Dose response for brain necrosis. Panel (a) No repair. Panel (b) Monoexponential repair.
Dose response for cervical spinal cord myelopathy. Panel (a) No repair. Panel (b) Monoexponential repair.
Dose response for thoracic spinal cord myelopathy. Panel (a) No repair. Panel (b) Monoexponential repair.
Clinical data used in model fitting for brain necrosis: Incidence of radiation necrosis after different radiation schedules for nasopharyngeal carcinoma. The fraction size and fractions per week represent values for the majority of fractions when schedules are mixed. RN: Radionecrosis.
Clinical data used in model fitting for spinal cord myelopathy: Incidence of myelopathy after different radiation schedules for a variety of primary disease sites.
Fitting results: Akaike weights for the 48 curve fits considered here. ME = monoexponential, BE = biexponential, RR = reciprocal repair. See text for definition of parameter sets A through F. The first column of Akaike weights is for a fit using weighting proportional to patient numbers, while the second column uses equal weighting for all data points. The third column considers a weighted fit, but the data points specified at a 10-year time point in Table I are excluded.
Model Parameters: Best-fit model parameters for the different repair models considered here. The ranges reported represent 95% confidence intervals.
Article metrics loading...
Full text loading...