Enhanced bone assessment of the shoulder using zero-echo time MRI with deep-learning image reconstruction.

OBJECTIVES: To assess a deep learning-based reconstruction algorithm (DLRecon) in zero echo-time (ZTE) MRI of the shoulder at 1.5 Tesla for improved delineation of osseous findings. METHODS: In this retrospective study, 63 consecutive exams of 52 patients (28 female) undergoing shoulder MRI at 1.5 Tesla in clinical routine were included. Coronal 3D isotropic radial ZTE pulse sequences were acquired in the standard MR shoulder protocol. In addition to standard-of-care (SOC) image reconstruction, the same raw data was reconstructed with a vendor-supplied prototype DLRecon algorithm. Exams were classified into three subgroups: no pathological findings, degenerative changes, and posttraumatic changes, respectively. Two blinded readers performed bone assessment on a 4-point scale (0-poor, 3-perfect) by qualitatively grading image quality features and delineation of osseous pathologies including diagnostic confidence in the respective subgroups. Quantitatively, signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) of bone were measured. Qualitative variables were compared using the Wilcoxon signed-rank test for ordinal data and the McNemar test for dichotomous variables; quantitative measures were compared with Student's t-testing. RESULTS: DLRecon scored significantly higher than SOC in all visual metrics of image quality (all, p < 0.03), except in the artifact category (p = 0.37). DLRecon also received superior qualitative scores for delineation of osseous pathologies and diagnostic confidence (p ≤ 0.03). Quantitatively, DLRecon achieved superior CNR (95 CI [1.4-3.1]) and SNR (95 CI [15.3-21.5]) of bone than SOC (p < 0.001). CONCLUSION: DLRecon enhanced image quality in ZTE MRI and improved delineation of osseous pathologies, allowing for increased diagnostic confidence in bone assessment.

Deep Learning Convolutional Neural Network Reconstruction and Radial k-Space Acquisition MR Technique for Enhanced Detection of Retropatellar Cartilage Lesions of the Knee Joint.

OBJECTIVES: To assess diagnostic performance of standard radial k-space (PROPELLER) MRI sequences and compare with accelerated acquisitions combined with a deep learning-based convolutional neural network (DL-CNN) reconstruction for evaluation of the knee joint. METHODS: Thirty-five patients undergoing MR imaging of the knee at 1.5 T were prospectively included. Two readers evaluated image quality and diagnostic confidence of standard and DL-CNN accelerated PROPELLER MR sequences using a four-point Likert scale. Pathological findings of bone, cartilage, cruciate and collateral ligaments, menisci, and joint space were analyzed. Inter-reader agreement (IRA) for image quality and diagnostic confidence was assessed using intraclass coefficients (ICC). Cohen's Kappa method was used for evaluation of IRA and consensus between sequences in assessing different structures. In addition, image quality was quantitatively evaluated by signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) measurements. RESULTS: Mean acquisition time of standard vs. DL-CNN sequences was 10 min 3 s vs. 4 min 45 s. DL-CNN sequences showed significantly superior image quality and diagnostic confidence compared to standard MR sequences. There was moderate and good IRA for assessment of image quality in standard and DL-CNN sequences with ICC of 0.524 and 0.830, respectively. Pathological findings of the knee joint could be equally well detected in both sequences (κ-value of 0.8). Retropatellar cartilage could be significantly better assessed on DL-CNN sequences. SNR and CNR was significantly higher for DL-CNN sequences (both p < 0.05). CONCLUSIONS: In MR imaging of the knee, DL-CNN sequences showed significantly higher image quality and diagnostic confidence compared to standard PROPELLER sequences, while reducing acquisition time substantially. Both sequences perform comparably in the detection of knee-joint pathologies, while DL-CNN sequences are superior for evaluation of retropatellar cartilage lesions.

Diagnostic performance of deep learning-based reconstruction algorithm in 3D MR neurography.

