatic fat fraction from radiofrequency (RF) data backscattered by the liver. To facilitate translation of such for clinical care research, we a standalone software tool that can automatically generate estimates using each four separate based on ultrasonic attenuation coefficient (AC) (Algorithm 1), backscatter (BSC) 2), both AC BSC 3), with uncalibrated raw RF 4). Reference phantom sonographer-drawn...
Qualitative sonography is used to assess nonalcoholic fatty liver disease (NAFLD), an important health issue worldwide. We B-mode image deep-learning objectively NAFLD in 4 views of the (hepatic veins at confluence with inferior vena cava, right portal vein, posterior vein and liver/kidney) 135 patients known or suspected NAFLD. Transfer learning a deep convolutional neural network (CNN) was applied for quantifying fat fraction diagnosing (≥ 5%) using contemporaneous MRI-PDFF as ground...
Nonalcoholic fatty liver disease (NAFLD) is a leading cause of chronic worldwide. Current diagnosis relies on invasive biopsy. Quantitative ultrasound (QUS) may provide noninvasive alternative, using parameters such as attenuation (AC) and backscatter (BSC) coefficients. This study examines how exclusion larger blood vessels from fields interest (FOI) affects AC BSC measurements in NAFLD patient livers. 201 total participants with known or suspected underwent right lobe QUS. Of these, 84...
The objective of soft-tissue quantitative ultrasound (QUS) is to improve diagnostic imaging via outcomes. Over the past three or so decades, there have been an increasing number QUS successes. In a UIUC-UCSD collaboration, nonalcoholic fatty liver disease (NAFLD) assessed from seven biomarkers [AC, BSC, Lizzi-Feleppa markers (slope, intercept, midband), two envelope parameters (k and mu)] derived RF data shows dependencies with fat content in human subjects. 102 participants underwent exams...
Quantitative ultrasound (QUS) aims to improve diagnostic imaging by extracting objective tissue parameters from backscattered signals. Deep learning can facilitate this process because of its ability extract high-level information the raw data. We have demonstrated that deep approaches applied signals accurately quantify liver fat noninvasively. will discuss three herein: (1) a one-dimensional convolutional neural network (1-D-CNN) for uncalibrated radiofrequency (RF) data; (2)...
Reducing the amount of manually annotated ground truth data needed to train deep learning-assisted medical image segmentation is desirable reduce time and cost labor by trained experts. This work aimed assess capability active learning techniques tackle this issue starting with a smaller dataset updating it through model feedback that most beneficial training. Brain MRIs 1251 patients pathologically confirmed glioma diagnosis from 2021 RSNA-ASNRMICCAI BraTS Challenge were used in study. A...