Lung Nodule Growth Rate Calculator

How to use this lung nodule malignancy risk calculator

What this calculator does

This calculator estimates the probability that a CT-detected pulmonary nodule is malignant using the full Brock University prediction model. The model combines patient characteristics with nodule morphology, location, count, and diameter to produce a log-odds value and convert it to a percentage probability. It is intended for reproducing a published statistical estimate, comparing how modeled risk changes when the documented inputs change, and preparing a structured spreadsheet of the calculation. It does not diagnose lung cancer, determine whether a biopsy or PET-CT is appropriate, or replace radiology and pulmonary evaluation. The underlying model was derived from screening CT cohorts; a CTSNet summary of the original PanCan/Brock study describes the derivation and validation cohorts and the main predictors included in the model.

When to use it

Use the calculator when you want to reproduce a Brock-model estimate from a radiology report, compare two documented nodule scenarios using the same model, check the numerical effect of a changed CT characteristic such as size or spiculation, or export the current assumptions and outputs to an XLSX workbook for discussion or record-keeping. For clinical pathways, use the model only in the context of applicable guidance; the British Thoracic Society pulmonary-nodule calculator places Brock estimates alongside other clinical tools and explicitly states that calculator output does not replace clinical judgment.

How to calculate

1. The calculator opens with a ready-to-use demonstration: age 65, female, no emphysema, family history present, spiculation present, upper-lobe location, one 8 mm part-solid nodule. The results and the example Excel workbook are immediately available.
2. Replace the demonstration values with the documented patient and CT findings. Numeric inputs use a period as the decimal separator. Do not enter commas, scientific notation, units, or free text into numeric fields.
3. Read Cancer probability as the model's percentage estimate and Log odds as the logistic-regression score before conversion to a percentage. Both update as soon as the current input set is complete and valid.
4. Select Download Excel to generate a fresh binary XLSX workbook from the current validated controls. The workbook contains Summary and Inputs sheets and stores numeric values as typed spreadsheet cells rather than copied display strings.
5. Select Reset to clear the demonstration and all entered semantic data. Reset deliberately does not restore the example. Excel export is disabled while the cleared state is incomplete and becomes available again as soon as a complete valid set is entered.

Input guide

Age is a required whole number in years from 18 through 100; 65 is the demonstration value. Higher age increases the model's log odds. Entering an age outside the supported range, a decimal age, or a value containing commas or letters is rejected rather than silently reinterpreted.

Sex is a required model category with Female and Male options. The published coefficient set assigns a positive term to Female and zero to Male. This is a statistical encoding from the derivation model, not a general statement about an individual person's diagnosis. Choose the category used by the source record rather than trying to enter a numeric coefficient directly.

Emphysema is required as Yes or No; the demonstration uses No. Selecting Yes adds the model's emphysema coefficient and raises the calculated log odds with all other inputs unchanged. Use the documented clinical or CT finding, and do not infer emphysema from smoking history alone.

Family history is required as Yes or No and refers to family history of lung cancer; the demonstration uses Yes. Selecting Yes adds the published family-history coefficient. A common error is to substitute any family cancer history for the lung-cancer-specific variable.

Spiculation is required as Yes or No and represents a spiculated nodule margin; the demonstration uses Yes. With other inputs fixed, Yes raises the estimated probability. This field should follow the radiology description rather than a visual guess from an image preview.

Upper lung nodule is required as Yes or No; the demonstration uses Yes. Choose Yes when the modeled nodule is in an upper lobe. The model assigns a positive coefficient to upper-lobe location. Do not use this field to describe whether the nodule appears high or low on a single displayed CT slice.

Nodule count is a required whole number of pulmonary nodules and must be at least 1; the demonstration uses 1. The Brock equation includes the total nodule count as a separate term, so the count should correspond to the modeled examination rather than always being set to one. Fractions, zero, negative values, and formatted strings are rejected.

Nodule size is the required maximum nodule diameter in millimeters, from 1 through 30 mm; the demonstration uses 8 mm. Larger diameter generally increases the modeled probability because the equation applies a nonlinear size transformation. Enter millimeters only. Values above 30 mm fall outside the nodule-size range used by this interface and should not be forced into the model.

Nodule type is required and offers Nonsolid / ground-glass, Part-solid, and Solid; the demonstration uses Part-solid. Each category has its published model coefficient. Use the CT attenuation category in the report. Do not treat these choices as interchangeable descriptions of growth rate or tissue diagnosis.

Output guide

Cancer probability is the primary result, shown as a percentage to two decimal places. It is driven by every input and is a statistical estimate from the Brock logistic model. A value near zero means the equation produces low modeled probability for that input combination; a higher value means higher modeled probability, but no percentage by itself establishes benignity or malignancy. Management thresholds come from separate clinical guidance and patient context.

Log odds is the unrounded internal logistic-regression score shown to four decimal places. Negative log odds correspond to probabilities below 50%, zero corresponds to 50%, and positive values correspond to probabilities above 50%. This is an exact mathematical intermediate of the implemented equation for the entered inputs, while its medical interpretation remains model-dependent.

Worked example

For the first-open demonstration, the model uses age 65; female sex; no emphysema; family history present; spiculation present; upper-lobe location; one nodule; 8 mm maximum diameter; and part-solid type. The size transformation is based on (8 ÷ 10) ^{– 0.5}. After adding the published patient and nodule coefficients and the model intercept, the log odds are – 1.2567. Applying the logistic conversion, probability = 100 × e^{log odds} ÷ (1 + e^{log odds}), gives 22.15%. Those same canonical values populate the initial results and the startup XLSX workbook.

Interpretation and limitations

The risk estimate should be read alongside the imaging pattern and the clinical setting. Nodule size, morphology, and upper-lobe location can materially affect assessment, and management guidance distinguishes solid from subsolid nodules and considers follow-up imaging, PET/CT, tissue sampling, or surveillance according to the complete case. The Fleischner Society 2017 pulmonary-nodule recommendations provide context on size, morphology, follow-up, and the limits of applying a single numerical estimate. A probability from this calculator should therefore be treated as one reproducible model output, not as a treatment recommendation.