OBJECTIVE: The study aims to evaluate the diagnostic performance of deep learning-based reconstruction method (DLRecon) in 3D MR neurography for assessment of the brachial and lumbosacral plexus. MATERIALS AND METHODS: Thirty-five exams (18 brachial and 17 lumbosacral plexus) of 34 patients undergoing routine clinical MR neurography at 1.5 T were retrospectively included (mean age: 49 ± 12 years, 15 female). Coronal 3D T2-weighted short tau inversion recovery fast spin echo with variable flip angle sequences covering plexial nerves on both sides were obtained as part of the standard protocol. In addition to standard-of-care (SOC) reconstruction, k-space was reconstructed with a 3D DLRecon algorithm. Two blinded readers evaluated images for image quality and diagnostic confidence in assessing nerves, muscles, and pathology using a 4-point scale. Additionally, signal-to-noise ratio (SNR) and contrast-to-noise ratios (CNR) between nerve, muscle, and fat were measured. For comparison of visual scoring result non-parametric paired sample Wilcoxon signed-rank testing and for quantitative analysis paired sample Student's t-testing was performed. RESULTS: DLRecon scored significantly higher than SOC in all categories of image quality (p < 0.05) and diagnostic confidence (p < 0.05), including conspicuity of nerve branches and pathology. With regard to artifacts there was no significant difference between the reconstruction methods. Quantitatively, DLRecon achieved significantly higher CNR and SNR than SOC (p < 0.05). CONCLUSION: DLRecon enhanced overall image quality, leading to improved conspicuity of nerve branches and pathology, and allowing for increased diagnostic confidence in evaluation of the brachial and lumbosacral plexus.

3D zero-echo time and 3D T1-weighted gradient-echo MRI sequences as an alternative to CT for the evaluation of the lumbar facet joints and lumbosacral transitional vertebrae.

BACKGROUND: Computed tomography (CT) is the reference standard for assessment of the bone. Magnetic resonance imaging (MRI) developments enable a CT-like visualization of the osseous structures. PURPOSE: To assess the diagnostic performance of 3D zero-echo time (3D-ZTE) and 3D T1-weighted gradient-echo (3D-T1GRE) MRI sequences for the evaluation of lumbar facet joints (LFJs) and the detection of lumbosacral transitional vertebrae (LSTV) using CT as the reference standard. MATERIAL AND METHODS: In total, 87 adult patients were included in this prospective study. Evaluation of degenerative changes of the facet joints at the L3/L4, L4/L5, and L5/S1 levels on both sides was performed by two readers using a 4-point Likert scale. LSTV were classified according to Castelvi et al. Image quality was quantitatively measured using the signal-to-noise (SNR) and contrast-to-noise (CNR) ratios. Intra-reader, inter-reader, and inter-modality reliability were calculated using Cohen's kappa statistic. RESULTS: Intra-reader agreement for 3D-ZTE, 3D-T1GRE, and CT was 0.607, 0.751, and 0.856 and inter-reader agreement was 0.535, 0.563, and 0.599, respectively. The inter-modality agreement between 3D-ZTE and CT was 0.631 and between 3D-T1GRE and CT 0.665. A total of LSTV were identified in both MR sequences with overall comparable accuracy compared to CT. Mean SNR for bone, muscle, and fat was highest for 3D-T1GRE and mean CNR was highest for CT. CONCLUSION: 3D-ZTE and 3D-T1GRE MRI sequences can assess the LFJs and LSTV and may serve as potential alternatives to CT.

Unravelling homologous recombination repair deficiency and therapeutic opportunities in soft tissue and bone sarcoma.

Defects in homologous recombination repair (HRR) in tumors correlate with poor prognosis and metastases development. Determining HRR deficiency (HRD) is of major clinical relevance as it is associated with therapeutic vulnerabilities and remains poorly investigated in sarcoma. Here, we show that specific sarcoma entities exhibit high levels of genomic instability signatures and molecular alterations in HRR genes, while harboring a complex pattern of chromosomal instability. Furthermore, sarcomas carrying HRDness traits exhibit a distinct SARC-HRD transcriptional signature that predicts PARP inhibitor sensitivity in patient-derived sarcoma cells. Concomitantly, HRDhigh sarcoma cells lack RAD51 nuclear foci formation upon DNA damage, further evidencing defects in HRR. We further identify the WEE1 kinase as a therapeutic vulnerability for sarcomas with HRDness and demonstrate the clinical benefit of combining DNA damaging agents and inhibitors of DNA repair pathways ex vivo and in the clinic. In summary, we provide a personalized oncological approach to treat sarcoma patients successfully.

