Supplementary data and code for: A study of the accuracy of smartphone augmented-reality measurement for early melanoma detection

Data and code that recompute every number derived by the authors in the paper "A study of the accuracy of smartphone augmented-reality measurement for early melanoma detection" (Bulletin of Cherkasy State Technological University). The deposit contains the 1,523 clinical records scored by the screening rules (the 1,934 records identified minus the 411 removed by the keyword prefilter), with their titles and abstracts, the PubMed records, the retrieval log, the record counts of the fourteen intersection-probe queries, the metric-type labels of the 83 papers in the extraction table, the reference list with the full-text status of each source, and the values quoted from cited studies that enter a calculation or a figure. recompute_statistics.py applies the screening rules and recomputes all 70 reported values. Its output can be compared byte for byte with expected_output.txt. make_figures.py draws Figures 1-4 from the same tables. Both scripts are deterministic, make no network requests and call no language model. Requirements: Python 3.9+, pandas, numpy and, for the figures, matplotlib. Scripts are released under the MIT licence and data under CC BY 4.0. The abstracts in data/screened_records.csv remain the copyright of their authors or publishers and are included only to verify the screening counts.