Implementation of an in vivo diode dosimetry program and changes in diode characteristics over a 4-year clinical history

An in vivo dosimetry system that used n-type semiconductor diodes with integral buildup caps was introduced into the clinic. Measurements were made on the entrance surface of the patient and were compared to calculated diode readings expected from monitor units delivered by each beam. A method is given for calibration and correction for changes in diode sensitivity, dose-per-pulse effects, collimated field-size (headscatter factor), wedges, compensators, and scatter from blocks and block trays. Clinically relevant temperature corrections are determined based on temperature measurements made with the diode used as a thermistor. Changes in diode characteristics over 4 years of clinical use are presented. With proper correction for clinical variables it is shown that difference between calculated and measured diode readings are within ±1% for phantom measurements and within ±3% for clinical measurements at a 95% confidence level. The correlation of dose measurements on the patient surface to dose inside a target volume is discussed.