A5070764516- Multiple Myeloma Research and Treatments
- Radiomics and Machine Learning in Medical Imaging
- AI in cancer detection
- Protein Degradation and Inhibitors
- Cancer Genomics and Diagnostics
- Melanoma and MAPK Pathways
- Electronic Health Records Systems
- Cancer Mechanisms and Therapy
- Peptidase Inhibition and Analysis
- Smoking Behavior and Cessation
- Biomedical Text Mining and Ontologies
- Health Systems, Economic Evaluations, Quality of Life
- Health, psychology, and well-being
- Generative Adversarial Networks and Image Synthesis
- Healthcare Systems and Technology
- Artificial Intelligence in Healthcare and Education
- Colorectal Cancer Screening and Detection
- Ethics in Clinical Research
- Social Media in Health Education
- Genomics and Rare Diseases
- Scarabaeidae Beetle Taxonomy and Biogeography
- Mental Health via Writing
- Medical Image Segmentation Techniques
- Genetic Neurodegenerative Diseases
- Colorectal Cancer Treatments and Studies
University of Arkansas for Medical Sciences
2015-2025
Winthrop Rockefeller Foundation
2020-2023
Wythenshawe Hospital
2020-2022
Manchester University NHS Foundation Trust
2021-2022
South West London and St George's Mental Health NHS Trust
2012-2017
Committee on Publication Ethics
2012
University of the West Indies
2010
Wilmington University
1998
Abbott (United States)
1998
Research Triangle Park Foundation
1998
Synthetic data generated by generative models can enhance the performance and capabilities of data-hungry deep learning in medical imaging. However, there is (1) limited availability (synthetic) datasets (2) are complex to train, which hinders their adoption research clinical applications. To reduce this entry barrier, we propose medigan, a one-stop shop for pretrained implemented as an open-source framework-agnostic Python library. medigan allows researchers developers create, increase,...
“Just Accepted” papers have undergone full peer review and been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, proof before it is published its final version. Please note that during production of the copyedited article, errors may be discovered which could affect content. Purpose To construct evaluate performance a machine learning model bone segmentation using whole-body CT images. Materials Methods In this retrospective...
Abstract As the COVID-19 pandemic unfolds, radiology imaging is playing an increasingly vital role in determining therapeutic options, patient management, and research directions. Publicly available data are essential to drive new into disease etiology, early detection, response therapy. In crisis, National Cancer Institute (NCI) has extended Imaging Archive (TCIA) include related images. Rural populations one population at risk for underrepresentation such public repositories. We have...
Copy-number changes and translocations have been studied extensively in many datasets with long-term follow-up. The impact of mutations remains debated given the short time to follow-up most datasets.We performed targeted panel sequencing covering 125 myeloma-specific genes loci involved 223 newly diagnosed myeloma samples recruited into one total therapy trials.As expected, commonly mutated were NRAS, KRAS, BRAF, making up 44% patients. Double-Hit BRAF DIS3 had an on outcome alongside...
Abstract We developed a DICOM dataset that can be used to evaluate the performance of de-identification algorithms. objects (a total 1,693 CT, MRI, PET, and digital X-ray images) were selected from datasets published in Cancer Imaging Archive (TCIA). Synthetic Protected Health Information (PHI) was generated inserted into Attributes mimic typical clinical imaging exams. The Standard TCIA curation audit logs guided insertion synthetic PHI standard non-standard data elements. A team tested...
Almost 40% of US adults provide informal caregiving, yet research gaps remain around what burdens affect caregivers. This study uses a novel social media site, Reddit, to mine and better understand online communities focus on as their caregiving burdens. These forums were accessed using an application programming interface, machine learning classifier was developed remove low information posts, topic modeling applied the corpus. An expert panel summarized forums’ themes into ten categories....
<h3>ABSTRACT</h3> <h3>Introduction</h3> Providing comprehensive tobacco addiction treatment to smokers admitted acute care settings represents an opportunity realise major health resource savings and population improvements. <h3>Methods</h3> The CURE project is a hospital-wide service piloted in Wythenshawe Hospital, Manchester, UK. core components of the are electronic screening all patients identify smokers; provision brief advice pharmacotherapy by frontline staff; opt-out referral...
Artificial intelligence (AI) is revolutionizing the field of medical imaging, holding potential to shift medicine from a reactive "sick-care" approach proactive focus on healthcare and prevention. The successful development AI in this domain relies access large, comprehensive, standardized real-world datasets that accurately represent diverse populations diseases. However, images data are sensitive, as such, before using them any way needs be modified protect privacy patients. This paper...
Direct extraction and use of electronic health record (EHR) data is a long-term multifaceted endeavor that includes design, development, implementation evaluation methods tools for semi-automating tasks in the research collection process, including, but not limited to, medical abstraction (MRA). A systematic mapping study elements was used to measure coverage Health Level Seven (HL7®) Fast Healthcare Interoperability Resources (FHIR®) standard federally sponsored, pragmatic cardiovascular...
AimsTo describe the characteristics of patients admitted over one year to two psychiatric intensive care units in large mental health trust. To establish number admissions, reasons for admission, incidents on PICU, bed days and destination after discharge. Differences gender these factors were explored. Secondary questions whether patient predicted aggressive or requiring long term care.MethodThe electronic notes admissions PICUs trust examined; sociodemographic clinical details recorded....
Treating tobacco dependency in patients admitted to acute care National Health Service (NHS) trusts is a key priority the NHS 10-year plan. This paper sets out results of health economic analysis for 'The CURE Project' pilot; new hospital-based service.A understand costs intervention (both inpatient service and postdischarge costs), return on investment (ROI) cost per quality-adjusted life year (QALY) Project pilot Greater Manchester. ROI QALY were calculated using European Study Quantifying...
The study of cancer genomics continually matures as the number patient samples sequenced increases. As more data is generated, oncogenic drivers for specific types are discovered along with their associated risks. This in turn leads to potential treatment strategies that pave way precision medicine. However, significant financial and analytical barriers make it infeasible sequence entire genome every patient. In contrast, targeted sequencing panels give reliable information on relevant...
ABSTRACT Background An unprecedented amount of personal health data, with the potential to revolutionise precision medicine, is generated at healthcare institutions worldwide. The exploitation such data using artificial intelligence relies on ability combine heterogeneous, multicentric, multimodal and multiparametric as well thoughtful representation knowledge availability. Despite these possibilities, significant methodological challenges ethico-legal constraints still impede real-world...
Abstract Objective This project demonstrates the feasibility of connecting medical imaging data and features, SARS‐CoV‐2 genome variants, with clinical in National Clinical Cohort Collaborative (N3C) repository to accelerate integrative research on detection, diagnosis, treatment COVID‐19‐related morbidities. The N3C curated a rich collection aggregated de‐identified electronic health records (EHR) over 18 million patients, including 7.5 COVID‐positive seen at hospitals across United States....
Abstract Purpose: De-identification of cancer imaging data is vitally important for sharing and the advancement research, however it a time consuming complex process that limits access to new sets such as those shared through NCI's Imaging Data Commons (IDC), built on Google Cloud Platform (GCP). Our research demonstrates how this can be automated using GCP-native services. Methods: We configured Medical Image De-Identification (MIDI) pipeline automate de-identification data. performed an...