Usage guide

Getting oriented in detectQS

Pick your reviewer name and dataset once, then open a case whenever you want to review it -- with AI-assisted suggestions and postmenstrual age estimation available alongside your own manual review. Every screenshot below is from the app running against a synthetic (fake, non-patient) recording, so you can compare it against your own screen without seeing anyone's real data here.

i Windows will likely show a blue “Windows protected your PC” screen the first time you run detectQS.exe — expected for a small independent tool without a paid publisher certificate, not a sign anything's wrong. Click More info, then Run anyway.

Open the dashboard

This is the app's home screen. Enter your reviewer name once (it's remembered for next time, and every SWC annotation you confirm gets attributed to it) -- cases can't be opened until it's filled in.

detectQS dashboard, listing one case under a synthetic dataset folder — The dashboard: reviewer name, dataset folder, and a list of cases found inside it.

Point it at your data folder

Click Change Dataset Folder... and browse to wherever your recordings live -- detectQS scans it for case folders (each one containing the EEG/impedance signal files and a studyinfo.xml) and lists what it finds. You only need to do this once; it's remembered between launches.

A folder picker showing only the synthetic_dataset folder, staged to avoid revealing unrelated local folder names — A folder picker -- nothing detectQS-specific here, just point it at your data. (Yours will look like your OS's own file browser; this one's been staged to keep unrelated local folder names off this page.)

Review a case

Clicking a case opens the full review window: aEEG traces for both hemispheres up top, a slider to move through the recording two hours at a time, and impedance or raw EEG underneath (toggle between them with the button in the bottom row). Click anywhere on an aEEG trace to inspect that moment's raw EEG; draw and confirm Sleep‑Wake Cyclicity intervals directly on the trace, and they save automatically as you go.

An open case in detectQS, showing left/right aEEG traces, the time slider, and impedance traces below — A case open for review -- this one is six hours of synthetic signal, so scrolling the slider only moves a little.

Let the AI suggest a starting point

Suggest Annotations (AI) proposes candidate Sleep‑Wake Cyclicity intervals for both hemispheres at once (shown in purple, distinct from your own confirmed intervals) -- click into one to accept, adjust, or reject it, or dismiss the whole batch with Clear AI Suggestions if it doesn't look right. Nothing is saved to the case until you confirm it yourself. If more than one model is available, a dropdown next to the button lets you pick between them.

The PMA Assessment panel below the annotation controls covers postmenstrual age two ways: enter the four Burdjalov scale components yourself for a reference range, or click Run AI - PMA Estimation for an independent model-based estimate -- useful as a sanity check against each other, or as a starting point on a case you haven't annotated yet.

An open case in detectQS showing AI-suggested SWC intervals in purple on both aEEG traces, and the PMA Assessment panel with a Burdjalov total, suggested PMA range, and AI PMA estimate filled in — AI-suggested intervals on both hemispheres, plus the PMA Assessment panel -- Burdjalov total and AI estimate shown side by side.

Try it yourself, no approval needed

Want to see how detectQS handles your own data folder structure before requesting access? This is the exact synthetic recording used in the screenshots above -- entirely fake signal, safe to poke at, no license key required.

Download synthetic EDF dataset