BrowserQC

Automated MRI quality control, entirely in your browser — no upload, no server. Drag in a NIfTI image (or a folder of DICOM files) and it runs automatically.

• Segmentation — a brainchop deep-learning model (TensorFlow.js) parcellates the brain into 17 gray/white-matter and subcortical regions, shown as a colour overlay on your scan (adjust with the Opacity slider)
• Quality metrics — niimath computes MRIQC-style image-quality metrics (CJV, CNR, SNR, WM2MAX, EFC, tissue composition) from the scan and its segmentation

Rendering uses NiiVue; DICOM import uses dcm2niix.

The image-quality metrics follow MRIQC: Esteban et al. (2017) MRIQC: Advancing the automatic prediction of image quality in MRI from unseen sites. PLoS One 12(9):e0184661. PMID: 28945803. With a masked background the CNR omits the air-noise term, so values are relative, not comparable to MRIQC normative data.

Licensed BSD-2-Clause. Requires a browser with WebGPU (recent desktop Chrome, Edge, or Safari).