Computer-Assisted Quality Assessment of Aortic CT Angiographies for Patient-Individual Dose Adjustment

Area of Science:

• Radiology
• Medical Imaging
• Nephrology

Background:

• Iodine-containing contrast agents (CA) enhance CT angiography (CTA) image quality but pose risks, particularly kidney damage.
• Current clinical practice often uses a standard CA dose, disregarding individual patient variations and leading to potential overdosing.

Purpose of the Study:

• To develop and evaluate an automated method for personalized CA dose adjustment in CTA.
• To identify excessive CA administration and recommend optimal dosages based on individual patient parameters.

Main Methods:

• Assessing image contrast in CTA images to detect CA overdosing.
• Implementing an automatic classification system using patient-specific clinical parameters to recommend CA doses.
• Utilizing a k-NN classifier with a leave-one-out method for evaluating classification accuracy.

Main Results:

• The study identified that a majority of patients receive excessive CA doses.
• The automated classification system achieved an overall error rate of 22% for CA dose adjustment recommendations.
• The positive predictive value for correctly identifying CA overdosing was 85.71%.

Conclusions:

• An automated, personalized approach to CA dosing in CTA is feasible and can help prevent overdosing.
• Individualized CA dose recommendations based on patient parameters are crucial for patient safety and effective imaging.
• Further refinement of the classification algorithm may improve accuracy and clinical implementation.