Background and Purpose: This study explores the use of deep learning (DL) techniques in MRI of the orbit to enhance imaging. Standard protocols, although detailed, have lengthy acquisition times. We investigate DL‐based methods for T2‐weighted and T1‐weighted, fat‐saturated, contrast‐enhanced turbo spin echo (TSE) sequences, aiming to improve image quality, reduce acquisition time, minimize artifacts, and enhance diagnostic confidence in orbital imaging.Methods: In a 3‐Tesla MRI study of 50 patients evaluating orbital diseases from March to July 2023, conventional (TSE_S) and DL TSE sequences (TSE_DL) were used. Two neuroradiologists independently assessed the image datasets for image quality, diagnostic confidence, noise levels, artifacts, and image sharpness using a randomized and blinded 4‐point Likert scale.Results: TSE_DL significantly reduced image noise and artifacts, enhanced image sharpness, and decreased scan time, outperforming TSE_S (p < .05). TSE_DL showed superior overall image quality and diagnostic confidence, with relevant findings effectively detected in both DL‐based and conventional images. In 94% of cases, readers preferred accelerated imaging.Conclusion: The study proved that using DL for MRI image reconstruction in orbital scans significantly cut acquisition time by 69%. This approach also enhanced image quality, reduced image noise, sharpened images, and boosted diagnostic confidence.

Before revascularization, moyamoya patients require hemodynamic evaluation. In this study, we evaluated the scoring system Prior Infarcts, Reactivity and Angiography in Moyamoya Disease (PIRAMID). We also devised a new scoring system, MRI-Based Assessment of Risk for Stroke in Moyamoya Angiopathy (MARS-MMA), and compared the scoring systems with respect to the capability to predict impaired [¹⁵O]water PET cerebral perfusion reserve capacity (CPR). We evaluated 69 MRI, 69 DSA and 38 [¹⁵O]water PET data sets. The PIRAMID system was validated by ROC curve analysis with neurological symptomatology as a dependent variable. The components of the MARS-MMA system and their weightings were determined by binary logistic regression analysis. The comparison of PIRAMID and MARS-MMA was performed by ROC curve analysis. The PIRAMID score correlated well with the symptomatology (AUC = 0.784). The MARS-MMA system, including impaired breath-hold-fMRI, the presence of the Ivy sign and arterial wall contrast enhancement, correlated slightly better with CPR impairment than the PIRAMID system (AUC = 0.859 vs. 0.827, Akaike information criterion 140 vs. 146). For simplified clinical use, we determined three MARS-MMA grades without loss of diagnostic performance (AUC = 0.855). The entirely MRI-based MARS-MMA scoring system might be a promising tool to predict the risk of stroke.

A fully diagnostic MRI glioma protocol is key to monitoring therapy assessment but is time-consuming and especially challenging in critically ill and uncooperative patients. Artificial intelligence demonstrated promise in reducing scan time and improving image quality simultaneously. The purpose of this study was to investigate the diagnostic performance, the impact on acquisition acceleration, and the image quality of a deep learning optimized glioma protocol of the brain. Thirty-three patients with histologically confirmed glioblastoma underwent standardized brain tumor imaging according to the glioma consensus recommendations on a 3-Tesla MRI scanner. Conventional and deep learning-reconstructed (DLR) fluid-attenuated inversion recovery, and T2- and T1-weighted contrast-enhanced Turbo spin echo images with an improved in-plane resolution, i.e., super-resolution, were acquired. Two experienced neuroradiologists independently evaluated the image datasets for subjective image quality, diagnostic confidence, tumor conspicuity, noise levels, artifacts, and sharpness. In addition, the tumor volume was measured in the image datasets according to Response Assessment in Neuro-Oncology (RANO) 2.0, as well as compared between both imaging techniques, and various clinical–pathological parameters were determined. The average time saving of DLR sequences was 30% per MRI sequence. Simultaneously, DLR sequences showed superior overall image quality (all p < 0.001), improved tumor conspicuity and image sharpness (all p < 0.001, respectively), and less image noise (all p < 0.001), while maintaining diagnostic confidence (all p > 0.05), compared to conventional images. Regarding RANO 2.0, the volume of non-enhancing non-target lesions (p = 0.963), enhancing target lesions (p = 0.993), and enhancing non-target lesions (p = 0.951) did not differ between reconstruction types. The feasibility of the deep learning-optimized glioma protocol was demonstrated with a 30% reduction in acquisition time on average and an increased in-plane resolution. The evaluated DLR sequences improved subjective image quality and maintained diagnostic accuracy in tumor detection and tumor classification according to RANO 2.0.

