How robust is a passing dose?
A biopsy can meet a dose criterion at one assumed position and miss it at another. Estimate how often the criterion holds across plausible locations.
Research connections
Does the biopsy still meet the dose criterion?
Near a dose threshold, uncertain sampling position can change the result. A pass probability describes how often the criterion holds across plausible locations.
How much room is there above the threshold?
Across 27 biopsies from 15 patients, the nominal margin captured much of the variation in pass probability. For D98, the fitted 95% pass probability required a 6.3 Gy margin above the 20 Gy threshold.
A nominal pass can still be borderline.
Across 27 biopsies, pass probabilities varied with the dose criterion. The coloured regions distinguish confident passes, borderline results and confident failures under the location model.
The required margin depends on the criterion.
Each dose criterion needs its own robustness assessment. A margin estimated for D98 should not be transferred to a different dose or volume threshold.
Interpretation. These are biopsy-assessment margins under the location model, not recommended treatment-dose increases.
Dose differences grow with separation.
The cohort extends the two-core observation: nearby positions tend to have more similar dose than positions farther apart. This motivates the next study's use of spatial correlation to reconstruct dose with attached uncertainty.
Interpretation. This is a descriptive cohort result. GPR fits a separate spatial relationship for each biopsy.
AAPM 2026 poster
Conference version of the threshold-robustness study. References retain their status at presentation.
