PSIRNet: Deep Learning-based Free-breathing Rapid Acquisition Late Enhancement Imaging
Summary: arXiv:2604.08781v1 Announce Type: cross
Abstract
Purpose: To develop and evaluate a deep learning (DL) method for free-breathing phase-sensitive inversion recovery (PSIR) late gadolinium enhancement (LGE) cardiac MRI that produces diagnostic-quality images from a single acquisition over two heartbeats, eliminating the need for 8 to 24 motion-corrected (MOCO) signal averages.
Materials and Methods
Raw data comprising 800,653 slices from 55,917 patients, acquired on 1.5T and 3T scanners across multiple sites from 2016 to 2024, were used in this retrospective study. The data were split by patient:
- 640,000 slices (42,822 patients) for training
- The remainder for validation and testing, without overlap
The training and testing data were sourced from different institutions. PSIRNet, a physics-guided DL network with 845 million parameters, was trained end-to-end to reconstruct PSIR images with surface coil correction from a single interleaved IR/PD acquisition over two heartbeats. Reconstruction quality was evaluated using:
- SSIM (Structural Similarity Index)
- PSNR (Peak Signal-to-Noise Ratio)
- NRMSE (Normalized Root Mean Square Error)
These metrics were compared against MOCO PSIR references. Additionally, two expert cardiologists performed an independent qualitative assessment, scoring image quality on a 5-point Likert scale across bright blood, dark blood, and wideband LGE variants. Paired superiority and equivalence (margin = 0.25 Likert points) were tested using exact Wilcoxon signed-rank tests at a significance level of 0.05 using R version 4.5.2.
Results
Both readers rated single-average PSIRNet reconstructions superior to MOCO PSIR for dark blood LGE (conservative P = .002). For bright blood and wideband, one reader rated it superior while the other confirmed equivalence (all P < .001). Inference time required approximately 100 milliseconds per slice compared to more than 5 seconds for MOCO PSIR.
Conclusion
PSIRNet demonstrates the capability to produce diagnostic-quality free-breathing PSIR LGE images from a single acquisition, enabling an 8- to 24-fold reduction in acquisition time. This advancement could significantly improve patient throughput and comfort during cardiac MRI procedures, representing a notable stride in the field of medical imaging.