Browsing by Author "Kuestner, Thomas"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    End-to-end deep learning nonrigid motion-corrected reconstruction for highly accelerated free-breathing coronary MRA
    (2021) Qi, Haikun; Hajhosseiny, Reza; Cruz, Gastao; Kuestner, Thomas; Kunze, Karl; Neji, Radhouene; Botnar, René Michael; Prieto Vásquez, Claudia
    Purpose: To develop an end-to-end deep learning technique for nonrigid motion-corrected (MoCo) reconstruction of ninefold undersampled free-breathing whole-heart coronary MRA (CMRA).
  • Thumbnail Image
    Item
    Generalized low-rank nonrigid motion-corrected reconstruction for MR fingerprinting
    (WILEY, 2021) Cruz, Gastao; Qi, Haikun; Jaubert, Olivier; Kuestner, Thomas; Schneider, Torben; Michael Botnar, Rene; Prieto, Claudia
    Purpose: Develop a novel low-rank motion-corrected (LRMC) reconstruction for nonrigid motion-corrected MR fingerprinting (MRF).
  • Thumbnail Image
    Item
    Self-supervised learning-based diffeomorphic non-rigid motion estimation for fast motion-compensated coronary MR angiography
    (ELSEVIER SCIENCE INC, 2022) Munoz, Camila; Qi, Haikun; Cruz, Gastao; Kuestner, Thomas; Botnar, Rene M.; Prieto, Claudia
    Purpose: To accelerate non-rigid motion corrected coronary MR angiography (CMRA) reconstruction by developing a deep learning based non-rigid motion estimation network and combining this with an efficient implementation of the undersampled motion corrected reconstruction.
  • Thumbnail Image
    Item
    Whole-heart non-rigid motion corrected coronary MRA with autofocus virtual 3D iNAV
    (ELSEVIER SCIENCE INC, 2022) Schneider, Alina; Cruz, Gastao; Munoz, Camila; Hajhosseiny, Reza; Kuestner, Thomas; Kunze, Karl P.; Neji, Radhouene; Botnar, Rene M.; Prieto, Claudia
    Purpose: Respiratory motion-corrected coronary MR angiography (CMRA) has shown promise for assessing coronary disease. By incorporating coronal 2D image navigators (iNAVs), respiratory motion can be corrected for in a beat-to-beat basis using translational correction in the foot-head (FH) and right-left (RL) directions and in a bin-to-bin basis using non-rigid motion correction addressing the remaining FH, RL and anterior-posterior (AP) motion. However, with this approach beat-to-beat AP motion is not corrected for. In this work we investigate the effect of remaining beat-to-beat AP motion and propose a virtual 3D iNAV that exploits autofocus motion correction to enable beat-to-beat AP and improved RL intra-bin motion correction. Methods: Free-breathing 3D whole-heart CMRA was acquired using a 3-fold undersampled variable-density Cartesian trajectory. Beat-to-beat 3D translational respiratory motion was estimated from the 2D iNAVs in FH and RL directions, and in AP direction with autofocus assuming a linear relationship between FH and AP movement of the heart. Furthermore, motion in RL was also refined using autofocus. This virtual 3D (v3D) iNAV was incorporated in a non-rigid motion correction (NRMC) framework. The proposed approach was tested in 12 cardiac patients, and visible vessel length and vessel sharpness for the right (RCA) and left (LAD) coronary arteries were compared against 2D iNAV-based NRMC. Results: Average vessel sharpness and length in v3D iNAV NRMC was improved compared to 2D iNAV NRMC (vessel sharpness: RCA: 56 +/- 1% vs 52 +/- 11%, LAD: 49 +/- 8% vs 49 +/- 7%; visible vessel length: RCA: 5.98 +/- 1.37 cm vs 5.81 +/- 1.62 cm, LAD: 5.95 +/- 1.85 cm vs 4.83 +/- 1.56 cm), however these improvements were not statistically significant. Conclusion: The proposed virtual 3D iNAV NRMC reconstruction further improved NRMC CMRA image quality by reducing artefacts arising from residual AP motion, however the level of improvement was subject-dependent.