Displaying all 3 publications

Abstract:
Sort:
  1. Using autoencoders for mammogram compression
    Tan CC, Eswaran C
    J Med Syst, 2011 Feb;35(1):49-58.
    PMID: 20703586 DOI: 10.1007/s10916-009-9340-3
    This paper presents the results obtained for medical image compression using autoencoder neural networks. Since mammograms (medical images) are usually of big sizes, training of autoencoders becomes extremely tedious and difficult if the whole image is used for training. We show in this paper that the autoencoders can be trained successfully by using image patches instead of the whole image. The compression performances of different types of autoencoders are compared based on two parameters, namely mean square error and structural similarity index. It is found from the experimental results that the autoencoder which does not use Restricted Boltzmann Machine pre-training yields better results than those which use this pre-training method.
    Matched MeSH terms: Radiographic Image Interpretation, Computer-Assisted/instrumentation
  2. Fulltext Image reconstruction of a charge coupled device based optical tomographic instrumentation system for particle sizing
    Idroas M, Rahim RA, Green RG, Ibrahim MN, Rahiman MH
    Sensors (Basel), 2010;10(10):9512-28.
    PMID: 22163423 DOI: 10.3390/s101009512
    This research investigates the use of charge coupled device (abbreviated as CCD) linear image sensors in an optical tomographic instrumentation system used for sizing particles. The measurement system, consisting of four CCD linear image sensors are configured around an octagonal shaped flow pipe for a four projections system is explained. The four linear image sensors provide 2,048 pixel imaging with a pixel size of 14 micron × 14 micron, hence constituting a high-resolution system. Image reconstruction for a four-projection optical tomography system is also discussed, where a simple optical model is used to relate attenuation due to variations in optical density, [R], within the measurement section. Expressed in matrix form this represents the forward problem in tomography [S] [R] = [M]. In practice, measurements [M] are used to estimate the optical density distribution by solving the inverse problem [R] = [S](-1)[M]. Direct inversion of the sensitivity matrix, [S], is not possible and two approximations are considered and compared-the transpose and the pseudo inverse sensitivity matrices.
    Matched MeSH terms: Image Interpretation, Computer-Assisted/instrumentation*
  3. A videoendoscopic evaluation of the upper airway in South East Asian adults with obstructive sleep apnea
    Abdullah B, Rajet KA, Abd Hamid SS, Mohammad WM
    Sleep Breath, 2011 Dec;15(4):747-54.
    PMID: 20957444 DOI: 10.1007/s11325-010-0431-7
    OBJECTIVES: We aimed to evaluate the severity of upper airway obstruction at the retropalatal and retroglossal regions in obstructive sleep apnea (OSA) patients.

    METHODOLOGY: This is a descriptive cross-sectional study at the Sleep Clinic, Department of Otorhinolaryngology-Head and Neck Surgery. Flexible nasopharyngolaryngoscopy was performed in seated erect and supine position. Retropalatal and retroglossal regions were continuously recorded during quiet breathing and Mueller's maneuver in both positions. Captured images were measured using Scion Image software and narrowing rate was calculated. Level of each site was classified based on Fujita classification and severity of obstruction using Sher scoring system for Mueller's maneuver.

    RESULTS: A total of 59 patients participated in this study. Twenty-nine (49.2%) participants had type 1 (retropalatal) obstruction, 23 (38.9%) had type 2 (retropalatal and retroglossal), and seven (11.9%) in type 3 (retroglossal) obstruction. Fifty (84.7%) of the patients have severe obstruction at the retropalatal region in supine position (SRP) followed by 35 (59.3%) at retropalatal region in erect position (ERP), 27 (45.8%) at retroglossal region in supine position (SRG) and eight (13.5%) at retroglossal region in erect position (ERG). The average oxygen saturation showed significant association in ERP (P = 0.012) and SRP (P < 0.001), but not significant in ERG and SRG.

    CONCLUSIONS: Videoendoscopy utilizing flexible nasopharyngolaryngoscopy and Scion Image software is reliable, minimally invasive, and useful as an office procedure in evaluating the multilevel obstruction of upper airway in OSA patients. The retropalatal region has more severe obstruction compared with retroglossal region either in erect or supine position.

    Matched MeSH terms: Image Interpretation, Computer-Assisted/instrumentation
Related Terms
    Try searching for something
Filters
Contact Us

Please provide feedback to Administrator ([email protected])

External Links