Displaying all 2 publications

Abstract:
Sort:
  1. Jayaraman J, Roberts GJ, Wong HM, King NM
    BMC Med Imaging, 2018 04 27;18(1):5.
    PMID: 29703180 DOI: 10.1186/s12880-018-0250-z
    BACKGROUND: The accuracy of estimated age should depend on the reference data sets (RDS) from which the maturity scores or Ages of Attainment (AoA) were obtained. This study aimed to test the accuracy of age estimation from three different population specific dental reference datasets (RDS).

    METHODS: Two hundred and sixty six dental panoramic radiographs of subjects belonging to southern Chinese ethnicity were scored and dental age (DA) was estimated from three reference datasets: French-Canadian, United Kingdom (UK) Caucasian and southern Chinese. Statistical significance was set at p  0.05). The southern Chinese RDS estimated the age of 80% of subjects within ±12 months range, and 90% of subjects within ±18 months range (p 

  2. Isaksson LJ, Pepa M, Summers P, Zaffaroni M, Vincini MG, Corrao G, et al.
    BMC Med Imaging, 2023 Feb 11;23(1):32.
    PMID: 36774463 DOI: 10.1186/s12880-023-00974-y
    BACKGROUND: Contouring of anatomical regions is a crucial step in the medical workflow and is both time-consuming and prone to intra- and inter-observer variability. This study compares different strategies for automatic segmentation of the prostate in T2-weighted MRIs.

    METHODS: This study included 100 patients diagnosed with prostate adenocarcinoma who had undergone multi-parametric MRI and prostatectomy. From the T2-weighted MR images, ground truth segmentation masks were established by consensus from two expert radiologists. The prostate was then automatically contoured with six different methods: (1) a multi-atlas algorithm, (2) a proprietary algorithm in the Syngo.Via medical imaging software, and four deep learning models: (3) a V-net trained from scratch, (4) a pre-trained 2D U-net, (5) a GAN extension of the 2D U-net, and (6) a segmentation-adapted EfficientDet architecture. The resulting segmentations were compared and scored against the ground truth masks with one 70/30 and one 50/50 train/test data split. We also analyzed the association between segmentation performance and clinical variables.

    RESULTS: The best performing method was the adapted EfficientDet (model 6), achieving a mean Dice coefficient of 0.914, a mean absolute volume difference of 5.9%, a mean surface distance (MSD) of 1.93 pixels, and a mean 95th percentile Hausdorff distance of 3.77 pixels. The deep learning models were less prone to serious errors (0.854 minimum Dice and 4.02 maximum MSD), and no significant relationship was found between segmentation performance and clinical variables.

    CONCLUSIONS: Deep learning-based segmentation techniques can consistently achieve Dice coefficients of 0.9 or above with as few as 50 training patients, regardless of architectural archetype. The atlas-based and Syngo.via methods found in commercial clinical software performed significantly worse (0.855[Formula: see text]0.887 Dice).

Related Terms
Filters
Contact Us

Please provide feedback to Administrator (afdal@afpm.org.my)

External Links