An end‐to‐end AI‐based framework for automated discovery of rapid CEST/MT MRI acquisition protocols and molecular parameter quantification (AutoCEST)

Abstract

To develop an automated machine-learning-based method for the discovery of rapid and quantitative chemical exchange saturation transfer (CEST) MR fingerprinting acquisition and reconstruction protocols.Methods: An MR physics-governed AI system was trained to generate optimized acquisition schedules and the corresponding quantitative reconstruction neural network. The system (termed AutoCEST) is composed of a CEST saturation block, a spin dynamics module, and a deep reconstruction network, all differentiable and jointly connected. The method was validated using a variety of chemical exchange phantoms and in vivo mouse brains at 9.4T. Results:The acquisition times for AutoCEST optimized schedules ranged from 35 to 71 s, with a quantitative image reconstruction time of only 29 ms. The resulting exchangeable proton concentration maps for the phantoms were in good agreement with the known solute concentrations for AutoCEST sequences (mean absolute error = 2.42 mM; Pearson's r = 0.992, p < 0.0001), but not for an unoptimized sequence (mean absolute error = 65.19 mM; Pearson's r = − 0.161, p = 0.522). Similarly, improved exchange rate agreement was observed between AutoCEST and quantification of exchange using saturation power (QUESP) methods (mean absolute error: 35.8 Hz, Pearson's r = 0.971, p < 0.0001) compared to an unoptimized schedule and QUESP (mean absolute error = 58.2 Hz; Pearson's r = 0.959, p < 0.0001). The AutoCEST in vivo mouse brain semisolid proton volume fractions were lower in the cortex (12.77% ± 0.75%) compared to the white matter (19.80% ± 0.50%), as expected.

calf – Software for CEST Analysis with Lorentzian Fitting

“…Recent studies by Cohen et al and Perlman et al have demonstrated the potential of deep learning for CEST imaging, but may not be accessible or transparent to a broader, non-technical audience [ 42 , 43 ]. However, we see the potential to incorporate AI-based methods into our system in the future to improve the performance and efficiency of the analysis.…”

Noninvasively differentiating acute and chronic nephropathies via multiparametric urea‐CEST, nuclear Overhauser enhancement‐CEST, and quantitative magnetization transfer MRI

“…Several acceleration methods such as compressed sensing algorithms and deep learning approaches may be considered to overcome this issue. [57][58][59][60] Some limitations of this study should be noted. Although we aimed to observe the progression of renal injuries toward the chronic end using the cisplatin model, histological findings suggest insufficient fibrosis development compared to previous mouse CKD models.…”

Molecular MRI-Based Monitoring of Cancer Immunotherapy Treatment Response

“…Loktyushin et al proposed a supervised learning framework to discover MRI sequences for different tasks, including k-space trajectory and flip angle optimization, specific absorption rate (SAR) mitigation, and quantitative T mapping [ 204 ]. Perlman et al developed an end-to-end framework for generating rapid (less than 1 min) CEST acquisition protocols while simultaneously training a reconstruction network that extracts quantitative molecular maps from the raw data [ 205 ]. Kang et al developed a learning-based approach for the optimization of semisolid magnetization transfer MRF protocols [ 206 ] and demonstrated its applicability on human subjects.…”