A5021933953- Pancreatic and Hepatic Oncology Research
- Gastric Cancer Management and Outcomes
- Esophageal and GI Pathology
- Machine Learning and Data Classification
- Domain Adaptation and Few-Shot Learning
- Gastrointestinal disorders and treatments
- Pancreatitis Pathology and Treatment
- Esophageal Cancer Research and Treatment
- Gastrointestinal Tumor Research and Treatment
- Gallbladder and Bile Duct Disorders
- COVID-19 diagnosis using AI
- Radiomics and Machine Learning in Medical Imaging
- Colorectal Cancer Screening and Detection
- Neuroendocrine Tumor Research Advances
- Multimodal Machine Learning Applications
- Medical Imaging Techniques and Applications
- Cholangiocarcinoma and Gallbladder Cancer Studies
- Organ Transplantation Techniques and Outcomes
- Water Systems and Optimization
- MRI in cancer diagnosis
- Liver Disease Diagnosis and Treatment
- Tracheal and airway disorders
- Gastrointestinal Bleeding Diagnosis and Treatment
- Anesthesia and Sedative Agents
- Eosinophilic Esophagitis
Siemens Healthcare (Germany)
2024
Mayo Clinic in Arizona
2011-2024
University of Surrey
2024
The University of Adelaide
2017-2023
Purdue University West Lafayette
2023
Australian Centre for Robotic Vision
2019-2023
Tufts University
2021
Hospital of the University of Pennsylvania
2017
Mayo Clinic in Florida
2017
Mayo Clinic
2003-2015
A new approach to optical fiber sensing is proposed and demonstrated that allows for specific measurement even in the presence of strong noise from undesired environmental perturbations. deep neural network model trained statistically learn relation complex interference output a multimode (MMF) with respect measurand interest while discriminating noise. This technique negates need carefully shield against, or compensate for, perturbations, as often case traditional sensors. achieved entirely...
Noisy labels are unavoidable yet troublesome in the ecosystem of deep learning because models can easily overfit them. There many types label noise, such as symmetric, asymmetric and instance-dependent noise (IDN), with IDN being only type that depends on image information. Such dependence information makes a critical to study, given labelling mistakes caused large part by insufficient or ambiguous about visual classes present images. Aiming provide an effective technique address IDN, we new...
Dysplasia in a Barrett's esophagus (BE) is associated with an increased risk for developing esophageal adenocarcinoma. Ablation using the HALO system has shown promise treatment of BE dysplasia. The objective this study was to assess safety and efficacy stepwise regimen circumferential focal ablation dysplasia.BE patients low-grade dysplasia (LGD) or high-grade (HGD) were enrolled. Primary followed every 3 months by further until complete endoscopic eradication achieved. At 3- 6-month...
The optimal interval of imaging studies for surveillance incidental pancreatic cystic neoplasms is not known.A retrospective analysis longitudinal medical records patients with was performed to examine the natural history incidentally detected respect development significant growth and identify predictors such growth.After excluding small (<10 mm) cysts (N = 144) inadequate clinical follow-up less than 6 months 79) those a diagnosis pseudocysts, serous cystadenoma, main duct intraductal...
We introduce a new, rigorously-formulated Bayesian meta-learning algorithm that learns probability distribution of model parameter prior for few-shot learning. The proposed employs gradient-based variational inference to infer the posterior parameters new task. Our can be applied any architecture and implemented in various machine learning paradigms, including regression classification. show models trained with our are well calibrated accurate, state-of-the-art calibration classification...
Robust training with noisy labels is a critical challenge in image classification, offering the potential to reduce reliance on costly clean-label datasets. Real-world datasets often contain mix of in-distribution (ID) and out-of-distribution (OOD) instance-dependent label noise, that rarely addressed simultaneously by existing methods further compounded lack comprehensive benchmarking Furthermore, even though current noisy-label learning approaches attempt find samples during training,...
Flumazenil is a competitive benzodiazepine antagonist that acts to reverse their sedative and hypnotic effects. It indicated in the management of overdose, but its role routine reversal endoscopic conscious sedation has not been defined.Patients undergoing diagnostic upper endoscopy who received with either diazepam or midazolam alone were given flumazenil 0.2 mg incrementally immediately following procedure until awake. They then asked repeat three psychomotor tests measuring cognitive...