Description
Phantom CNR using an explicitly supplied common noise standard deviation; no modality-dependent noise estimator or quality threshold.
An image grid has array indices as well as physical spacing, timing, and signal conventions. A calculation on supplied metadata describes one part of that grid or signal; it does not inspect an image, validate a complete file, or interpret anatomy.
Phantom CNR using an explicitly supplied common noise standard deviation; no modality-dependent noise estimator or quality threshold.
When to use this tool
Use it for radiology phantom contrast noise term when the required inputs and conventions are already available.
How to use the tool and how it works
- Enter firstMean, secondMean, noiseSd. Use the stated units and boundary conventions.
- Run the calculation. Correct any input error before interpreting a result.
- Read contrast, cnr, signedCnr. Inspect any known, defined, inclusion, or boundary flags before interpreting numeric values.
Require positive noiseSd. Subtract secondMean from firstMean for signed contrast, divide that difference by noiseSd for signedCnr, and divide its absolute value by noiseSd for cnr.1
| Input field | Meaning and units | Accepted input |
|---|---|---|
| firstMean | First Mean | Required. Finite number. |
| secondMean | Second Mean | Required. Finite number. |
| noiseSd | Noise Sd | Required. Finite number; strictly above 0. |
| Output field | Interpretation | Returned shape |
|---|---|---|
| contrast | Contrast | Required. Finite number. |
| cnr | Cnr | Required. Finite number. |
| signedCnr | Signed Cnr | Required. Finite number. |
{
"input": {
"firstMean": 120,
"secondMean": 100,
"noiseSd": 5
},
"output": {
"contrast": 20,
"cnr": 4,
"signedCnr": 4
}
}Limitations and assumptions
- The caller supplies one noise standard deviation in the same signal units as both means. No noise estimator, region selection, diagnostic threshold, or protocol-specific contrast-to-noise definition is inferred.
- The field tables describe structural bounds; the mechanics also state cross-field requirements. Malformed input shapes, disallowed keys, and invalid domains cause input errors. Results must remain finite and within the declared output contract. A successful false flag or zero placeholder is distinct from an error.
- Numeric arithmetic uses binary floating point. Exact comparisons have no hidden tolerance or decimal rounding; negative zero is exported as zero. Decimal-text integer operations retain the exactness specifically stated above.
Alternative or Complementary methods
For radiology phantom contrast noise term, check the source records and the supplied rule before comparing the result with another method. Use a full image reader and geometry or signal analysis when the task requires interpreting image data.
References
Similar or alternative tools
- Radiology Phantom Uniformity One Pair Term
One extrema-based uniformity convention 1-(max-min)/(max+min), with no modality acceptance threshold or automatic ROI selection.
- Radiology Bench Dose Area Unit Factor
Bench dose-area product unit arithmetic using explicit gray and square-meter factors, without patient dose inference or exposure recommendations.