Real-Time AI Monitoring · Physics-Grounded · Clinical Imaging

You know when your scanners are failing.
Do you know when your AI is?

gammametric.com

GammaMetric monitors deployed radiology AI in real time and alerts when acquisition drift is degrading model performance. Physics-grounded sensitivity prediction from DICOM metadata — not statistical anomaly detection.

Request a Demo → CT Dose Tool →
⚠ RED Alert — AI Sensitivity Critically Degraded
Study UID: 1.2.840.SOMATOM.20260419.083142
Scanner: Siemens SOMATOM Definition AS+
Sensitivity Estimate
Estimated Sensitivity 56.2%
Degradation −22.0pp
Acquisition Parameters
Slice Thickness 5.0 mm
Kernel B40f
CTDIvol 8.2 mGy
Sensitivity for 3–6mm nodules is critically degraded under current acquisition parameters. Review protocol immediately. Primary concern: missed nodules in the clinically critical detection window.
GammaMetric AI Reliability Monitor 2026-04-19 18:03 UTC
Live alert — fired automatically on RED classification
Credentials
Yale School of Medicine Diagnostic Physics Residency
ABR Board-Eligible Diagnostic Medical Physicist
arXiv:2603.26785 — Under Review at Academic Radiology
154-Case LIDC-IDRI Perturbation Study
Live Engine Output

Real inputs. Real outputs.
No mock data.

Three studies run through the sensitivity engine. Each shows raw input, raw output, and what it means clinically.

RED — Immediate Alert Siemens SOMATOM · 5.0mm · B40f · 8.2 mGy
Input
{
  "slice_thickness_mm": 5.0,
  "reconstruction_kernel": "B40f",
  "ctdivol_mgy": 8.2,
  "scanner_model": "Siemens SOMATOM",
  "model_version": "1.0.0"
}
Output
{
  "classification": "RED",
  "baseline_sensitivity": 0.782,
  "estimated_sensitivity": 0.562,
  "degradation_pp": 22.0,
  "confidence_interval": [0.512, 0.612],
  "drivers": [
    { "parameter": "slice_thickness",
      "value": "5.0mm",
      "contribution_pp": -13.2 },
    { "parameter": "kernel",
      "value": "B40f",
      "contribution_pp": -7.9 }
  ]
}
Sensitivity for 3–6mm nodules estimated at 56.2% — 22pp below validated baseline. Immediate alert fired. Primary driver: 5mm slice thickness (−13.2pp) compounded by soft kernel (−7.9pp).
The actual email — landed in Gmail within seconds of the RED classification above
RED alert email in Gmail
YELLOW — Daily Digest GE Revolution · 3.75mm · B30f · 7.5 mGy
Input
{
  "slice_thickness_mm": 3.75,
  "reconstruction_kernel": "B30f",
  "ctdivol_mgy": 7.5,
  "scanner_model": "GE Revolution",
  "model_version": "1.0.0"
}
Output
{
  "classification": "YELLOW",
  "baseline_sensitivity": 0.782,
  "estimated_sensitivity": 0.689,
  "degradation_pp": 9.3,
  "confidence_interval": [0.639, 0.739],
  "drivers": [
    { "parameter": "slice_thickness",
      "value": "3.75mm",
      "contribution_pp": -8.1 },
    { "parameter": "dose",
      "value": "7.5mGy",
      "contribution_pp": -1.2 }
  ]
}
Sensitivity estimated at 68.9% — 9.3pp below baseline. Queued for daily digest. Slice thickness at 3.75mm is the primary driver. Protocol review recommended; no immediate action required.
GREEN — No Action Philips IQon · 1.25mm · B30f · 9.0 mGy
Input
{
  "slice_thickness_mm": 1.25,
  "reconstruction_kernel": "B30f",
  "ctdivol_mgy": 9.0,
  "scanner_model": "Philips IQon",
  "model_version": "1.0.0"
}
Output
{
  "classification": "GREEN",
  "baseline_sensitivity": 0.782,
  "estimated_sensitivity": 0.777,
  "degradation_pp": 0.5,
  "confidence_interval": [0.727, 0.827],
  "drivers": [
    { "parameter": "dose",
      "value": "9.0mGy",
      "contribution_pp": -0.5 }
  ]
}
Sensitivity estimated at 77.7% — within 0.5pp of validated baseline. Acquisition parameters are within the characterized envelope. No alert generated.
Note — These are real outputs from the live engine. Sensitivity deltas are derived from 154-case LIDC-IDRI perturbation experiments (arXiv:2603.26785). Baseline: MONAI RetinaNet, LUNA16-trained, v1.0.0.
Services

Monitoring deployed AI.
Continuously.

Real-time AI reliability monitoring is live. CT dose analytics runs alongside it — the same physics infrastructure, two clinical problems solved.

