Tutorials
This page contains unreviewed content created with large language model (LLM) tools.
This tutorial converts a BIDS dataset to BITS, queries the persistent BIDS-equivalent index, reads images and metadata, and constructs batches for model training. The second part demonstrates an isolated workspace containing a quantised version of one image and examines the resulting lineage records.
The examples assume familiarity with Python, BIDS, and common neuroimaging file formats. No prior knowledge of BITS is required.
The tutorial is intentionally circumscribed with the goal of developing working knowledge, not exhaustively enumerating use cases; consult the how-to guides for specific tasks, explanation for design details, and reference for the API.
1 Chapters
1.1 Part 1: ingestion and batched reads
Starting from an OpenNeuro ds000114 subset, create a BITS dataset, query its BIDS index, read images and metadata, and iterate through model-ready batches.
1.2 Part 2: workspaces and provenance
Create a copy-on-write workspace over the dataset. Quantise one BOLD image to int8, read the stored codes, inspect the recorded lineage, and then discard the workspace.