Background <p>In the early phase after mechanical thrombectomy, differentiating between intracranial hemorrhage and contrast medium extravasation on CT imaging remains a&#xa0;major diagnostic challenge. Both can appear hyperdense on standard CT scans, however, they have significantly different clinical implications, particularly regarding anticoagulation decisions. The Virtual Non-Contrast (VNC) technique using photon-counting computed tomography (PCCT), offers the potential to reliably distinguish between blood and iodine-based contrast material in a&#xa0;single scan by virtually removing the iodine component.</p> Purpose <p>To evaluate the clinical feasibility and diagnostic performance of the VNC technique in patients after thrombectomy, with focus on the differentiation between hemorrhage and contrast extravasation.</p> Methods <p>In this retrospective study, we analyzed hyperdense areas of 36&#xa0;patients who underwent PCCT imaging shortly after mechanical thrombectomy. Two experienced neuroradiologists independently assessed the presence of intracranial hemorrhage and residual contrast extravasation using the VNC datasets. Findings, including measured volumes, were compared to follow-up scans.</p> Results <p>VNC imaging enabled reliable differentiation between hemorrhage and contrast staining. The sensitivity for detecting hemorrhage was 0.93 (25&#xa0;of 27&#xa0;patients, confidence interval 0.77–0.98), with a&#xa0;specificity of 100%. Contrast extravasation was correctly diagnosed in the 9&#xa0;remaining cases using both VNC and native CT.</p> Conclusion <p>The VNC technique using PCCT demonstrates high clinical utility in the early post-thrombectomy setting, offering reliable differentiation between intracranial hemorrhage and contrast extravasation in the majority of cases. However, isolated cases with extensive contrast extravasation or overlapping imaging features may still pose diagnostic limitations.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Differentiating Hemorrhage and Contrast Extravasation After Mechanical Thrombectomy Using Virtual Non-Contrast Photon-Counting CT

  • Eya Khadhraoui,
  • Roland Schwab,
  • Mathias Becker,
  • Elie Diamandis,
  • Maciej Pech,
  • Daniel Behme,
  • Sebastian Johannes Müller

摘要

Background

In the early phase after mechanical thrombectomy, differentiating between intracranial hemorrhage and contrast medium extravasation on CT imaging remains a major diagnostic challenge. Both can appear hyperdense on standard CT scans, however, they have significantly different clinical implications, particularly regarding anticoagulation decisions. The Virtual Non-Contrast (VNC) technique using photon-counting computed tomography (PCCT), offers the potential to reliably distinguish between blood and iodine-based contrast material in a single scan by virtually removing the iodine component.

Purpose

To evaluate the clinical feasibility and diagnostic performance of the VNC technique in patients after thrombectomy, with focus on the differentiation between hemorrhage and contrast extravasation.

Methods

In this retrospective study, we analyzed hyperdense areas of 36 patients who underwent PCCT imaging shortly after mechanical thrombectomy. Two experienced neuroradiologists independently assessed the presence of intracranial hemorrhage and residual contrast extravasation using the VNC datasets. Findings, including measured volumes, were compared to follow-up scans.

Results

VNC imaging enabled reliable differentiation between hemorrhage and contrast staining. The sensitivity for detecting hemorrhage was 0.93 (25 of 27 patients, confidence interval 0.77–0.98), with a specificity of 100%. Contrast extravasation was correctly diagnosed in the 9 remaining cases using both VNC and native CT.

Conclusion

The VNC technique using PCCT demonstrates high clinical utility in the early post-thrombectomy setting, offering reliable differentiation between intracranial hemorrhage and contrast extravasation in the majority of cases. However, isolated cases with extensive contrast extravasation or overlapping imaging features may still pose diagnostic limitations.