A multiple regression model for peak skin dose using principal component analysis in interventional radiology.

Area of Science:

• Medical Physics
• Radiology
• Radiation Dosimetry

Background:

• Increasing complexity of interventional radiology procedures leads to higher patient radiation doses.
• Effective patient radiation dose management is crucial but underutilized in many facilities.
• Accurate prediction of peak skin dose (PSD) is essential for radiation safety.

Purpose of the Study:

• To develop and validate models for predicting peak skin dose (PSD) in interventional radiology.
• To assess the utility of dose parameters from digital imaging and communication in medicine-radiation dose structured reports for PSD prediction.
• To compare the predictive accuracy of air kerma (K_a,r) and air kerma area product (KAP) for PSD.

Main Methods:

• Utilized data from radio-photoluminescence glass dosimeters and five dose parameters.
• Employed single and multiple regression analyses to develop PSD prediction models.
• Applied principal component analysis (PCA) to consolidate data and enhance regression models.

Main Results:

• Air kerma (K_a,r) was found to be more accurate in predicting PSD than air kerma area product (KAP).
• Rotational digital subtraction angiography showed a minimal impact on peak skin dose.
• Principal component analysis-enhanced multiple regression models significantly improved PSD prediction accuracy.

Conclusions:

• Predictive models using dose parameters from DICOM-RDSR are feasible for estimating patient radiation doses.
• K_a,r is a more reliable predictor of PSD than KAP in complex interventional procedures.
• PCA-enhanced models offer superior accuracy for peak skin dose estimation, aiding radiation dose management.