The influence of partial union on the mechanical strength of scaphoid fractures: a finite element study.

Assessment of scaphoid fracture union on computed tomography scans is not currently standardized. We investigated the extent of scaphoid waist fracture union required to withstand physiological loads in a finite element model, based on a high-resolution CT scan of a cadaveric forearm. For simulations, the scaphoid waist was partially fused at the radial and ulnar sides. A physiological load of 100 N was transmitted to the scaphoid and the minimal amount of union to maintain biomechanical stability was recorded. The orientation of the fracture plane was varied to analyse the effect on biomechanical stability. The results indicate that the scaphoid is more prone to re-fracture when healing occurs on the ulnar side, where at least 60% union is required. Union occurring from the radial side can withstand loads with as little as 25% union. In fractures more parallel to the radial axis, the scaphoid seems less resistant on the radial side, as at least 50% union is required. A quantitative CT scan analysis with the proposed cut-off values and a consistently applied clinical examination will guide the clinician as to whether mid-waist scaphoid fractures can be considered as truly united.

Added value of ultra-short echo time and fast field echo using restricted echo-spacing MR imaging in the assessment of the osseous cervical spine.

PURPOSE: To evaluate the added value of ultra-short echo time (UTE) and fast field echo resembling a CT using restricted echo-spacing (FRACTURE) MR sequences in the assessment of the osseous cervical spine using CT as reference. MATERIALS AND METHODS: Twenty-seven subjects underwent postmortem CT and MRI within 48 h. Datasets were anonymized and analyzed retrospectively by two radiologists. Morphological cervical spine alterations were rated on CT, UTE and FRACTURE images. Afterward, neural foraminal stenosis was graded on standard MR and again after viewing additional UTE/FRACTURE sequences. To evaluate interreader and intermodality reliability, intra-class correlation coefficients (ICC) and for stenosis grading Wilcoxon-matched-pairs testing with multiple comparison correction were calculated. RESULTS: Moderate interreader reliability (ICC = 0.48-0.71) was observed concerning morphological findings on all modalities. Intermodality reliability was good between modalities regarding degenerative vertebral and joint alterations (ICC = 0.69-0.91). Compared to CT neural stenosis grades were more often considered as nonsignificant on all analyzed MR sequences. Neural stenosis grading scores differed also significantly between specific bone imaging sequences, UTE and FRACTURE, to standard MR sequences. However, no significant difference was observed between UTE and FRACTURE sequences. CONCLUSION: Compared to CT as reference, UTE or FRACTURE sequence added to standard MR sequences can deliver comparable information on osseous cervical spine status. Both led to changes in clinically significant stenosis gradings when added to standard MR, mainly reducing the severity of neural foramina stenosis.

The Effect of Paraspinal Fatty Muscle Infiltration and Cumulative Lumbar Spine Degeneration on the Outcome of Patients With Lumbar Spinal Canal Stenosis: Analysis of the Lumbar Stenosis Outcome Study (LSOS) Data.

STUDY DESIGN: Prospective. OBJECTIVE: To investigate the influence of paraspinal fatty muscle infiltration (FMI) and cumulative lumbar spine degeneration as assessed by magnetic resonance imaging on long-term clinical outcome measures in patients with lumbar spinal canal stenosis (LSCS) of the Lumbar Stenosis Outcome Study (LSOS) cohort. SUMMARY OF BACKGROUND DATA: Past studies have tried to establish correlations of morphologic imaging findings in LSCS with clinical endpoints. However, the impact of FMI and overall lumbar spinal degeneration load has not been examined yet. MATERIALS AND METHODS: Patients from the LSOS cohort with moderate to severe LSCS were included. Two radiologists assessed the degree of LSCS as well as cumulative degeneration of the lumbar spine. FMI was graded using the Goutallier scoring system. Spinal Stenosis Measure (SSM) was used to measure the severity level of symptoms and disability. European Quality of Life 5 Dimensions 3 Level Version (EQ-5D-3L) was used to measure health-related quality of life. RESULTS: The nonsurgically treated group consisted of 116 patients (age 74.8±8.5 yr), whereas the surgically treated group included 300 patients (age 72.3±8.2 yr). Paraspinal FMI was significantly different between the groups (54.3% vs. 32.0% for Goutallier grade ≥2; P <0.001). Total degeneration score was comparable in both groups (9.5±2.0 vs. 9.3±2.0; P =0.418). FMI was associated with lower SSM function and lower EQ-5D-3L (all P <0.05), but not with SSM symptoms. Total degeneration of the lumbar spine was associated neither with SSM symptoms, nor with SSM function, nor with EQ-5D-3L (all P >0.05). CONCLUSIONS: FMI is associated with higher disability and worse health-related quality of life of LSCS patients in the LSOS cohort. There was no significant association between total cumulative lumbar spine degeneration and the outcome of either surgically or nonsurgically treated patients. LEVEL OF EVIDENCE: 3.

