🩻

Opportunistic CT Screening Under development

Turn a routine non-contrast abdominal CT into added-value body-composition screening — bone density, muscle, aortic calcium, visceral fat, and liver fat — with a copy-ready report block.

Best used once, and strongest on non-contrast. This is a high-value one-time add-on, ideally applied opportunistically to a non-contrast abdominal CT already obtained for another reason, in an adult roughly 50–70 years old who has not had prior body-composition/bone-density screening on an earlier CT — every measurement below is validated or preferred on non-contrast. Contrast-enhanced studies are supported too if that's what you have: select the phase below and the measurements that don't apply will gray out or carry a caution note. Apply to routine outpatient CTs only — not emergency-department or inpatient scans, where acute illness, fluid shifts, and positioning confound body-composition measures. (Yield for osteoporosis, cardiovascular calcium, and sarcopenia climbs after ~50 — this range is a starting suggestion, adjustable to local practice.)

Sex

Height cm (for SMI)

Study phase

Select the phase actually acquired. Measurements not validated for that phase gray out and disable below; ones still usable but read differently on that phase stay active with a caution note.

1 · Bone density

L1 trabecular attenuationNon-contrast or PV

L1 vertebral trabecular attenuation HU (oval ROI in mid-vertebral trabecular bone)

Prevalent vertebral compression fracture present

Pickhardt 2013: IV contrast did not significantly affect L1 performance, so the same thresholds apply to contrast-enhanced scans. A prevalent vertebral compression fracture is the single strongest predictor of future fracture — and DXA is often falsely negative when one is present.

2 · Muscle (L3)

myosteatosis + sarcopeniaArea: any · Attenuation: phase-sensitive

Measure on a single axial slice at mid-L3 (both transverse processes visible). Outline the total skeletal muscle — paraspinal (erector spinae, psoas, quadratus lumborum) plus the abdominal wall (rectus, obliques, transversus). Attenuation = the mean HU of that region. Area = the total cross-sectional area with the muscle threshold set to -29 to +150 HU. Segmentation software gives both at once; otherwise a large freehand ROI over the muscle approximates the mean HU.

Mean skeletal muscle attenuation (L3) HU (mean of the outlined muscle)

Total skeletal muscle area (L3) cm2 (needs sex + height above)

3 · Abdominal aortic calcium

cardiovascular riskNon-contrast preferred

Aortic calcification (visual)

Why a visual estimate, not an Agatston score?

A true Agatston score is produced by dedicated calcium-scoring software: on each slice, every calcified focus (peak ≥130 HU, area ≥1 mm²) has its area (mm²) multiplied by a density factor based on peak HU — 1 for 130–199, 2 for 200–299, 3 for 300–399, 4 for ≥400 HU — then summed across all foci and slices.

That needs a dedicated, usually ECG-gated, thin-section acquisition — a routine abdominal CT is not validated to produce one. Because this is an opportunistic screen, we report a qualitative none / mild / moderate / severe estimate instead; any calcification is still a cardiovascular-risk marker.

4 · Visceral adiposity (L3)

advisory metabolic markerAny phase

Reference — drawn to scale: the subcutaneous (SAT) rim area and the visceral (VAT) fat area equal the labeled ratio; the organs — colon, renal poles, aorta — and the muscle wall are subtracted, so the fat-area ratio stays exact. Calibrate your area estimate — not a risk threshold: VAT/SAT is strongly sex-dependent (higher in men) with no validated cutoff.

Visceral vs subcutaneous fat (L3)

Qualitative eyeball estimate at L3 — no quantification tool needed. A visceral-predominant (apple) pattern carries higher cardiometabolic risk, but normal VAT/SAT is strongly sex-dependent (higher in men) with no validated universal cutoff — interpret the pattern in the context of sex, not against a fixed ratio.

5 · Hepatic steatosis

screen only — quantify with MRI-PDFFNon-contrast or PV

Liver attenuation HU

Spleen attenuation (optional, non-contrast) HU

Non-contrast or portal-venous CT screen only; each phase has its own threshold. For the occasional case that needs fat quantification or monitoring, use the PDFF Liver Fat Assistant — not routine for every fatty liver.

Report (copy)

Enter any measurement above.

