Iterative Reconstruction Improves Both Objective and Subjective Image Quality in Acute Stroke CTP

Area of Science:

• Radiology
• Medical Imaging
• Neurology

Background:

• Computed tomography perfusion (CTP) is crucial for acute ischemic stroke (AIS) diagnosis.
• Image noise in standard-dose CTP can lead to measurement errors, impacting diagnostic accuracy.
• Iterative reconstruction (IR) algorithms offer potential solutions for noise reduction in medical imaging.

Purpose of the Study:

• To evaluate the effectiveness of iterative reconstruction (IR) algorithms in improving the diagnostic value of standard-dose CTP for AIS.
• To assess the impact of IR on image quality metrics and diagnostic performance in AIS patients.

Main Methods:

• Twenty-three AIS patients underwent standard-dose CTP.
• Raw CTP data were reconstructed using filtered back projection (FBP) and IR algorithms at varying intensity levels.
• Image quality was assessed objectively (SNR, CNR, perfusion values) and subjectively.
• Discriminative power for infarct prediction was determined using receiver operating characteristic (ROC) curves.

Main Results:

• IR significantly improved objective (SNR, CNR) and subjective image quality with increasing intensity levels.
• IR led to higher mean transit time (MTT) in ischemic lesions, without significant changes in cerebral blood volume (CBV) or cerebral blood flow (CBF).
• IR increased the discriminative power (AUC) for infarct prediction in CBF and MTT, enhancing diagnostic accuracy.

Conclusions:

• Iterative reconstruction (IR) algorithms enhance both objective and subjective image quality in standard-dose CTP for acute stroke patients.
• Improved CTP image quality with IR increases confidence in acute stroke triage and diagnosis.
• IR shows promise in improving the prediction of infarct outcomes in AIS.