AI Diagnostic Accuracy Meta-Analysis and Statistical Outputs
Publicada el 2026-07-31
Descripción de la oferta
Statistician Required to Produce Forest Plots and SROC Figures for Diagnostic Accuracy Meta-analysis Job description I am seeking an experienced biostatistician or diagnostic test accuracy meta-analysis specialist to produce a confidential, publication-quality statistical figures and summary outputs for a postgraduate biomedical science thesis. The research examines the diagnostic accuracy of artificial intelligence methods used to detect malaria from peripheral blood films. A completed Excel spreadsheet will be provided containing the extracted study characteristics and diagnostic accuracy data. The systematic review includes 115 studies, of which 36 have currently been identified as suitable for quantitative synthesis because they report complete or derivable two-by-two diagnostic data, including true positives, false positives, true negatives, and false negatives. The work mainly involves checking the supplied quantitative data and generating professionally formatted meta-analysis figures. Extensive thesis writing is not required. Required work The freelancer will be expected to: 1. Confirm an appropriate grouping of studies by unit of analysis, particularly: cell or parasite-level studies and patient, or slide-level studies 2. Conduct an appropriate diagnostic test accuracy meta-analysis using a bivariate random-effects or hierarchical summary receiver operating characteristic model. 3. Generate separate coupled forest plots for sensitivity and specificity for each relevant subgroup. 4. Generate SROC or HSROC plots showing: study-level estimates pooled operating point 95% confidence region prediction region 5. Provide pooled sensitivity and specificity with 95% confidence intervals. Provide positive and negative likelihood ratios and diagnostic odds ratios where appropriate. 6.Report relevant measures of between-study heterogeneity. Deliverables The completed work should include: a. Publication-quality forest plots for each analytical subgroup. SROC or HSROC figures with confidence and prediction regions. b. A concise summary table of pooled diagnostic accuracy estimates. c. A clean list of studies included in each analysis. d. High-resolution figures suitable for insertion into Microsoft Word. e. Complete R, Stata, SAS, or equivalent statistical code used to generate the analysis. f. A brief technical explanation of the statistical model, assumptions, exclusions, and findings. Required expertise Applicants should have proven experience in: Diagnostic test accuracy meta-analysis Bivariate random-effects or HSROC modelling Sensitivity and specificity forest plots SROC confidence and prediction regions R, Stata, SAS, MetaDTA, or equivalent software Producing publication-quality figures for theses or peer-reviewed papers Experience in medical diagnostics, artificial intelligence, microscopy, infectious diseases, or systematic reviews would be advantageous. Proposal requirements Please include: Your qualifications and relevant experience An example of a diagnostic accuracy forest plot or SROC analysis you have produced The statistical software and packages you will use Confirmation that the spreadsheet and research data will remain confidential
Skills
Fuente original: freelancer