TL;DR
This study demonstrates that using an ultra-low dose of contrast medium (224 mgI/kg) for abdominal CT angiography (CTA) is feasible for visualizing pancreatic arteries. By combining advanced virtual monoenergetic imaging (VMI+) with a high-concentration contrast agent, the reduced image quality from the lower dose can be effectively compensated, potentially saving on contrast costs and reducing patient exposure.
What Problem Does This Paper Address?
Abdominal CTA is a crucial tool for diagnosing vascular issues and planning surgeries, particularly for pancreatic conditions. However, the use of contrast media (CM) in CTA is associated with risks like contrast-induced nephropathy and other chemotoxic reactions. While reducing CM dose is desirable, it can compromise the visualization of small vessels, like the pancreatic arteries, which are critical for surgical planning and can be a source of intraoperative bleeding. This study aims to find a way to significantly reduce CM dose while maintaining or improving the visualization of these small, important arteries.
Methods
The researchers conducted an intra-patient study where each participant received both a standard dose of contrast medium (ND-CM, 320 mgI/kg) and an ultra-low dose (LD-CM, 224 mgI/kg) in a randomized order. The key innovation here is the use of advanced virtual monoenergetic imaging (VMI+) applied to the LD-CM scans. VMI+ reconstructs images at specific energy levels (in this case, 40 keV, close to iodine's K-edge) to enhance contrast. This technique was combined with a high-concentration contrast medium (400 mgI/mL). The study compared quantitative metrics (CT attenuation, SNR, CNR) and qualitative scores of image quality and artery depiction across the three imaging scenarios: ND-CM, LD-CM, and VMI+ LD-CM.
Datasets
The study included 41 patients who underwent thoracoabdominal enhanced CT scans for tumor follow-up. These patients served as their own controls, receiving both ND-CM and LD-CM scans. No new datasets were introduced or made public.
Key Insights and Ideas
- VMI+ as a Compensator: The most significant insight is that VMI+ can effectively compensate for the reduced image quality resulting from an ultra-low CM dose. This suggests VMI+ is not just for enhancing standard scans but can be a critical tool for dose reduction strategies.
- High Concentration Matters: Using a high-concentration CM (400 mgI/mL) alongside VMI+ is crucial for achieving adequate iodine delivery rate (IDR) even with a lower total CM volume. This is a key factor in maintaining enhancement.
- Cost and Safety Implications: The ability to use significantly less CM (saving up to half the cost) while maintaining diagnostic quality has direct implications for patient safety and healthcare economics, especially for patients requiring frequent follow-up imaging.
- Pancreatic Arteries are Challenging: The study highlights the difficulty in visualizing small pancreatic arteries even with standard protocols, underscoring the need for advanced techniques when reducing CM.
Key Results
- Quantitative Improvements: VMI+ LD-CM images showed significantly higher mean CT attenuation and CNR values for all arteries compared to both ND-CM and LD-CM images (p < 0.0001).
- Qualitative Equivalence: There were no statistically significant differences in the qualitative scores for abdominal aorta, its main branches, and pancreatic arteries between the ND-CM and VMI+ LD-CM groups (p > 0.05). This means the visual quality was comparable.
- Depiction Rate Trade-offs: While VMI+ compensated for image quality, the depiction rates of some pancreatic arteries were slightly lower in VMI+ LD-CM images compared to ND-CM images. However, they were still comparable to or better than LD-CM alone.
- CM Volume Reduction: The LD-CM group used significantly less CM volume (37.0±6.7 mL) compared to the ND-CM group (52.6±9.4 mL), representing a substantial reduction.
Limitations and Future Work
The authors acknowledge several limitations:
- Disease Progression: Patients with progressive disease or partial response were not excluded, which could theoretically affect artery conditions over time. However, they randomized the scan order to mitigate this.
- Injection Protocol: The study used a constant injection rate rather than a constant injection duration for ND-CM and LD-CM groups. This makes it difficult to perfectly match the time-to-peak enhancement and might not be ideal for all patients. Future studies could explore constant injection duration or patient-specific trigger-delay techniques.
- High BMI Patients: The study did not specifically group patients by Body Mass Index (BMI), so the protocol's effectiveness in patients with high BMI remains to be investigated.
Future work could focus on refining injection protocols for even better time-to-peak matching and evaluating the protocol's performance in specific patient populations like those with high BMI.
Summary updated