Ischemic stroke enhancement using a variational model and the expectation maximization method

Abstract

Objectives: In order to enable less experienced physicians to reliably detect early signs of stroke, A novel approach was proposed to enhance the visual perception of ischemic stroke in non-enhanced CT. Methods: A set of 39 retrospective CT scans were used, divided into 23 cases of acute ischemic stroke and 16 normal patients. Stroke cases were obtained within 4.5 h of symptom onset and with a mean NIHSS of 12.9±7.4. After selection of adjunct slices from the CT exam, image averaging was performed to reduce the noise and redundant information. This was followed by a variational decomposition model to keep the relevant component of the image. The expectation maximization method was applied to generate enhanced images. Results: We determined a test to evaluate the performance of observers in a clinical environment with and without the aid of enhanced images. The overall sensitivity of the observer’s analysis was 64.5 % and increased to 89.6 % and specificity was 83.3 % and increased to 91.7 %. Conclusion: These results show the importance of a computational tool to assist neuroradiology decisions, especially in critical situations such as the diagnosis of ischemic stroke. Key Points: • Diagnosing patients with stroke requires high efficiency to avoid irreversible cerebral damage. • A computational algorithm was proposed to enhance the visual perception of stroke. • Observers’ performance was increased with the aid of enhanced images.