Advancing fMRI Acquisition through Dissemination of EPTI- An Efficient Distortion-Free Multi-Contrast Imaging Technology

PROJECT SUMMARY / ABSTRACT Functional MRI (fMRI) is today the predominant tool for noninvasive imaging of brain function, which has revolutionized our understanding of the human brain. To date, echo-planar imaging (EPI) has been the standard fMRI acquisition method, but suffers from intrinsic limitations such as static and dynamic distortion, image/contrast blurring, signal voids, suboptimal CNR and physiological noises. These limitations compromise the sensitivity and reliability of task and resting-state fMRI, and hinder the effective spatial resolution and achievable functional specificity for state-of-the-art high-resolution laminar or columnar fMRI studies, especially given that these problems become more severe at higher resolutions and field strengths. Echo-planar time-resolved imaging (EPTI) has recently been introduced to address these limitations. Conventional image formation generates a single echo image from k-space samples evolving across time with accumulated imperfections. EPTI recognizes the strong spatiotemporal correlation of the data during the image encoding process, and exploits it to replace conventional image formation and resolve multi-echo images from the fully-recovered spatiotemporal data, making the acquisition immune to these imperfections. This turns a deficit – the evolution of the MR signal with time – into an asset. It achieves high SNR/CNR efficiency using continuous readout with minimal deadtime, while providing images completely free from both static and dynamic distortions caused by field inhomogeneity and its variations due to subject motion. The multi-echo images sampled at a wide range of TEs (e.g., ~4–70 ms) not only allow for optimal CNR across the whole brain and mitigated signal dropout at challenging short T2* regions, but also enable effective removal of unwanted physiological noise. Moreover, they provide pure contrast at the exact TEs with minimal contaminations that can be used to improve the specificity for high-resolution laminar or columnar fMRI studies. These improvements provided by EPTI have been demonstrated at both 3T and 7T in a variety of applications. The goal of this project is therefore to broadly disseminate EPTI as the next-generation fMRI acquisition tool. Despite its demonstrated high value, EPTI has only been used in a small number of sites due to the lack of available pulse sequence and image reconstruction software and the challenge of cross-vendor implementation. In this project, we will overcome these barriers by assembling and disseminating the whole EPTI acquisition, reconstruction and pre-processing package developed in different vendor platforms (Siemens and GE) and in open-source and cross-vendor frameworks accessible to multiple other vendors. The sequence will be refined by incorporating multiple functionalities, and integrated with the optimized reconstruction and pre-processing in a user-friendly workflow. We will also leverage the existing resources and mechanisms at Martinos Center for technology dissemination and quality assurance, which have been refined over the years through large-scale studies (e.g., HCP, ABCD), to maximize the impact of this technology and ensure its sustainability.