Background/Objectives: Postoperative MRI is crucial for detecting residual tumor, identifying complications, and planning subsequent therapy. This study evaluates accelerated deep learning reconstruction (DLR) versus standard clinical protocols for early postoperative MRI following tumor resection. Methods: This study uses a multidisciplinary approach involving a neuroradiologist, neurosurgeon, neuro-oncologist, and radiotherapist to evaluate qualitative aspects using a 5-point Likert scale, the preferred reconstruction variant and potential residual tumor of DLR and conventional reconstruction (CR) of FLAIR, T1-weighted non-contrast and contrast-enhanced (T1), and coronal T2-weighted (T2) sequences for 1.5 and 3 T MRI. Quantitative analysis included the image quality metrics Structural Similarity Index (SSIM), Multi-Scale SSIM (MS-SSIM), Feature Similarity Index (FSIM), Noise Quality Metric (NQM), signal-to-noise ratio (SNR), and Peak SNR (PSNR) with CR as a reference. Results: All raters strongly preferred DLR over CR. This was most pronounced for FLAIR images at 1.5 and 3 T (91% at 1.5 T and 97% at 3 T) and least pronounced for T1 at 1.5 T (79% for non-contrast-enhanced and 84% for contrast-enhanced sequences) and for T2 at 3 T (69%). DLR demonstrated superior qualitative image quality for all sequences and field strengths (p < 0.001), except for T2 at 3 T, which was observed across all raters (p = 0.670). Diagnostic confidence was similar at 3 T with better but non-significant differences for T2 (p = 0.134) and at 1.5 T with better but non-significant differences for non-contrast-enhanced T1 (p = 0.083) and only marginally significant results for FLAIR (p = 0.033). Both the SSIM and MS-SSIM indicated near-perfect similarity between CR and DLR. FSIM performs worse in terms of consistency between CR and DLR. The image quality metrics NQM, SNR, and PSNR showed better results for DLR. Visual assessment of residual tumor was similar at 3 T but differed at 1.5 T, with more residual tumor detected with DLR, especially by the neurosurgeon (n = 4). Conclusions: An accelerated DLR protocol demonstrates clinical feasibility, enabling high-quality reconstructions in challenging postoperative MRIs. DLR sequences received strong multidisciplinary preference, underscoring their potential to improve neuro-oncologic decision making and suitability for clinical implementation.

Background/Objectives: Iterative reconstruction (IR) techniques were developed to address the shortcomings of filtered back projection (FBP), yet research comparing different types of IR is still missing. This work investigates how reducing radiation dose influences both image quality and noise profiles when using two iterative reconstruction techniques—Sinogram-Affirmed Iterative Reconstruction (SAFIRE) and Advanced Modeled Iterative Reconstruction (ADMIRE)—in comparison to filtered back projection (FBP) in non-enhanced head CT (NECT). Methods: In this retrospective single-center study, 21 consecutive patients underwent standard NECT on a 128-slice CT scanner. Raw data simulated dose reductions to 90% and 70% of the original dose via ReconCT software. For each dose level, images were reconstructed with FBP, SAFIRE 3, and ADMIRE 3. Image noise power spectra quantified objective image noise. Two blinded neuroradiologists scored overall image quality, image noise, image contrast, detail, and artifacts on a 10-point Likert scale in a consensus reading. Quantitative Hounsfield unit (HU) measurements were obtained in white and gray matter regions. Statistical analyses included the Wilcoxon signed-rank test, mixed-effects modeling, ANOVA, and post hoc pairwise comparisons with Bonferroni correction. Results: Both iterative reconstructions significantly reduced image noise compared to FBP across all dose levels (p < 0.001). ADMIRE exhibited superior image noise suppression at low (1.31 1/mm) spatial frequencies, whereas SAFIRE performed better in the mid-frequency range (0.51–1.31 1/mm). Subjective scores for overall quality, image noise, image contrast, and detail were higher for ADMIRE and SAFIRE versus FBP at the original dose and simulated doses of 90% and 70% (all p < 0.001). ADMIRE outperformed SAFIRE in artifact reduction (p < 0.001), while SAFIRE achieved slightly higher image contrast scores (p < 0.001). Objective HU values remained stable across reconstruction methods, although SAFIRE yielded marginally higher gray and white matter (WM) attenuations (p < 0.01). Conclusions: Both IR techniques—ADMIRE and SAFIRE—achieved substantial noise reduction and improved image quality relative to FBP in non-enhanced head CT at standard and reduced dose levels on the specific CT system and reconstruction strength tested. ADMIRE showed enhanced suppression of low- and high-frequency image noise and fewer artifacts, while SAFIRE preserved image contrast and reduced mid-frequency noise. These findings support the potential of iterative reconstruction to optimize radiation dose in NECT protocols in line with the ALARA principle, although broader validation in multi-vendor, multi-center settings is warranted.