Thresholds & references

Bone density — L1 trabecular attenuation

>200 HU: osteoporosis effectively excluded (NPV ~99%). 110–160 HU: indeterminate (the 90%-sensitive-to-90%-specific zone) — consider DXA. <110 HU: suspicious for osteoporosis (>90% specific); <100 HU PPV ≥68%. Most vertebral fractures occurred at ≤145 HU. IV contrast did not significantly affect performance. Pickhardt et al., Ann Intern Med 2013 (doi:10.7326/0003-4819-158-8-201304160-00003).

Muscle attenuation (L3) — myosteatosis

Low muscle attenuation independently predicts mortality. 90%-specificity thresholds: 23 HU (men), 13 HU (women); the men's death-optimized cut was ~31 HU. Lee et al., Radiology 2023 (doi:10.1148/radiol.220574).

Sarcopenia — L3 skeletal muscle index (SMI = area ÷ height²)

Cutpoints men <52.4, women <38.5 cm²/m² (Martin et al., J Clin Oncol 2013). Population-specific cutoffs vary; interpret with clinical context.

Abdominal aortic calcium

Reported here as a qualitative visual estimate (none/mild/moderate/severe) — a routine abdominal CT is not validated for a true Agatston score. With dedicated scoring, the sex-specific 90%-specificity Agatston thresholds for 10-year death were ≥1475 (men) and ≥735 (women) — Lee et al. note these values “would all be considered a severe or advanced category” of aortic calcium, so the mortality signal is best established at moderate-to-severe burden. Lee et al., Radiology 2023 (doi:10.1148/radiol.220574). A pooled meta-analysis found any-or-advanced calcification associated with higher cardiovascular and all-cause mortality, though this is a heterogeneous, less granular estimate: Leow et al., J Am Heart Assoc 2021 (doi:10.1161/JAHA.120.017205).

Visceral adiposity

The VAT/SAT ratio reflects fat distribution but is strongly sex-dependent — men normally have much higher ratios than women (apple vs pear habitus) — and has no validated universal cutoff. A ratio of ~1 is therefore not a risk threshold; interpret a visceral-predominant (apple) pattern in the context of sex. Reported here descriptively. Pickhardt, Radiology 2022 (doi:10.1148/radiol.211561).

Hepatic steatosis

On non-contrast CT, liver attenuation <40 HU (which corresponds to ~15% MRI-PDFF) indicates at least moderate steatosis; a liver−spleen difference <−10 HU is a supporting sign. A screen only — for MASLD, fibrosis is the outcome that matters, so stratify risk non-invasively first (e.g., FIB-4) and reserve MRI-PDFF for selected cases needing fat quantification or monitoring rather than imaging every fatty liver. Pickhardt, Radiology 2022 (liver <40 HU ≈ 15% PDFF; doi:10.1148/radiol.211561); Starekova/Reeder/Pickhardt, Radiology 2021 (doi:10.1148/radiol.2021204288); AASLD 2023.

Portal-venous phase. Contrast-enhanced CT can still flag moderate steatosis with its own cutoffs: automated liver attenuation <80 HU gives 78% sensitivity and 93% specificity, and <90 HU gives 91% sensitivity and 78% specificity. On this phase absolute liver HU outperforms the liver−spleen difference, so the spleen is not used. Pickhardt et al., AJR 2023;221:748–758; Pickhardt et al., Eur Radiol 2024;34:7041–7052; summarized in Dogra et al., RadioGraphics 2026 (doi:10.1148/rg.260009). Arterial phase remains unvalidated.

Overview: Pickhardt et al., RadioGraphics 2021 (doi:10.1148/rg.2021200056); Pickhardt, Radiology 2022 (doi:10.1148/radiol.211561). Educational tool — thresholds are population-derived and must be integrated with the clinical context.

1 · Bone density

2 · Muscle (L3)

3 · Abdominal aortic calcium

4 · Visceral adiposity (L3)

5 · Hepatic steatosis

Report (copy) Copy

Bone density — L1 trabecular attenuation

Muscle attenuation (L3) — myosteatosis

Sarcopenia — L3 skeletal muscle index (SMI = area ÷ height²)

Abdominal aortic calcium

Visceral adiposity

Hepatic steatosis

Report (copy)