Search the collection
Filter by modality and research focus. The ring distribution is fixed — categories excluded by the current filter are dimmed rather than removed. The panel on the right shows what remains in this combination. Only aggregate counts appear here: no paths, no names.
Where these two dimensions come from
Modality and research focus are not hand-entered — they are parsed from directory names. The group already uses a project_modality_date_version convention (e.g. sca3_cj_fmri_20261109), and the parser only accepts words that are actually present in the name; anything it cannot read is filed as Unlabelled rather than guessed. When the name gives no clue, the second-level subdirectory is used as secondary evidence (e.g. AD数据汇总\AD原始脑电 → EEG), and the evidence word is recorded. Currently 29 datasets have no modality clue and 51 no focus clue; both appear as grey segments on the ring and can be clicked to list them on their own. Filling these in is the next task.