• Login
    View Item 
    •   UMB Digital Archive
    • UMB Open Access Articles
    • UMB Open Access Articles
    • View Item
    •   UMB Digital Archive
    • UMB Open Access Articles
    • UMB Open Access Articles
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    Browse

    All of UMB Digital ArchiveCommunitiesPublication DateAuthorsTitlesSubjectsThis CollectionPublication DateAuthorsTitlesSubjects

    My Account

    LoginRegister

    Statistics

    Display statistics

    Automatic hemorrhage detection from color Doppler ultrasound using a Generative Adversarial Network (GAN)-based anomaly detection method

    • CSV
    • RefMan
    • EndNote
    • BibTex
    • RefWorks
    Author
    Mitra, Jhimli
    Qiu, Jianwei
    MacDonald, Michael
    Venugopal, Prem
    Wallace, Kirk
    Abdou, Hossam
    Richmond, Michael
    Elansary, Noha
    Edwards, Joseph
    Patel, Neerav
    Morrison, Jonathan
    Marinelli, Luca
    Show allShow less

    Date
    2022-01-01
    Journal
    IEEE Journal of Translational Engineering in Health and Medicine
    Publisher
    IEEE
    Type
    Article
    
    Metadata
    Show full item record
    See at
    https://doi.org/10.1109/JTEHM.2022.3199987
    Abstract
    Hemorrhage control has been identified as a priority focus area both for civilian and military populations in the United States because exsanguination is the most common cause of preventable death in hemorrhagic injury. Non-compressible torso hemorrhage (NCTH) has high mortality rate and there are currently no broadly available therapies for NCTH outside of a surgical room environment. Novel therapies, which include High Intensity Focused Ultrasound (HIFU) have emerged as promising methods for hemorrhage control as they can non-invasively cauterize bleeding tissue deep within the body without injuring uninvolved regions. A major challenge in the application of HIFU with color Doppler US guidance is the interpretation and optimization of the blood flow images in real-time to identify the hemorrhagic focus. Today, this task requires an expert sonographer, limiting the utility of this therapy in non-clinical environments. In this work, we investigated the feasibility of an automated hemorrhage detection method using a Generative Adversarial Network (GAN) for anomaly detection that learns a manifold of normal blood flow variability and subsequently identifies anomalous flow patterns that fall outside the learned manifold. As an initial feasibility study, we collected ultrasound color Doppler images of femoral arteries in an animal model of vascular injury (N = 11 pigs). Velocity information of the blood flow were extracted from the color Doppler images that were used for training and testing the anomaly detection network. Normotensive images from 8 pigs were used for training, and testing was performed on normotensive, immediately after injury, 10 minutes post-injury and 30 minutes post-injury images from 3 other pigs. The residual images or the reconstructed error maps show promise in detecting hemorrhages with an AUC of 0.90, 0.87, 0.62 immediately, 10 minutes post-injury and 30 minutes post-injury respectively with an overall AUC of 0.83.
    Keyword
    Anomaly detection
    color Doppler ultrasound
    deep learning
    Doppler effect
    generative adversarial network
    Generative adversarial networks
    hemorrhage detection
    Hemorrhaging
    Image color analysis
    Training
    Ultrasonic imaging
    unsupervised anomaly detection
    Show allShow less
    Identifier to cite or link to this item
    http://hdl.handle.net/10713/19687
    ae974a485f413a2113503eed53cd6c53
    10.1109/JTEHM.2022.3199987
    Scopus Count
    Collections
    UMB Open Access Articles

    entitlement

     

    Related items

    Showing items related by title, author, creator and subject.