Available Now

AI Reliability Monitoring

  • Physics-grounded sensitivity prediction from DICOM acquisition metadata
  • GREEN / YELLOW / RED classification per study — no manual review required
  • RED alert fires immediately to AI governance contact
  • YELLOW studies batched into daily digest
  • 7-page PDF reliability report — acquisition profile, degradation analysis, protocol recommendations
  • Full audit log — every study classified and timestamped
  • Orthanc DICOM listener integration — passive ingestion, no workflow disruption
  • Based on published research: arXiv:2603.26785, under review at Academic Radiology
Request a Demo →
Also Available

CT Dose Analytics & Protocol Optimization

  • Self-serve tool at dose.gammametric.com — free, no account required
  • Leapfrog Section 8B pediatric CT dose reporting
  • Adult CT DLP percentiles — all body regions
  • ACR DIR national benchmark comparisons with status flags
  • Outlier detection with methodology notes
  • Protocol optimization observations — actionable recommendations
  • Physicist-reviewed PDF report — 24–48 hour turnaround
  • Direct physicist support via email
Try the Free Tool →
Research & Advanced Imaging

Support for imaging performance and variability in research, multi-site, and non-routine clinical contexts.

  • MRI & Fluoroscopy Protocol Review — Acquisition parameter evaluation, consistency assessment, and variability characterization across systems and operators
  • Multi-site Harmonization — Protocol difference identification and impact on cross-site comparability in trials or distributed imaging environments
  • Quantitative Imaging & Endpoint Validation — Assessment of how acquisition variability affects measurement stability, AI outputs, and derived imaging biomarkers
Inquire →
Case Studies

The research behind the work.

Two analyses showing exactly what GammaMetric measures — and what it finds.

Protocol Optimization

Your Protocols Are Costing You on Three Fronts Simultaneously

Dose compliance. Image quality. AI performance. Most CT protocol reviews address one. This analysis shows how the parameters interact — and which ones actually matter.

  • 5mm slice thickness: −13.2pp AI sensitivity loss
  • Soft reconstruction kernel: −10.5pp AI sensitivity loss
  • mAs reduction: only −4pp — the least destructive lever
  • Leapfrog compliance and AI performance are different problems
Read the case study →
AI Validation

What Your AI Misses After Deployment

Post-deployment validation of a CT lung nodule detection algorithm across six real-world imaging perturbations. Based on LIDC-IDRI (154 cases). Methodology: arXiv:2603.26785.

  • Baseline sensitivity: 84.8% under reference protocol
  • Combined perturbation: ~65–68% — a 20pp gap
  • Effect most pronounced in the 3–6mm nodule range
  • Vendor benchmarks do not reflect site-specific conditions
Read the case study →
The Problem

There's a gap between what your
equipment is supposed to do —
and what it's actually doing.

Most imaging providers and research organizations are flying blind. Dose settings drift. Protocols go unreviewed. AI tools get deployed without validation against real acquisition conditions. The gap is invisible until something goes wrong.

01

Protocol Creep Is Silent

Dose settings from years ago are still running on most outpatient CT scanners. Nobody decided to keep them — nobody decided to change them either. The result is unnecessary dose, inconsistent image quality, and benchmarks that quietly drift out of range.

02

Compliance ≠ Optimization

Passing Leapfrog Section 8B means your numbers are below the threshold — not that your protocols are optimized. Most facilities stop at compliance. The gap between "acceptable" and "excellent" is where dose reduction opportunities live.

03

Your AI Wasn't Validated On Your Conditions

FDA-cleared imaging AI is tested at controlled dose levels and standard protocols. Your facility runs lower doses, thicker slices, and varied techniques. Nobody tested what happens in that gap — until now.

The Process

Passive by design.
No workflow disruption.

Once configured, GammaMetric runs continuously in the background. Every study that passes through your scanner is classified automatically — no uploads, no manual review, no intervention required.

01

Forward DICOM Headers

Configure your scanner to forward studies to a lightweight Orthanc DICOM listener. Only acquisition metadata is extracted — no pixel data, no image storage, no PHI transmitted.

02

Engine Classifies Every Study

Each study runs through the sensitivity engine. Slice thickness, reconstruction kernel, and dose level are mapped to an estimated sensitivity value. The study is classified GREEN, YELLOW, or RED.

03

Alert Fires Automatically

RED studies trigger an immediate email to your AI governance contact. YELLOW studies appear in a daily digest. Every study is logged with its classification, parameters, and sensitivity estimate.

On Demand

PDF Reliability Report

Generate a 7-page site-specific reliability report from any study — acquisition profile, sensitivity degradation analysis, and prioritized protocol recommendations. Suitable for post-market surveillance documentation.

Request a Demo →
Also Available

CT Dose Analytics

Self-serve CT dose monitoring at dose.gammametric.com. Leapfrog Section 8B reporting, ACR DIR benchmarking, drift alerts. Free to use.

