An objective scoring method to evaluate image quality of middle cerebral artery Doppler

Objective: To validate an objective scoring system for middle cerebral artery (MCA) pulsed wave Doppler images. Method: From an image database of routine 36-week scans, a random sample of MCA Doppler images was selected. Two reviewers rated the images subjectively as acceptable or unacceptable. Subsequently they used an objective 6-point image scoring system and awarded one point for each of the following: (1) anatomical site, (2) magnification, (3) angle of insonation, (4) image clarity, (5) sweep speed adjustment, and (6) velocity scale and baseline adjustment. Image scores 4-6 were defined as good quality whereas 0-3 as poor. The subjective and objective agreement between the two reviewers was compared using the adjusted Kappa statistic. Results: A total of 124 images were assessed. Using objective scoring the agreement rate between reviewers increased to 91.9% (kappa = 0.839) compared to subjective agreement 75.8% (kappa = 0.516). The agreement for each criterion was: anatomical site 91.1% (kappa = 0.823), magnification 95.2% (kappa = 0.903), clarity 83.9% (kappa = 0.677), angle 96.0% (kappa = 0.919), sweep speed 98.4% (kappa = 0.968), and velocity scale and baseline 94.4% (kappa = 0.887). Conclusion: Objective assessment of MCA Doppler images using a 6-point scoring system has greater interobserver agreement than subjective assessment and could be used for MCA Doppler quality assurance.