PETDetectorPairNormalizationCrossChecker

FS — Functional Specification

1. Run PETDetectorPairNormalizationCrossChecker.exe in demonstration mode or with input.csv output.json arguments.
2. Read UTF-8 CSV and map columns to the data contract.
3. Convert values to the expected types and validate mandatory fields.
4. Perform calculations and rule checks for “PET Detector-Pair Normalization Factor Cross-Check”.
5. Create individual checks and derive the overall status from the worst result.
6. Write machine-readable output.json; populate errors when a valid calculation cannot be produced.

input.csv example

ScannerID,NormalizationID,CheckDate,TotalLORs,MeanVendorFactor,MeanCalculatedFactor,StdDevFactors,MaxFactor,MinFactor,ZeroOrInfiniteCount,DuplicateLORCount
PET-SIEMENS-01,NORM-2026-07-22-V01,2026-07-22,1500000,1.002,1.001,0.05,1.25,0.85,0,0
PET-GE-02,NORM-2026-07-20-V05,2026-07-22,1200000,1.050,1.000,0.18,2.50,0.40,5,0
PET-PHILIPS-03,NORM-2026-07-21-V02,2026-07-22,1400000,0.990,0.995,0.04,1.10,0.90,0,12

Minimum output.json structure

{
  "utility": "PETDetectorPairNormalizationCrossChecker",
  "source": "input.csv",
  "status": "PASS|WARNING|FAIL",
  "derivedValues": {},
  "checks": [
    {
      "parameter": "example",
      "value": 0,
      "limit": "configured profile",
      "status": "PASS",
      "message": "criterion satisfied"
    }
  ],
  "errors": []
}

Minimum OQ scenarios

• A row within the approved profile shall return PASS.
• A non-critical limit violation shall return WARNING or a review status.
• A critical violation shall return FAIL or the applicable operational decision.
• A missing column, invalid type or impossible date shall populate errors and shall not return PASS.

Before operational use

• Reconcile formulas and limits with manufacturer documentation and the approved SOP.
• Fix the scanner model, software version and profile version.
• After an executable, formula or profile change, repeat affected tests.
• Retain the source CSV and JSON result together.