Try It Free →
Report Contents

Everything your quality
program needs

Compliance

Leapfrog Section 8B Reporting

Median DLP for routine head and abdomen-pelvis CT across all five Leapfrog pediatric age groups (<1, 1–4, 5–9, 10–14, 15–17) — formatted and ready for Section 8B reference.

Benchmarking

ACR DIR Benchmark Comparisons

Your facility's dose percentiles compared against ACR Dose Index Registry national reference levels. Clear status flags — Excellent, Acceptable, or Above Benchmark — for every body region.

Quality

Outlier Detection

Automatic identification of exams with unusually high DLP — repeat acquisitions, wrong protocols, or multi-phase studies — with transparent methodology notes for your physics team.

Optimization

Protocol Observations

Physicist observations on protocol consistency, scanner variability, and dose reduction opportunities — useful context for your quality improvement program beyond compliance reporting.

Trend

Dose Trends Across Reporting Period

Dose trends visualized across your full reporting period. Identify protocol changes, scanner drift, or technologist variability — supporting ongoing QA program development beyond Leapfrog season.

Deliverable

Professional PDF Report

Publication-quality output with percentile tables, benchmark charts, methodology documentation, and your facility name — suitable for quality committee presentation or Leapfrog submission reference.

Context

The performance gap
is measurable

GammaMetric's own pilot study quantifies how acquisition variability affects imaging AI — and why protocol optimization matters beyond compliance.

~4pp

Sensitivity drop in CT lung nodule detection AI at 25% dose reduction — a protocol choice many facilities make routinely, with unquantified impact.

~19pp

Sensitivity drop at 5mm slice thickness versus standard. The gap between your protocol and the vendor's validation conditions is rarely measured.

<48 hrs

GammaMetric turnaround from data submission to physicist-reviewed dose analytics or protocol optimization report.

Pricing

Built for vendors.
Priced per site.

AI monitoring is the primary product. CT dose analytics runs alongside it, free.

AI Reliability Monitoring
Pilot
Pricing
per site · monthly · contact for terms
  • Continuous DICOM header monitoring — passive, no workflow disruption
  • Physics-grounded sensitivity prediction per study
  • RED immediate alert to AI governance contact
  • YELLOW daily digest
  • 7-page PDF reliability report on demand
  • Full audit log — every study classified and timestamped
  • Based on published research — defensible for post-market surveillance documentation
Request a Demo →
CT Dose Analytics
$0
free tool · always available
  • Self-serve at dose.gammametric.com
  • Adult CT DLP percentiles — all body regions
  • Pediatric CT — all five Leapfrog age strata
  • ACR DIR national benchmark comparisons
  • Drift alerts and QA acknowledgment workflow
  • Physicist-reviewed reports available — $1,500/facility/year
Try It Free →
FAQ

Common questions

What data format do you accept?
CSV exports from Radimetrics (Bayer), DoseWatch (GE), or any dose monitoring system. Manual PACS query exports are also accepted. Common column naming conventions are auto-detected. Non-standard formats are welcome — format mapping is handled before analysis begins.
Is my data secure?
De-identifying patient data before sending is strongly recommended — the analysis only requires dose metrics, exam descriptions, and patient age. No PHI is needed or requested. Raw data files are not retained after analysis is complete.
What Leapfrog section does this cover?
Section 8B: Pediatric Computed Tomography (CT) Radiation Dose. This requires reporting median DLP for routine head and abdomen-pelvis CT across five pediatric age groups. Reports provide exactly those data points, plus adult CT analysis as a value-add for your quality program.
Is this a replacement for a medical physicist?
No. Every report includes review by a diagnostic medical physicist, but final interpretation, regulatory compliance, and clinical protocols remain the responsibility of your institution and its qualified physics staff. GammaMetric is a reporting and analytics service, not a substitute for physics oversight.
How are benchmarks determined?
Dose percentiles are compared against ACR Dose Index Registry (DIR) national reference levels, maintained and updated by a diagnostic medical physicist to reflect current national practice.
What does an AI validation engagement look like?
You provide de-identified DICOM data or model outputs across your facility's acquisition conditions (dose levels, slice thicknesses, protocols). GammaMetric applies systematic degradation and inference to quantify sensitivity loss and failure modes under each condition — delivered as a physicist-reviewed report with methodology documentation suitable for quality committee or regulatory review.
What does the AI monitoring system actually do?
It reads DICOM acquisition metadata from incoming CT studies, runs them through a physics-grounded sensitivity engine, classifies each study GREEN / YELLOW / RED, and fires an alert if parameters are degrading AI performance. RED studies trigger an immediate email. YELLOW studies appear in a daily digest. Every study is logged with its classification, acquisition parameters, and sensitivity estimate — full audit trail. No manual review required.
Get Started

See it
working.

15-minute demo. Live system, real DICOM parameters, real alert. No slides.

Request a Demo → CT Dose Tool →

Or email directly: daniel@gammametric.com