Application of deep learning-based image reconstruction in MR imaging of the shoulder joint to improve image quality and reduce scan time.

OBJECTIVES: To compare the image quality and diagnostic performance of conventional motion-corrected periodically rotated overlapping parallel line with enhanced reconstruction (PROPELLER) MRI sequences with post-processed PROPELLER MRI sequences using deep learning-based (DL) reconstructions. METHODS: In this prospective study of 30 patients, conventional (19 min 18 s) and accelerated MRI sequences (7 min 16 s) using the PROPELLER technique were acquired. Accelerated sequences were post-processed using DL. The image quality and diagnostic confidence were qualitatively assessed by 2 readers using a 5-point Likert scale. Analysis of the pathological findings of cartilage, rotator cuff tendons and muscles, glenoid labrum and subacromial bursa was performed. Inter-reader agreement was calculated using Cohen's kappa statistic. Quantitative evaluation of image quality was measured using the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). RESULTS: Mean image quality and diagnostic confidence in evaluation of all shoulder structures were higher in DL sequences (p value = 0.01). Inter-reader agreement ranged between kappa values of 0.155 (assessment of the bursa) and 0.947 (assessment of the rotator cuff muscles). In 17 cases, thickening of the subacromial bursa of more than 2 mm was only visible in DL sequences. The pathologies of the other structures could be properly evaluated by conventional and DL sequences. Mean SNR (p value = 0.01) and CNR (p value = 0.02) were significantly higher for DL sequences. CONCLUSIONS: The accelerated PROPELLER sequences with DL post-processing showed superior image quality and higher diagnostic confidence compared to the conventional PROPELLER sequences. Subacromial bursa can be thoroughly assessed in DL sequences, while the other structures of the shoulder joint can be assessed in conventional and DL sequences with a good agreement between sequences. KEY POINTS: • MRI of the shoulder requires long scan times and can be hampered by motion artifacts. • Deep learning-based convolutional neural networks are used to reduce image noise and scan time while maintaining optimal image quality. The radial k-space acquisition technique (PROPELLER) can reduce the scan time and has potential to reduce motion artifacts. • DL sequences show a higher diagnostic confidence than conventional sequences and therefore are preferred for assessment of the subacromial bursa, while conventional and DL sequences show comparable performance in the evaluation of the shoulder joint.

Molecular and immunophenotypic characterization of SMARCB1 (INI1) - deficient intrathoracic Neoplasms.

The switch/sucrose-non-fermenting (SWI/SNF) complex is an ATP-dependent chromatin remodeling complex that plays important roles in DNA repair, transcription and cell differentiation. This complex consists of multiple subunits and is of particular interest in thoracic malignancies due to frequent subunit alteration of SMARCA4 (BRG1). Much less is known about SMARCB1 (INI1) deficient intrathoracic neoplasms, which are rare, often misclassified and understudied. In a retrospective analysis of 1479 intrathoracic malignant neoplasms using immunohistochemistry for INI1 (SMARCB1) on tissue micro arrays (TMA) and a search through our hospital sarcoma database, we identified in total nine intrathoracic, INI1 deficient cases (n = 9). We characterized these cases further by additional immunohistochemistry, broad targeted genomic analysis, methylation profiling and correlated them with clinical and radiological data. This showed that genomic SMARCB1 together with tumor suppressor alterations drive tumorigenesis in some of these cases, rather than epigenetic changes such as DNA methylation. A proper diagnostic classification, however, remains challenging. Intrathoracic tumors with loss or alteration of SMARCB1 (INI1) are highly aggressive and remain often underdiagnosed due to their rarity, which leads to false diagnostic interpretations. A better understanding of these tumors and proper diagnosis is important for better patient care as clinical trials and more targeted therapeutic options are emerging.

