Search

Your search keyword '"Williams, L.J."' showing total 2 results
Start Over Author "Williams, L.J." Topic discriminant analysis
2 results on '"Williams, L.J."'

1. Repurposing a fracture risk calculator (FRAX) as a screening tool for women at risk for sarcopenia.

Author

Pasco, J.A., Mohebbi, M., Tembo, M.C., Holloway-Kew, K.L., Hyde, N.K., Williams, L.J., and Kotowicz, M.A.

Subjects
DIAGNOSIS of bone fractures, DISCRIMINANT analysis, BONE fractures, GRIP strength, HIP joint injuries, LONGITUDINAL method, OSTEOPOROSIS, PROBABILITY theory, RISK assessment, WOMEN'S health, SARCOPENIA, RECEIVER operating characteristic curves, LEAN body mass, PHOTON absorptiometry, DISEASE risk factors
Abstract

Summary: Osteoporosis and sarcopenia share risk profiles, so we tested a fracture risk assessment tool (FRAX) as a screening tool for sarcopenia. FRAX probabilities without bone mineral density predicted sarcopenia with high sensitivity and reasonable specificity. There is potential to use this FRAX as a screening tool for sarcopenia. Purpose: There is a need for simple screening tools for sarcopenia. As osteoporosis and sarcopenia share risk profiles, we tested the performance of a fracture risk assessment tool for discriminating individuals at risk for sarcopenia. Methods: In this longitudinal study, FRAX (Australia) probabilities were calculated for 354 women (ages 40–90 years) in the Geelong Osteoporosis Study. Sarcopenia was assessed a decade later using DXA-derived low appendicular lean mass (Lunar; ALM/height2 < 5.5 kg/m2) and low handgrip strength (Jamar; HGS < 16 kg), according to EWGSOP2. We determined FRAX probabilities (%) for hip fracture (HF-FRAX) and major osteoporotic fracture (MOF-FRAX), with and without BMD. Area under the receiver operator characteristic (AUROC) curves quantified the performance of FRAX for predicting sarcopenia. Results: Baseline median (IQR) values for HF-FRAX without BMD were 0.4 (0.1–1.3) and for MOF-FRAX without BMD, 2.4 (1.2–5.2); comparable figures for HF-FRAX with BMD were 0.2 (0.0–0.7) and for MOF-FRAX with BMD, 2.1 (1.1–4.4). At follow-up, sarcopenia was identified for 11 (3.1%) women. When FRAX was calculated without BMD, the AUROC was 0.90 for HF-FRAX and 0.88 for MOF-FRAX. Optimal thresholds were 0.9 for HF-FRAX (sensitivity 90.9%, specificity 62.4%) and 5.3 for MOF-FRAX (sensitivity 81.8%, specificity 71.7%). Calculating FRAX with BMD did not improve the predictive performance of FRAX for sarcopenia. Conclusion: Here we provide preliminary evidence to suggest that FRAX probabilities without BMD might predict sarcopenia with high sensitivity and reasonable specificity. Given that FRAX clinical risk factors are identified without equipment, there is potential to use this or a modified version of the FRAX tool to screen for individuals at risk of sarcopenia. [ABSTRACT FROM AUTHOR]

Published
2020
Full Text
View/download PDF

2. Multiple Subject Barycentric Discriminant Analysis (MUSUBADA): How to Assign Scans to Categories without Using Spatial Normalization

Author

James V. Haxby, Joseph Dunlop, M. Ida Gobbini, Lynne J. Williams, Andrew C. Connolly, Hervé Abdi, Abdi H., Williams L.J., Connolly A.C., Gobbini M.I., Dunlop J.P., and Haxby J.V.

Subjects
Male, Factorial, Article Subject, CEREBRAL CORTEX, Barycentric coordinate system, lcsh:Computer applications to medicine. Medical informatics, computer.software_genre, CLASSIFICATION, General Biochemistry, Genetics and Molecular Biology, Dogs, Voxel, Statistical inference, Animals, Humans, FUNCTIONAL MAGNETIC RESONANCE IMAGING (FMRI), Mathematics, General Immunology and Microbiology, business.industry, Applied Mathematics, Brain, Discriminant Analysis, PARTIAL LEAST SQUARES, Pattern recognition, Haplorhini, General Medicine, Linear discriminant analysis, Magnetic Resonance Imaging, Data Interpretation, Statistical, Face, Modeling and Simulation, Face (geometry), Spatial normalization, lcsh:R858-859.7, Female, Artificial intelligence, business, computer, Jackknife resampling, Research Article
Abstract

We present a new discriminant analysis (DA) method called Multiple Subject Barycentric Discriminant Analysis (MUSUBADA) suited for analyzing fMRI data because it handles datasets with multiple participants that each provides different number of variables (i.e., voxels) that are themselves grouped into regions of interest (ROIs). Like DA, MUSUBADA (1) assigns observations to predefined categories, (2) gives factorial maps displaying observations and categories, and (3) optimally assigns observations to categories. MUSUBADA handles cases with more variables than observations and can project portions of the data table (e.g., subtables, which can represent participants or ROIs) on the factorial maps. Therefore MUSUBADA can analyze datasets with different voxel numbers per participant and, so does not require spatial normalization. MUSUBADA statistical inferences are implemented with cross-validation techniques (e.g., jackknife and bootstrap), its performance is evaluated with confusion matrices (for fixed and random models) and represented with prediction, tolerance, and confidence intervals. We present an example where we predict the image categories (houses, shoes, chairs, and human, monkey, dog, faces,) of images watched by participants whose brains were scanned. This example corresponds to a DA question in which the data table is made of subtables (one per subject) and with more variables than observations.

Published
2012
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results