Background: Stroke is a leading cause of disability and death worldwide1, necessitating accurate and efficient diagnostic imaging techniques for prompt clinical management and research. Magnetic Resonance Imaging (MRI) is a vital tool for assessing stroke lesions, and its combination with Positron Emission Tomography (PET) can provide valuable information about tissue perfusion and metabolism. MAGiC (Magnetic resonance image compilation), as one type of synthetic MRI, is an innovative and rapid quantitative MRI technique recently introduced. This method enables the simultaneous acquisition of relaxometry measures T1, T2 relaxation time and proton-weighted (PD) through multi-echo and multi-delay acquisition method. With multiple relaxometry maps, MAGiC allows for tailored contrast optimization to enhance stroke lesion delineation2, 3. Moreover, the quantitative relaxation values within the lesion area offer a potential avenue for predicting PET SUV, which can help in characterizing metabolic activity within the stroke lesion. The purpose of this study was to explore the feasibility of applying synthetic MRI technology to improve stroke lesion characterization and to predict regional metabolic activity of the lesion area.
Methods: We conducted a pilot study on ten stroke patients. Each patient underwent Conventional T2-FLAIR, MAGiC and PET scanning with a 3.0 T scanner (SIGNA PET/MR, GE Healthcare). MAGiC offers flexibility in adjusting parameters such as repetition time (TR), echo time (TE), and inversion time (TI) to optimize imaging contrast for specific targets. In this study, we extracted the T1 value of the stroke lesion and surrounding tissue from the T1 map generated from MAGiC. Then, based on T1 measurements, two inversion recovery (IR) images were generated using different TI values to suppress stroke lesion (TI=800ms) and surrounding tissue (TI=500ms), respectively (Figure 1A). The two IR images were merged to create a composite image that enhanced stroke lesion contrast while maintaining anatomical context. These images were compared to traditional T2 FLAIR images for tissue contrast. Additionally, we leveraged the quantitative relaxation values from the lesion areas to build predictive models for PET Standardized Uptake Values (SUV), enabling the prediction of metabolic activity within the stroke lesion region (Figure 2A).
Results: The MAGiC-generated images provided significantly higher tissue contrast compared to conventional T2 FLAIR images (p<0.01, Figure 1B), allowing for improved lesion visualization. This increased contrast facilitated more accurate delineation of stroke lesions, enhancing their visibility for clinical diagnosis and research purposes. Additionally, quantitative relaxation values derived from the lesion areas in the T1, T2, and PD maps were utilized to establish a predictive model for PET Standardized Uptake Values (SUV). This model effectively predicted regional metabolic activity within the stroke lesion area (Figure 2B).
Discussion: The findings of this study demonstrate the potential of synthetic MRI using the MAGiC sequence in stroke patients. The improved tissue contrast provided by the generated images enhances the visualization of stroke lesion areas. Additionally, the prediction model based on the relaxation values offers a non-invasive method to estimate the metabolic level in the stroke regions. These findings contribute to the understanding and management of stroke patients, potentially leading to improved diagnostic accuracy and treatment strategies. However, further validation and clinical exploration with larger patient cohorts are still required.

Acknowledgements

No acknowledgement found.