Detailed bone assessment of the sacroiliac joint in a prospective imaging study: comparison between computed tomography, zero echo time, and black bone magnetic resonance imaging.

OBJECTIVES: To compare the value of zero echo time (ZTE) and gradient echo "black bone" (BB) MRI sequences for bone assessment of the sacroiliac joint (SI) using computed tomography (CT) as the reference standard. MATERIALS AND METHODS: Between May 2019 and January 2021, 79 patients prospectively underwent clinically indicated 3-T MRI including ZTE and BB imaging. Additionally, all patients underwent a CT scan covering the SI joints within 12 months of the MRI examination. Two blinded readers performed bone assessment by grading each side of each SI joint qualitatively in terms of seven features (osteophytes, subchondral sclerosis, erosions, ankylosis, joint irregularity, joint widening, and gas in the SI joint) using a 4-point Likert scale (0 = no changes-3 = marked changes). Scores were compared between all three imaging modalities. RESULTS: Interreader agreement was largely good (k values: 0.5-0.83). Except for the feature "gas in SI joint" where ZTE exhibited significantly lower scores than CT (p < 0.001), ZTE and BB showed similar performance relative to CT for all other features (p > 0.52) with inter-modality agreement being substantial to almost perfect (Krippendorff's alpha coefficients: 0.724-0.983). When combining the data from all features except for gas in the SI joint and when binarizing grading scores, combined sensitivity/specificity was 76.7%/98.6% for ZTE and 80.8%/99.1% for BB, respectively, compared to CT. CONCLUSIONS: The performance of ZTE and BB sequences was comparable to CT for bone assessment of the SI joint. These sequences may potentially serve as an alternative to CT yet without involving exposure to ionizing radiation.

Impact of acceleration on bone depiction quality by ultrashort echo time magnetic resonance bone imaging sequences in medication-related osteonecrosis of the jaw.

Objectives: To assess the impact on bone depiction quality by decreasing number of radial acquisitions (RA) of a UTE MR bone imaging sequence in MRONJ. Material and methods: UTE MR bone imaging sequences using pointwise encoding time reduction with RA (PETRA) with 60'000, 30'000 and 10'000 RA were acquired in 16 patients with MRONJ and 16 healthy volunteers. Blinded readout sessions were performed by two radiologists. Qualitative analysis compared the detection of osteolytic lesions and productive bony changes in the PETRA sequences of the patients with MRONJ. Quantitative analysis assessed the differences in image artifacts, contrast-to-noise ratio (CNR) and image noise. Results: Acquisition times were reduced from 315 to 165 and 65 s (60'000, 30'000, 10'000 RA, respectively), resulting in a fewer number of severe motion artifacts. Bone delineation was increasingly blurred when reducing the number of RA but without any trade-off in terms of diagnostic performance. Interreader agreement for the detection of pathognomonic osteolysis was moderate (κ = 0.538) for 60'000 RA and decreased to fair (κ = 0.227 and κ = 0.390) when comparing 30'000 and 10'000 RA, respectively. Image quality between sequences was comparable regarding CNR, image noise and artifact dimensions without significant differences (all P > 0.05). Conclusions: UTE MR bone imaging sequences with a lower number of RA provide sufficient image quality for detecting osteolytic lesions and productive bony changes in MRONJ subjects at faster acquisition times compared to the respective standard UTE MR bone imaging sequence.

Added value of combined acromiohumeral distance and critical shoulder angle measurements on conventional radiographs for the prediction of rotator cuff pathology.

