Inspectable measurements
Regional attenuation, spatial texture, patient-relative features, and quality checks.
An existing scan can open a new line of inquiry. We’re exploring the patterns inside bone—and what they might add to our understanding of fragility.
CT records values throughout a three-dimensional volume. Explore how we turn those values into measurements we can study.
Anatomical masks identify the sacrum and left/right iliac regions. Image geometry checks help ensure the measurements come from the intended tissue.
Both grids contain equal numbers of light and dark cells. Their averages are identical. Their neighbouring values tell a different story.
A teaching example of texture. These patterns do not represent healthy or diseased bone.
Our current experiments study regions with existing fractures. The longer-term question is whether CT features can improve prediction of future fragility fractures.
Inspect the experiment ↗Regional attenuation, spatial texture, patient-relative features, and quality checks.
Matched model comparisons with patient-locked splits and reported uncertainty.
Pre-fracture scans, follow-up, paired DXA and clinical comparators, and external validation.
Research use only. Routine CT texture is resolution-dependent; it does not directly resolve trabecular connectivity. No validated clinical risk score or superiority to DXA/FRAX is established.
Use the browser to explore the demo. Use the Python toolkit to extract measurements from native CT volumes and supplied region masks.
Browse curated pelvic CT images, reference annotations, feature summaries, and the current experiment.
Open research workspace ↗BUILD / WITH THE TOOLKITLocal extraction, input checks, JSON and CSV exports, and source fingerprints for reproducible research.
Read the quickstart ↗Help build the evidence for a richer view of bone health.