GlenometriX

Research Access

Quantify What Others
Estimate.

Glenometrix measures glenoid bone loss from 3D CT scans with ICC 0.87 agreement against expert clinicians. The objective measurement every shoulder instability case requires.

ICC 1.00perfect reproducibility
ICC 0.87agreement with expert clinicians
88%concordance across bone loss categories
Evidence base
Khan et al. · GlenometriX Validation · McMaster University 2026Balg & Boileau · ISIS Score · JBJS 2007Di Giacomo et al. · Glenoid Track · Arthroscopy 2014Burkhart & De Beer · Critical Threshold · Arthroscopy 2000

The Problem

Bone loss changes everything. Most tools don't measure it reliably or accurately.

Shoulder instability outcomes hinge on many factors. Glenoid bone loss percentage is one of the most important variables in determining recurrent instability. Yet most clinical assessments rely on visual estimation, which carries documented poor inter- and intra-rater reliability. Glenometrix measures it objectively.

40%
Recurrence rate with Bankart repair when ISIS score ≥ 7Balg & Boileau, JBJS 2007
20%
Critical bone loss threshold mandating a bone block procedureBurkhart & De Beer, Arthroscopy 2000
0.87
Inter-rater ICC against 5 expert clinicians: 3 MSK radiologists, 2 orthopaedic surgeonsKhan et al., McMaster University 2026

How It Works

The Glenometrix Score: Objective. Reproducible. Instant.

Upload a de-identified 3D CT scan. GlenometriX uses a research algorithm to automatically segment the glenoid from the affected side alone. No contralateral imaging required. It then applies a best-fit circle algorithm, and delivers a structured clinical output including ISIS scoring, on/off-track assessment, and a decision pathway.

Upload CT Scan

De-identified 3D CT imaging only. En face glenoid reconstruction. DICOM, JPEG, or PNG accepted.

Automated Segmentation

A research algorithm segments the glenoid, identifies anterior and posterior margins, and applies a best-fit circle algorithm to calculate percentage bone loss.

Glenometrix Score

Bone loss %, ISIS score, on/off-track status, and a rule-based decision pathway: Bankart, remplissage, or bony augmentation.

The Science

Validated against five expert clinicians.

GlenometriX was developed on 90 annotated 3D-CT scans and validated on a separate cohort of 121 scans against five blinded, fellowship-trained experts: three MSK radiologists and two orthopaedic surgeons.

ICC 1.00
Perfect intra-rater reproducibility. Deterministic algorithm: identical output on repeat measurements.
ICC 0.8795% CI 0.82–0.91
Inter-rater agreement with the combined expert average
88%
Concordance across minimal, subcritical, and critical bone loss categories
−0.17%
Near-zero mean bias vs. the expert average on Bland–Altman analysis
121
3D-CT scans in the independent validation cohort, separate from training data
5
Blinded expert raters: 3 MSK radiologists and 2 orthopaedic surgeons

A research algorithm segments the glenoid and applies a best-fit circle to the affected side alone. No contralateral scan required.

Khan et al. "GlenometriX: A Reliable Tool for Quantifying Glenoid Bone Loss." McMaster University, 2026. Manuscript in submission.

Sample analysis. Research use only. Not for clinical use.

From 3D CT scan to structured surgical decision.

Glenometrix Analysis Report
Research Use Only. Not For Clinical Use.
Quantitative Output
Glenoid Bone Loss18.4%
0%20% critical
ISIS Score7/10
0≥7 high risk
Clinical Status
High Risk
Off-Track
ISIS Factors
Age < 20y2/2
Competitive sport2/2
GBL > 20%2/2
Off-track Hill-Sachs2/2
Decision Pathway
Glenoid Track21.4 / 24.8mm
Glenoid track widthHSI

Bone block procedure indicated (Latarjet or Eden-Hybinette). Arthroscopic repair contraindicated at this bone loss level.

Recurrence Risk> 40%

with soft-tissue Bankart repair at this bone loss level

Access Required

Automated CT Bone Loss Analysis

GlenometriX quantifies glenoid bone loss directly from a 3D CT scan using a validated AI model (ICC 0.87 agreement with expert clinicians). Full analysis is currently limited to research partners.

Drop de-identified CT imaging here

DICOM · JPEG · PNG · Max 50 MB

AI CT analysis is rolling out to research partners. Request access to enable it at your institution.

Research Access

Join Mailing List

Sign up to stay updated on Glenometrix developments. Open to orthopedic surgeons, sports medicine physicians, and shoulder researchers.