Purpose: To investigate the role of acromiohumeral distance (AHD) and critical shoulder angle (CSA) measurements from conventional radiographs (CR) in isolation and combined (prognostic index PIAHD-CSA) as predictors of full thickness rotator cuff tendon tears (RCT) and critical fatty degeneration (CFD; i.e. as much fat as muscle). Method: In this retrospective study AHD and CSA were measured in 127 CR. MR arthrograms served as reference standard and were screened for RCT and CFD. Statistical analysis for inter-reader agreement, Spearman's rank correlation, linear stepwise regression and logistic regression for AHD and CSA with ROC analyses including PIAHD-CSA were performed. Results: In 90 subjects (17 females, mean age 36.1 ± 14.1) no RCT were found on MR imaging and served as control group. In 37 patients (13 females, mean age 58.7 ± 13.2) ≥ one RCT was found. Inter-reader agreements rated between к = 0.42-0.82 for categorical and 0.91-0.96 for continuous variables. No significant correlation of AHD and CSA with either age or sex was seen (p = 0.28 and p = 0.74, respectively). Case group had significantly smaller mean AHD (8.7 ± 3.2 vs. 10.8 ± 2.2 mm; p < 0.001) and larger mean CSA (36.5 ± 4.5° vs. 33.1 ± 4.0°; p < 0.001). PIAHD-CSA increased diagnostic performance for prediction of RCT and CFD (AUC = 0.78 and 0.71), compared to isolated AHD (0.74 and 0.71) and CSA (0.71 and 0.66). Conclusions: AHD and CSA do not depend on age or sex but differ significantly between healthy and pathologic rotator cuffs. A decreased AHD is most influenced by infraspinatus muscle atrophy and fatty degeneration. Combined PIAHD-CSA increases diagnostic performance for predicting RCT and CFD.

Comparison of muscle fat fraction measurements in the lower spine musculature with non-contrast-enhanced CT and different MR imaging sequences.

PURPOSE: To assess whether two-point Dixon (TPD) MRI, true fast imaging with steady-state free precession (TRUFI) MRI and non-contrast-enhanced CT (NECT) can accurately measure muscle fat fraction (FF) in the autochthonous back muscles (AM) and the psoas muscle (PM) compared to multi-point Dixon (MPD) MRI. METHOD: 29 oncological patients who received MRI including MPD, TPD and NECT imaging in a period of three months were analyzed retrospectively. A sub-cohort of 16 patients additionally underwent TRUFI MRI and were included in a sub-analysis. Region of interest (ROI) measurements for each muscle compartment of the AM and PM were conducted by two examiners. Additionally, the Goutallier classification was used to quantify the amount of fatty infiltration of each muscle. Intermodality correlations were assessed with the Pearson correlation coefficient (r), and interreader and intrareader agreements with the intraclass correlation coefficient (ICC). RESULTS: Good intermodality correlations were found for NECT (r = 0.969), TPD (r = 0.942) and TRUFI (r = 0.904, all P < 0.001) when assessing FF in the AM and slightly lower in the PM. Interreader agreement showed good correlations and low median deviations (1.1 - 4.1 %, depending on the modality). The Goutallier classification of the AM showed good separation between grades with substantial interreader agreement (κ = 0.627, P < 0.001). CONCLUSIONS: ROI measurements of the AM in NECT, TPD and TRUFI highly correlate with muscle FF measurements in MPD MRI and may be used to assess sarcopenia in oncological patients.

Correlation of dynamic contrast-enhanced bone perfusion with morphologic ultra-short echo time MR imaging in medication-related osteonecrosis of the jaw.

OBJECTIVES: To investigate whether dynamic contrast-enhanced (DCE)-MR bone perfusion could serve as surrogate for morphologic ultra-short echo time (UTE) bone images and to correlate perfusion with morphologic hallmarks in histologically proven foci of medication-related osteonecrosis of the jaw (MRONJ). METHODS: Retrospective study including 20 patients with established diagnosis of MRONJ. Qualitative consensus assessment of predefined jaw regions by two radiologists was used as reference standard using Likert scale (0-3) for standard imaging hallmarks in MRONJ (osteolysis, sclerosis, periosteal thickening). DCE-MRI measurements performed in corresponding regions of the mandible were then correlated with qualitative scores. Regions were grouped into "non-affected" and "pathologic" based on binarized Likert scores of different imaging hallmarks (0-1 vs 2-3). DCE-MRI measurements among hallmarks were compared using Mann-Whitney-U-testing. ROC (receiver-operating-characteristic) analysis was performed for each of the perfusion parameters to assess diagnostic performance for identification of MRONJ using morphologic ratings as reference standard. RESULTS: Median perfusion measurements of "pathologic" regions in wash-in, peak enhancement intensity and integrated area under the curve are significantly higher than those of "non-affected" regions, irrespective of reference imaging hallmark (p < 0.05). No significant perfusion differences were found between "pathologic" regions with and without osteolysis (p = 0.180). ROC analysis showed fair diagnostic performance of DCE-MRI parameters for identification of MRONJ (AUC 0.626-0.727). CONCLUSIONS: DCE bone perfusion parameters are significantly increased in MRONJ compared to non-affected regions, irrespective of osteolysis. Due to certain overlap DCE-MRI bone perfusion cannot serve as full surrogate for UTE bone imaging but may enhance reader confidence.

