STEP BY STEP CT PERFUSION ON GE CT SCANNER, Processing settings,

 Awesome—here’s a practical, step-by-step CT Brain Perfusion (CTP) workflow on GE (Revolution/Discovery/Optima family with AW/ADW “CT Perfusion 4D/BrainWave” apps). I’ve kept it scanner-room → console → post-processing, with exactable numbers you can use today.

1) Patient + IV prep

• Consent & checks: contrast allergy, eGFR, pregnancy, metformin policy per site.

• IV: 18–20G antecubital, power-injector–rated; test flush.

• Immobilize: head pad + strap; remove dentures/hearing aids; coach “stay still.”

2) Scout + Non-contrast CT head (NCCT)

• Purpose: rule out hemorrhage; define levels.

• Typical: 120 kVp, automated mA (or ~200–300 mAs eff.), 5 mm axial + 0.625–1.25 mm thin recon; iterative recon (ASiR-V) on.

3) Choose your CTP acquisition mode

A. 64-slice (≈4 cm slab; shuttle up to 8–10 cm if available)

• Level: basal ganglia (include M1/M2 + thalami).

• Mode: axial cine (or shuttle for extended coverage).

• Typical parameters

  • kVp 80 (70–80), 100–200 mA (Smart mA OK)

  • Rotation 0.4–0.5 s; slice 5 mm (8 slices → 40 mm)

  • Temporal sampling: 1–2 s/vol for ~45–60 s (extend to 70 s if slow flow)

• Contrast: 40–50 mL non-ionic 320–370 mgI/mL @ 5 mL/s, then 30–40 mL saline @ same rate.

• Timing: Start the series and trigger injector with ~5 s delay (captures baseline + full bolus).

B. Wide-coverage (Revolution/256+ / shuttle or 16 cm volume)

• Aim for whole-brain (≥8–16 cm).

• Typical parameters

  • kVp 70–80, 100–200 mA, rotation 0.28–0.5 s

  • Dynamic volumes: early dense sampling (every 1–2 s for first 30 s), then sparser (every 3–5 s) until 60–75 s total.

• Same contrast plan as above.

Dose tips (both modes): low kVp, iterative recon (ASiR-V), smaller FOV to head, minimize repeat runs, consider sparser late sampling.

4) Quick CTA (optional but common in stroke bundle)

• If your pathway is NCCT → CTP → CTA or NCCT → CTA → CTP, keep injector connected. CTA: 60–70 mL @ 5–6 mL/s + saline.

5) Send to AW/ADW and launch perfusion app

• Open CT Perfusion 4D / BrainWave workspace.

• Verify the correct series (CTP, not NCCT/CTA).

• Motion correction: run auto; if residual, apply manual rigid registration.

6) Arterial & Venous input selection (critical!)

• AIF (arterial): small VOI on M2/M3 MCA (or A2 ACA) on the non-stenosed side. Look for sharp upslope, high narrow peak, minimal dispersion.

• VOF (venous): large VOI in superior sagittal sinus (straight segment). Tall, delayed, clean venous curve.

7) Processing settings

• Algorithm: Delay-corrected deconvolution (oSVD) (avoids arterial delay pitfalls).

• Generate maps: CBF, CBV, MTT, TTP, Tmax. Check for truncation (bolus not fully washed out) or noisy maps.

8) Derive core & penumbra (stroke)

(Confirm your site’s validated thresholds with neurology; below are commonly used.)

• Core (infarct): relative CBF < 30% of contralateral (preferred), or very low CBV region.

• Penumbra (tissue at risk): Tmax ≥ 6 s (optionally grade 6–10 s vs >10 s for severity).

• Mismatch volume = Penumbra − Core; Mismatch ratio = Penumbra / Core (often treatment criteria need ratio ≥1.8 and core below a size threshold—follow your protocol).

9) QA checklist before you sign off

• AIF/VOF curves look physiologic (AIF peaks ~15–25 s; VOF later and taller).

• No major motion after correction (ring/edge ghosts = motion).

• Adequate duration (late venous phase sampled; if not, repeat with extended time).

• Consistent anatomy across timepoints (table didn’t drift; shuttle registered).

• Maps agree with CTA (e.g., Tmax prolongation in the vascular territory of an M1 occlusion).

10) Troubleshooting (fast fixes)

• Flat AIF / weak enhancement: check IV, rate, or contrast concentration; re-run with 5–6 mL/s and good antecubital line.

• Severe carotid/M1 stenosis: pick contralateral AIF; ensure delay-corrected model; interpret Tmax carefully.

• Patient moved: head strap; repeat with better immobilization; use motion correction.

• “Patchy” core on rCBF: re-check AIF/VOF placement; look for truncation (too short acquisition).

• Beam-hardening from skull base: set slab slightly higher; use iterative recon.

11) Suggested acquisition “starter” presets (tune per site)

64-slice cine (4 cm): 80 kVp, 150 mA, 0.5 s rot, 8×5 mm, 1.5 s/vol, 60 s total; 45 mL @ 5 mL/s + 35 mL saline.
Shuttle (8–10 cm): 80 kVp, 200 mA, 0.5 s, 2 positions alternating, effective ~2–3 s per volume per slab, 70 s total.
16-cm volume (Revolution): 80 kVp, auto mA (cap ~200), 0.28–0.5 s, sampling every 1–2 s (0–30 s) then every 4 s (30–70 s).

12) One-minute reporting template

• Technique: NCCT + CTP (coverage X cm, duration Y s, contrast Z mL @ R mL/s, delay-corrected deconvolution).

• Quality: motion ☐none ☐mild ☐mod; AIF/VOF adequate ☐yes ☐no.

• Findings:

  • Core (rCBF<30%): ___ mL, location/territory.

  • Penumbra (Tmax≥6 s): ___ mL, territory.

  • Mismatch: ___ mL; ratio ___.

  • CTA correlation (if done): site of occlusion/stenosis.

• Impression: Eligible/not eligible per local criteria; urgent neuro-intervention alert if indicated.