A5004965117- Radiomics and Machine Learning in Medical Imaging
- AI in cancer detection
- Biomedical Text Mining and Ontologies
- Medical Imaging Techniques and Applications
- Retinal Imaging and Analysis
- Machine Learning in Healthcare
- Semantic Web and Ontologies
- MRI in cancer diagnosis
- Radiology practices and education
- Retinal Diseases and Treatments
- Lung Cancer Diagnosis and Treatment
- Artificial Intelligence in Healthcare and Education
- Digital Radiography and Breast Imaging
- Topic Modeling
- Colorectal Cancer Screening and Detection
- COVID-19 diagnosis using AI
- Advanced MRI Techniques and Applications
- Bioinformatics and Genomic Networks
- Medical Image Segmentation Techniques
- Glaucoma and retinal disorders
- Image Retrieval and Classification Techniques
- Advanced X-ray and CT Imaging
- Global Cancer Incidence and Screening
- Radiation Dose and Imaging
- Lung Cancer Treatments and Mutations
Stanford University
2015-2024
Emory University
2024
Stanford Medicine
2007-2023
Lucile Packard Children's Hospital
2021-2023
Palo Alto University
2011-2022
University of Illinois Chicago
2022
Intel (United States)
2020-2021
Association for the Advancement of Artificial Intelligence
2021
University of California, San Francisco
2021
St. Vincent's Birmingham
2020
Abstract Lung cancer is the most prevalent worldwide, and histopathological assessment indispensable for its diagnosis. However, human evaluation of pathology slides cannot accurately predict patients’ prognoses. In this study, we obtain 2,186 haematoxylin eosin stained histopathology whole-slide images lung adenocarcinoma squamous cell carcinoma patients from The Cancer Genome Atlas (TCGA), 294 additional Stanford Tissue Microarray (TMA) Database. We extract 9,879 quantitative image...
Biomedical ontologies provide essential domain knowledge to drive data integration, information retrieval, annotation, natural-language processing and decision support. BioPortal (http://bioportal.bioontology.org) is an open repository of biomedical that provides access via Web services browsers developed in OWL, RDF, OBO format Protégé frames. functionality includes the ability browse, search visualize ontologies. The interface also facilitates community-based participation evaluation...
Artificial intelligence (AI) continues to garner substantial interest in medical imaging. The potential applications are vast and include the entirety of imaging life cycle from image creation diagnosis outcome prediction. chief obstacles development clinical implementation AI algorithms availability sufficiently large, curated, representative training data that includes expert labeling (eg, annotations). Current supervised methods require a curation process for optimally train, validate,...
Abstract Published research results are difficult to replicate due the lack of a standard evaluation data set in area decision support systems mammography; most computer-aided diagnosis (CADx) and detection (CADe) algorithms for breast cancer mammography evaluated on private sets or unspecified subsets public databases. This causes an inability directly compare performance methods prior results. We seek resolve this substantial challenge by releasing updated standardized version Digital...
Purpose To conduct a comprehensive analysis of radiologist-made assessments glioblastoma (GBM) tumor size and composition by using community-developed controlled terminology magnetic resonance (MR) imaging visual features as they relate to genetic alterations, gene expression class, patient survival. Materials Methods Because all study patients had been previously deidentified the Cancer Genome Atlas (TCGA), publicly available data set that contains no linkage identifiers is HIPAA compliant,...
To identify prognostic imaging biomarkers in non-small cell lung cancer (NSCLC) by means of a radiogenomics strategy that integrates gene expression and medical images patients for whom survival outcomes are not available leveraging data public sets.A associating image features with clusters coexpressed genes (metagenes) was defined. First, correlation map is created pairwise association between metagenes. Next, predictive models metagenes built terms using sparse linear regression....
<h3>Importance</h3> Mammography screening currently relies on subjective human interpretation. Artificial intelligence (AI) advances could be used to increase mammography accuracy by reducing missed cancers and false positives. <h3>Objective</h3> To evaluate whether AI can overcome interpretation limitations with a rigorous, unbiased evaluation of machine learning algorithms. <h3>Design, Setting, Participants</h3> In this diagnostic study conducted between September 2016 November 2017, an...
Deep learning has become a promising approach for automated support clinical diagnosis. When medical data samples are limited, collaboration among multiple institutions is necessary to achieve high algorithm performance. However, sharing patient often limitations due technical, legal, or ethical concerns. In this study, we propose methods of distributing deep models as an attractive alternative data.We simulate the distribution across 4 using various training heuristics and compare results...
Quantitative imaging stratifies glioblastoma into three different phenotypes with distinct molecular activities independent of established markers and clinical status.
Abstract Medical image biomarkers of cancer promise improvements in patient care through advances precision medicine. Compared to genomic biomarkers, provide the advantages being non-invasive, and characterizing a heterogeneous tumor its entirety, as opposed limited tissue available via biopsy. We developed unique radiogenomic dataset from Non-Small Cell Lung Cancer (NSCLC) cohort 211 subjects. The comprises Computed Tomography (CT), Positron Emission (PET)/CT images, semantic annotations...
Purpose To assess the ability of convolutional neural networks (CNNs) to enable high-performance automated binary classification chest radiographs. Materials and Methods In a retrospective study, 216 431 frontal radiographs obtained between 1998 2012 were procured, along with associated text reports prospective label from attending radiologist. This data set was used train CNNs classify as normal or abnormal before evaluation on held-out 533 images hand-labeled by expert radiologists. The...
Purpose: To develop an automated method of localizing and discerning multiple types findings in retinal images using a limited set training data without hard-coded feature extraction as step toward generalizing these methods to rare disease detection which number are available. Methods: Two ophthalmologists verified 243 images, labeling important subsections the image generate 1324 patches containing either hemorrhages, microaneurysms, exudates, neovascularization, or normal-appearing...
Characterization of carotid plaque composition, more specifically the amount lipid core, fibrous tissue, and calcified is an important task for identification plaques that are prone to rupture, thus early risk estimation cardiovascular cerebrovascular events. Due its low costs wide availability, ultrasound has potential become modality choice characterization in clinical practice. However, significant image noise, coupled with small size their complex appearance, makes it difficult automated...