Comparison of metal artifact reduction techniques in magnetic resonance imaging of carbon-reinforced PEEK and titanium spinal implants.

BACKGROUND: Carbon-reinforced PEEK (C-FRP) implants are non-magnetic and have increasingly been used for the fixation of spinal instabilities. PURPOSE: To compare the effect of different metal artifact reduction (MAR) techniques in magnetic resonance imaging (MRI) on titanium and C-FRP spinal implants. MATERIAL AND METHODS: Rod-pedicle screw constructs were mounted on ovine cadaver spine specimens and instrumented with either eight titanium pedicle screws or pedicle screws made of C-FRP and marked with an ultrathin titanium shell. MR scans were performed of each configuration on a 3-T scanner. MR sequences included transaxial conventional T1-weighted turbo spin echo (TSE) sequences, T2-weighted TSE, and short-tau inversion recovery (STIR) sequences and two different MAR-techniques: high-bandwidth (HB) and view-angle-tilting (VAT) with slice encoding for metal artifact correction (SEMAC). Metal artifact degree was assessed by qualitative and quantitative measures. RESULTS: There was a much stronger effect on artifact reduction with using C-FRP implants compared to using specific MRI MAR-techniques (screw shank: P < 0.001; screw tulip: P < 0.001; rod: P < 0.001). VAT-SEMAC sequences were able to reduce screw-related signal loss artifacts in constructs with titanium screws to a certain degree. Constructs with C-FRP screws showed less artifact-related implant diameter amplification when compared to constructs with titanium screws (P < 0.001). CONCLUSION: Constructs with C-FRP screws are associated with significantly less artifacts compared to constructs with titanium screws including dedicated MAR techniques. Artifact-reducing sequences are able to reduce implant-related artifacts. This effect is stronger in constructs with titanium screws than in constructs with C-FRP screws.

AI MSK clinical applications: spine imaging.

Recent investigations have focused on the clinical application of artificial intelligence (AI) for tasks specifically addressing the musculoskeletal imaging routine. Several AI applications have been dedicated to optimizing the radiology value chain in spine imaging, independent from modality or specific application. This review aims to summarize the status quo and future perspective regarding utilization of AI for spine imaging. First, the basics of AI concepts are clarified. Second, the different tasks and use cases for AI applications in spine imaging are discussed and illustrated by examples. Finally, the authors of this review present their personal perception of AI in daily imaging and discuss future chances and challenges that come along with AI-based solutions.

Whole-body Composition Profiling Using a Deep Learning Algorithm: Influence of Different Acquisition Parameters on Algorithm Performance and Robustness.

OBJECTIVES: To develop, test, and validate a body composition profiling algorithm for automated segmentation of body compartments in whole-body magnetic resonance imaging (wbMRI) and to investigate the influence of different acquisition parameters on performance and robustness. MATERIALS AND METHODS: A segmentation algorithm for subcutaneous and visceral adipose tissue (SCAT and VAT) and total muscle mass (TMM) was designed using a deep learning U-net architecture convolutional neuronal network. Twenty clinical wbMRI scans were manually segmented and used as training, validation, and test datasets. Segmentation performance was then tested on different data, including different magnetic resonance imaging protocols and scanners with and without use of contrast media. Test-retest reliability on 2 consecutive scans of 16 healthy volunteers each as well as impact of parameters slice thickness, matrix resolution, and different coil settings were investigated. Sorensen-Dice coefficient (DSC) was used to measure the algorithms' performance with manual segmentations as reference standards. Test-retest reliability and parameter effects were investigated comparing respective compartment volumes. Abdominal volumes were compared with published normative values. RESULTS: Algorithm performance measured by DSC was 0.93 (SCAT) to 0.77 (VAT) using the test dataset. Dependent from the respective compartment, similar or slightly reduced performance was seen for other scanners and scan protocols (DSC ranging from 0.69-0.72 for VAT to 0.83-0.91 for SCAT). No significant differences in body composition profiling was seen on repetitive volunteer scans (P = 0.88-1) or after variation of protocol parameters (P = 0.07-1). CONCLUSIONS: Body composition profiling from wbMRI by using a deep learning-based convolutional neuronal network algorithm for automated segmentation of body compartments is generally possible. First results indicate that robust and reproducible segmentations equally accurate to a manual expert may be expected also for a range of different acquisition parameters.

