@article{DawoodBreuerStebanietal.2023,
  author    = {Dawood, Peter and Breuer, Felix and Stebani, Jannik and Burd, Paul and Homolya, Istv{\´a}n and Oberberger, Johannes and Jakob, Peter M. and Blaimer, Martin},
  title     = {Iterative training of robust k-space interpolation networks for improved image reconstruction with limited scan specific training samples},
  series = {Magnetic Resonance in Medicine},
  volume    = {89},
  journal   = {Magnetic Resonance in Medicine},
  number    = {2},
  doi       = {10.1002/mrm.29482},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-312306},
  pages     = {812 -- 827},
  year      = {2023},
  abstract  = {To evaluate an iterative learning approach for enhanced performance of robust artificial-neural-networks for k-space interpolation (RAKI), when only a limited amount of training data (auto-calibration signals [ACS]) are available for accelerated standard 2D imaging. Methods In a first step, the RAKI model was tailored for the case of limited training data amount. In the iterative learning approach (termed iterative RAKI [iRAKI]), the tailored RAKI model is initially trained using original and augmented ACS obtained from a linear parallel imaging reconstruction. Subsequently, the RAKI convolution filters are refined iteratively using original and augmented ACS extracted from the previous RAKI reconstruction. Evaluation was carried out on 200 retrospectively undersampled in vivo datasets from the fastMRI neuro database with different contrast settings. Results For limited training data (18 and 22 ACS lines for R = 4 and R = 5, respectively), iRAKI outperforms standard RAKI by reducing residual artifacts and yields better noise suppression when compared to standard parallel imaging, underlined by quantitative reconstruction quality metrics. Additionally, iRAKI shows better performance than both GRAPPA and standard RAKI in case of pre-scan calibration with varying contrast between training- and undersampled data. Conclusion RAKI benefits from the iterative learning approach, which preserves the noise suppression feature, but requires less original training data for the accurate reconstruction of standard 2D images thereby improving net acceleration.},
  language  = {en}
}
@article{StebaniBlaimerZableretal.2023,
  author    = {Stebani, Jannik and Blaimer, Martin and Zabler, Simon and Neun, Tilmann and Pelt, Dani{\"e}l M. and Rak, Kristen},
  title     = {Towards fully automated inner ear analysis with deep-learning-based joint segmentation and landmark detection framework},
  series = {Scientific Reports},
  volume    = {13},
  journal   = {Scientific Reports},
  doi       = {10.1038/s41598-023-45466-9},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-357411},
  year      = {2023},
  abstract  = {Automated analysis of the inner ear anatomy in radiological data instead of time-consuming manual assessment is a worthwhile goal that could facilitate preoperative planning and clinical research. We propose a framework encompassing joint semantic segmentation of the inner ear and anatomical landmark detection of helicotrema, oval and round window. A fully automated pipeline with a single, dual-headed volumetric 3D U-Net was implemented, trained and evaluated using manually labeled in-house datasets from cadaveric specimen (N = 43) and clinical practice (N = 9). The model robustness was further evaluated on three independent open-source datasets (N = 23 + 7 + 17 scans) consisting of cadaveric specimen scans. For the in-house datasets, Dice scores of 0.97 and 0.94, intersection-over-union scores of 0.94 and 0.89 and average Hausdorf distances of 0.065 and 0.14 voxel units were achieved. The landmark localization task was performed automatically with an average localization error of 3.3 and 5.2 voxel units. A robust, albeit reduced performance could be attained for the catalogue of three open-source datasets. Results of the ablation studies with 43 mono-parametric variations of the basal architecture and training protocol provided task-optimal parameters for both categories. Ablation studies against single-task variants of the basal architecture showed a clear performance beneft of coupling landmark localization with segmentation and a dataset-dependent performance impact on segmentation ability.},
  language  = {en}
}
@article{BillerCholiBlaimeretal.2014,
  author    = {Biller, Armin and Choli, Morwan and Blaimer, Martin and Breuer, Felix A. and Jakob, Peter M. and Bartsch, Andreas J.},
  title     = {Combined Acquisition Technique (CAT) for Neuroimaging of Multiple Sclerosis at Low Specific Absorption Rates (SAR)},
  series = {PLOS ONE},
  volume    = {9},
  journal   = {PLOS ONE},
  number    = {3},
  issn      = {1932-6203},
  doi       = {10.1371/journal.pone.0091030},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-117179},
  pages     = {e91030},
  year      = {2014},
  abstract  = {Purpose: To compare a novel combined acquisition technique (CAT) of turbo-spin-echo (TSE) and echo-planar-imaging (EPI) with conventional TSE. CAT reduces the electromagnetic energy load transmitted for spin excitation. This radiofrequency (RF) burden is limited by the specific absorption rate (SAR) for patient safety. SAR limits restrict high-field MRI applications, in particular. Material and Methods: The study was approved by the local Medical Ethics Committee. Written informed consent was obtained from all participants. T2- and PD-weighted brain images of n = 40 Multiple Sclerosis (MS) patients were acquired by CAT and TSE at 3 Tesla. Lesions were recorded by two blinded, board-certificated neuroradiologists. Diagnostic equivalence of CAT and TSE to detect MS lesions was evaluated along with their SAR, sound pressure level (SPL) and sensations of acoustic noise, heating, vibration and peripheral nerve stimulation. Results: Every MS lesion revealed on TSE was detected by CAT according to both raters (Cohen's kappa of within-rater/across-CAT/TSE lesion detection kappa(CAT) = 1.00, at an inter-rater lesion detection agreement of kappa(LES) = 0.82). CAT reduced the SAR burden significantly compared to TSE (p<0.001). Mean SAR differences between TSE and CAT were 29.0 (+/- 5.7) \% for the T2-contrast and 32.7 (+/- 21.9) \% for the PD-contrast (expressed as percentages of the effective SAR limit of 3.2 W/kg for head examinations). Average SPL of CAT was no louder than during TSE. Sensations of CAT-vs. TSE-induced heating, noise and scanning vibrations did not differ. Conclusion: T2-/PD-CAT is diagnostically equivalent to TSE for MS lesion detection yet substantially reduces the RF exposure. Such SAR reduction facilitates high-field MRI applications at 3 Tesla or above and corresponding protocol standardizations but CAT can also be used to scan faster, at higher resolution or with more slices. According to our data, CAT is no more uncomfortable than TSE scanning.},
  language  = {en}
}