MR fluoroscopy: Technical feasibility

TLDR

The acquisition time per image normally exceeds the interval between successive images, creating a temporal lag. The study proposes a magnetic‑resonance imaging method that enables high imaging rates and rapid reconstruction. The method modifies the FLASH sequence by acquiring a limited number N of phase encodings, applying them sequentially, periodically, or continuously, and reconstructing images via a sliding window of width N that updates each new echo, allowing reconstruction times of about 1 s and display rates of 12.5 images/s in phantom experiments, and a hardware implementation can achieve real‑time acquisition. Experimental phantom studies demonstrated a 12.5 images/s display rate and real‑time reconstruction capability, confirming the sliding‑window technique’s suitability for high‑speed MR imaging. © 1988 Academic Press, Inc.

Abstract

Abstract A method of magnetic resonance image acquisition and reconstruction is described in which high imaging rates and fast reconstruction times are allowed. The acquisition is a modification of the basic FLASH sequence but with a restricted number N of phase encodings. The encodings are applied sequentially, periodically, and continuously. Images are formed by sliding a window of width N encodings along the acquired data and reconstructing an image for each position of the window. In general the acquisition time per image exceeds the time between successive images, and the method thus has a temporal lag. Experimental studies were performed with a dynamic phantom using 48 phase encodings and a TR of 20 ms, for an image acquisition time of about 1 s. The image display rate in the reconstructed sequence was 12.5 images/s, and the image sequence portrayed the motion of the phantom. Additional studies were done with 24 encodings. It is shown how the sliding window technique lends itself to high‐speed reconstruction, with each newly acquired echo used to quickly update the image on display. The combination of the acquisition technique described and a hardware implementation of the reconstruction algorithm can result in realtime MR image acquisition and reconstruction. © 1988 Academic Press, Inc.