Evaluation of ultrashort echo-time (UTE) and fast-field-echo (FRACTURE) sequences for skull bone visualization and fracture detection - A postmortem study.

BACKGROUND AND PURPOSE: CT is considered the modality of choice in the assessment of the skull due to the fast and accurate depiction of bone structures. Nevertheless, MRI has evolved into a possible alternative due to optimal soft tissue contrast and recent advances with the ability to visualize tissues with shortest T2 times, such as osseous structures. In this study we compare skull bone visualization and fracture detection across two MRI sequences to CT as reference standard. MATERIAL AND METHODS: Twenty subjects underwent CT and MRI with less than 72 h between examination. The MRI protocol included a 2D ultrashort echo time (UTE) and a 3D multi-echo in-phase fast-field-echo (FRACTURE) sequence. Independent raters evaluated qualitative characteristics and fracture detectability in different skull subregions (skull vault, skull base and viscerocranium). Interrater and intermodality agreement was evaluated by calculating intraclass coefficients (ICC). RESULTS: FRACTURE ICC indicated a good agreement in all subregions (ICC = 0.83 - 0.88), whereas UTE had excellent results calculated in the skull vault and viscerocranium (ICC = 0.91 - 0.94). At the skull vault, both MRI sequences received an overall good rating (UTE: 2.63 ± 0.42 FRACTURE. 2.81 ± 0.32). Fracture detection using MRI sequences for the skull vault, was highest compared to other subregions. CONCLUSIONS: Both MRI sequences may provide an alternative e.g. for surgical planning or follow up exams of the osseous neurocranium; although, at the skull base and viscerocranium bone visualization with MRI bone imaging sequences perform inferior to CT standard imaging.

Comparison of AI-powered 3D automated ultrasound tomography with standard handheld ultrasound for the visualization of the hands-clinical proof of concept.

OBJECTIVE: To assess the ability of a newly developed AI-powered ultrasound 3D hand scanner to visualize joint structures in healthy hands and detect degenerative changes in cadaveric hands. MATERIALS AND METHODS: Twelve individuals (6 males, 6 females, age 43.5 ± 17.8 years) underwent four scans with the 3D ultrasound tomograph (right and left hand, dorsal and palmar, respectively) as well as four sets of handheld ultrasound of predefined anatomic regions. The 3D ultrasound tomographic images and the standard handheld ultrasound images were assessed by two radiologists with regard to visibility of bone contour, joint capsule and space, and tendons. In addition, three cadaveric hands were scanned with the 3D ultrasound tomograph and CT. RESULTS: Mean scan time for both hands was significantly faster with handheld ultrasound (10 min 30 s ± 95 s) compared to 3D ultrasound tomography (32 min 9 s ± 6 s; p < 0.001). Interreader and intermodality agreement was moderate (0.4 < κ ≤ 0.6) to substantial (0.6 < κ ≤ 0.8). Overall visibility of joint structures was comparable between the modalities at the level of the wrist (p = 0.408), and significantly better with handheld ultrasound at the level of the finger joints and the thumb (both p < 0.001). The 3D ultrasound tomograph was able to detect osteophytes in cadaveric hands which were confirmed by CT. CONCLUSION: The AI-powered 3D ultrasound tomograph was able to visualize joint structures in healthy hands and singular osteophytes in cadaveric hands. Further technical improvements are necessary to shorten scan times and improve automated scanning of the finger joints and the thumb.