TL;DR
This paper introduces a novel technique, time maximum intensity projection (t-MIP), applied to dynamic CT angiography (dyn-CTA) for a faster and more accurate diagnosis of lower leg peripheral arterial disease (PAD). By merging all phases of dyn-CTA, t-MIP significantly improves vascular attenuation and shows strong correlation with standard CTA (s-CTA) in assessing stenosis, offering a more efficient clinical workflow.
What Problem Does This Paper Address?
Peripheral arterial disease (PAD) affecting the lower extremities is often underdiagnosed due to a wide range of symptoms and a high prevalence of asymptomatic cases. While computed tomographic angiography (CTA) is a common diagnostic tool, standard CTA (s-CTA) can face challenges with accurate stenosis assessment, especially in cases of severe calcification, inaccurate contrast timing due to long vessel distances, or asymmetric stenoses. Dynamic CTA (dyn-CTA) offers improved contrast enhancement and diagnostic confidence but requires reviewing multiple time phases, making it time-consuming and limiting its clinical utility. This study aims to address the need for a faster, yet accurate, method to evaluate lower leg vascular stenosis in PAD patients.
Methods
The study introduces and evaluates a post-processing technique called time maximum intensity projection (t-MIP) applied to dynamic CT angiography (dyn-CTA) data. Dyn-CTA scans were acquired using a dual-source CT system with a 45 cm scan range and 9 phases. The t-MIP technique, as illustrated in Figure 1, involves merging the maximum intensity value from each matrix across all time points of the dyn-CTA acquisition. This process automatically generates images that capture the optimal enhancement of the lower leg arteries, regardless of their specific peak enhancement times.
For comparison, standard CTA (s-CTA) scans were also performed. The study involved both objective and subjective analyses. Objective analysis focused on measuring CT attenuation in specific arterial segments (anterior tibial, posterior tibial, and fibular arteries) for t-MIP, dyn-CTA (best enhancement phase), and s-CTA. Subjective analysis involved evaluating vessel stenosis and occlusion using a "runoff score" assigned by two radiologists. This score ranged from 0 to 9, with higher scores indicating more severe disease. The study also assessed interobserver agreement and performed Bland-Altman analysis to compare the consistency of runoff scores between t-MIP and s-CTA.
Datasets
The study utilized data from 34 patients with known PAD, encompassing 68 lower extremities and 204 vascular segments. These patients were scanned using a combined s-CTA and dyn-CTA protocol. The data was collected at Peking Union Medical College Hospital between November 2015 and March 2016. No new benchmark datasets were introduced; the study evaluated existing patient data.
Key Insights and Ideas
- t-MIP as a "best of all worlds" approach: By merging all dynamic phases, t-MIP effectively captures the optimal contrast enhancement across different arterial segments, overcoming the timing challenges inherent in single-phase CTA.
- Improved vascular attenuation: t-MIP CTA demonstrated significantly higher CT attenuation in key lower leg arteries compared to both the best enhancement phase of dyn-CTA and s-CTA. This suggests better visualization of the vessels.
- Strong correlation with established methods: The runoff scores derived from t-MIP CTA showed a strong correlation with those from s-CTA, indicating that t-MIP can reliably assess the severity of vascular stenosis.
- Efficiency gain: While dyn-CTA offers better image quality, its multiphase review is time-consuming. t-MIP, by providing a single, optimized image set, offers a much quicker evaluation, making it more clinically practical.
- Potential for early detection: The study's findings suggest t-MIP could be a valuable tool for assisting in the diagnosis of lower leg vascular stenosis, potentially leading to earlier intervention.
Key Results
- CT Attenuation: t-MIP CTA showed significantly higher CT attenuation in the posterior tibial artery and fibular artery compared to s-CTA (p < 0.05). The average CT attenuation across all three segments was also higher for t-MIP than for s-CTA.
- Runoff Scores: There were strong correlations (r ≥ 0.75, p < 0.05) between the runoff scores evaluated by t-MIP and s-CTA for both radiologists. Bland-Altman analysis showed good agreement, with mean differences of 1.79 and 1.38 for the two radiologists.
- Interobserver Agreement: Good interobserver agreement was found for the runoff scores assigned to t-MIP images (Cronbach's alpha = 0.833).
- Clinical Stage Correlation: Runoff scores from both t-MIP and s-CTA were higher in patients with more severe clinical stages of PAD (Fontaine stage III+IV), indicating that the scoring system reflects disease severity.
Limitations and Future Work
The authors acknowledge several limitations:
- Small Patient Population: The study's small sample size, particularly the limited number of patients with severe clinical stages (Fontaine stage III and IV), restricts subgroup analysis and generalizability.
- Lack of Gold Standard: A gold-standard analysis (e.g., surgical findings or invasive angiography) was not performed for comparison, which is a limitation given the pilot nature of the study.
- Muscle CT Attenuation: The t-MIP technique, while enhancing arterial visualization, also selects muscle CT attenuation. This mild muscle enhancement might slightly reduce the contrast-to-noise ratio of the t-MIP images.
Future work should involve larger, multi-center studies, especially focusing on patients with more advanced PAD, to further validate the findings and explore the full clinical utility of t-MIP in PAD diagnosis and treatment planning.
Summary updated