Establishing the Optimal Standard for Preprocessing Head CT Data in Diagnostic Analysis

Abstract

Acute ischemic stroke (AIS) is a significant cause of mortality and disability in adults. Diagnosis involves patient anamnesis, clinical examination, and brain and imaging. Thrombi on non-contrast CT (NCCT) may manifest as the hyperdense artery sign (HAS), while on CT angiography (CTA), an identification of thrombus can be detected as disruption of the contrast-enhanced artery. The preprocessing steps such as image registration are usually required. However, to the best of our knowledge, there is no definitive standard for the pretreatment of head CT data. To adress this, we used the stroke protocol including a multiphase CTA (mCTA) of the head and neck, and a brain perfusion CT. We conducted the registration process using SmartBrain and Elastix programs. Subsequently, the data were fused utilising a time maximum intensity projection technique (tMIP) over contrast phases. The manual thrombi segmentation was done on tMIP data using MITK software.Acute ischemic stroke (AIS) is a significant cause of mortality and disability in adults. Diagnosis involves patient anamnesis, clinical examination, and brain and imaging. Thrombi on non-contrast CT (NCCT) may manifest as the hyperdense artery sign (HAS), while on CT angiography (CTA), an identification of thrombus can be detected as disruption of the contrast-enhanced artery. The preprocessing steps such as image registration are usually required. However, to the best of our knowledge, there is no definitive standard for the pretreatment of head CT data. To adress this, we used the stroke protocol including a multiphase CTA (mCTA) of the head and neck, and a brain perfusion CT. We conducted the registration process using SmartBrain and Elastix programs. Subsequently, the data were fused utilising a time maximum intensity projection technique (tMIP) over contrast phases. The manual thrombi segmentation was done on tMIP data using MITK software.