    • Thumbnail

      Detection of Adulteration in Commercial Fruit Juices

      Wilson, Melinda K.; LaCourse, William R., 1957- (2011)
      Food safety and security remains an under investigated forensic problem. The US Department of Agriculture (USDA) and the Food and Drug Administration (FDA) examine close to 42 million imports per year, with a mere one percent being tested in a laboratory. These tests are often laborious, complicated, and time consuming with the consequence of economically motivated adulteration (EMA) products entering the consumer marketplace. EMA Is estimated by the Grocery Manufacturers Association to cost from $10-$15 billion per year, much of which is passed directly to the consumer. Fruit juice adulteration occurs when a more expensive product is diluted with a less expensive juice for an economic advantage. Acai juice is made from the berries of the acai palm, and is a major cash crop in the Amazon River region of South America. Due to the fragile nature of these berries, the expensive processing to create the juice, and the high importation costs, acai juice remains a target for adulteration by unscrupulous suppliers. The juice of these berries has a unique organic acid profile, which imparts the tartness of flavor and stability to the juice. This research uses high performance anion exchange chromatography followed by suppressed conductivity detection to profile the organic acids as a means to detect product authenticity. The linear range of this method spanned 0.2 μg/mL to 100 μg/mL for the nine investigated organic acids. The limit of detection for this assay was 0.1 μg/mL with a limit of quantitation of 0.2 μg/mL. Novel aspects of this research include the use of relative retention times to improve reproducibility, high throughput, and little to no sample preparation. This method allows for the establishment of the organic acid profile of a?ai juice as well as other juices. It also allows for the determination of the single strength, mixed juice, and adulterated juices. Importantly, this work can be expanded into other juices. The forensic utility of this method is highlighted by the successful determination of adulteration in a two juice blend.
    • Thumbnail

      Quantitative Detection of Cold Flow in Transdermal Drug Delivery System

      Zhao, Yuxiang; Hoag, Stephen W. (2014)
      Cold flow, by definition is the migration of adhesive beyond the boundaries of the transdermal drug delivery system (TDDS). It can result in poor transdermal product performance on the skin, decrease in the drug flux across the skin and undesirable drug exposure. These problems usually show up on storage and can only be detected by stability studies. However, once a transdermal product is approved, a company typically does not test the stability of every production batch. Hence, the first aim of this study was to investigate the relationship between the cold flow degree and the physical-mechanical properties of the TDDSs. The second aim was to investigate the feasibility of near infrared spectroscopy (NIRS) to quantitatively predict cold flow. For this study, Durotak87-4098® was selected as a model pressure sensitive adhesive (PSA). Triacetin and triethyl citrate were used as plasticizers, respectively, mixing with Durotak87-4098® to formulate the prototype transdermal products. Thermally induced cold flow studies, static/dynamic shear adhesion tests and a pressure induced cold flow study were conducted to understand the relationship between the product shear adhesion and cold flow degree. A near infrared (NIR) model was developed; shear adhesion testing data and the amount of induced cold flow were used as the reference values to calibrate the model. This study shows that plasticizer content in a PSA can affect the product shear adhesion; the product shear adhesion is highly related to the cold flow degree. The NIR data was analyzed by using a series of chemometric methods like principal components analysis (PCA) and partial least square regression (PLS) showing that the NIR spectroscopy is able to quantitatively predict the shear adhesion of the transdermal product at the time of production.
    • Thumbnail

      An improved algorithm of white matter hyperintensity detection in elderly adults

      Ding, T.; Cohen, A.D.; O'Connor, E.E. (Elsevier Inc., 2020)
      Automated segmentation of the aging brain raises significant challenges because of the prevalence, extent, and heterogeneity of white matter hyperintensities. White matter hyperintensities can be frequently identified in magnetic resonance imaging (MRI) scans of older individuals and among those who have Alzheimer's disease. We propose OASIS-AD, a method for automatic segmentation of white matter hyperintensities in older adults using structural brain MRIs. OASIS-AD is an approach evolved from OASIS, which was developed for automatic lesion segmentation in multiple sclerosis. OASIS-AD is a major refinement of OASIS that takes into account the specific challenges raised by white matter hyperintensities in Alzheimer's disease. In particular, OASIS-AD combines three processing steps: 1) using an eroding procedure on the skull stripped mask; 2) adding a nearest neighbor feature construction approach; and 3) applying a Gaussian filter to refine segmentation results, creating a novel process for WMH detection in aging population. We show that OASIS-AD performs better than existing automatic white matter hyperintensity segmentation approaches. Copyright 2020 The Authors
    DSpace software (copyright © 2002 - 2023)  DuraSpace
    Quick Guide | Policies | Contact Us | UMB Health Sciences & Human Services Library
    Open Repository is a service operated by 
    Atmire NV
     

    Export search results

    The export option will allow you to export the current search results of the entered query to a file. Different formats are available for download. To export the items, click on the button corresponding with the preferred download format.

    By default, clicking on the export buttons will result in a download of the allowed maximum amount of items.

    To select a subset of the search results, click "Selective Export" button and make a selection of the items you want to export. The amount of items that can be exported at once is similarly restricted as the full export.

    After making a selection, click one of the export format buttons. The amount of items that will be exported is indicated in the bubble next to export format.