Clinical Scorecard: Automated Assessment of Net Water Uptake in Acute Ischemic Stroke Utilizing Non-Contrast CT Imaging

At a Glance

Key Highlights

• NWU is a prognostic biomarker for AIS that can be calculated from NCCT hypodensity without reliance on CTP or DWI.
• The proposed method uses a fully automated, explainable image processing pipeline avoiding deep learning, enhancing reproducibility and interpretability.
• Validation was performed on two datasets comparing NWU maps to expert-annotated lesions from CTP and DWI, demonstrating feasibility of NCCT-only NWU estimation.

Guideline-Based Recommendations

Diagnosis

• Use NCCT imaging within 6 hours of symptom onset for initial assessment of ischemic stroke.
• Consider automated NWU quantification from NCCT as a biomarker to estimate infarct volume and tissue viability.
• Complement NCCT NWU assessment with expert-annotated lesion masks from CTP or DWI when available for validation.

Management

• Utilize NWU values to help distinguish reversible from irreversible ischemic tissue, potentially refining treatment decisions.
• Adopt NCCT-only NWU quantification to reduce radiation exposure and streamline imaging workflows following ALARA principles.

Monitoring & Follow-up

• Monitor infarct progression using serial NCCT scans and NWU quantification to assess treatment response and tissue viability.

Risks

• Be aware of limitations in early infarct lesion visibility on NCCT due to subtle hypodensity and limited contrast.
• Exclude cases with poor image quality or very small lesion volumes (<1 mL) to ensure reliable NWU quantification.

Patient & Prescribing Data

Patients with acute ischemic stroke confirmed by large vessel occlusion and visible early ischemic lesions on imaging.

Automated NWU quantification from NCCT may aid in identifying patients with salvageable brain tissue, potentially influencing eligibility for reperfusion therapies.

Clinical Best Practices

• Perform NCCT imaging promptly within 6 hours of symptom onset for accurate NWU assessment.
• Use automated, explainable image processing pipelines to quantify NWU, minimizing operator dependence and analysis time.
• Validate NWU-based lesion maps against expert annotations from CTP or DWI when possible to ensure accuracy.
• Exclude images with artifacts or low signal-to-noise ratio to maintain data quality.
• Apply affine image registration to align lesion masks from different modalities to NCCT space for consistent analysis.

References

• Infarct lesion volume assessment in clinical trials
• Limitations of computed tomography perfusion in ischemic core estimation
• Overestimation of irreversible damage by CTP
• Net water uptake as a prognostic biomarker in AIS
• Diffusion-weighted imaging in stroke lesion delineation
• Challenges in CTP-based infarct segmentation
• Subjectivity in CTP interpretation
• Manual infarct delineation limitations
• ALARA principle in radiation exposure
• Challenges in infarct volume estimation from NCCT
• Kumar et al. infarct volume estimation using CBF maps
• Semi-automatic deep learning methods for NCCT infarct segmentation
• Limitations of deep learning in infarct segmentation
• Nowinski et al. stroke imaging marker-based techniques
• Ethical standards and institutional review board approvals
• Affine image registration for lesion mask mapping