Search

This review examines the latest advancements in compositional and quantitative cartilage MRI techniques, addressing both their potential and challenges. The integration of these advancements promises to improve disease detection, treatment monitoring, and overall patient care. We want to highlight the pivotal task of translating these techniques into widespread clinical use, the transition of cartilage MRI from technical validation to clinical application, emphasizing its critical role in identifying early signs of degenerative and inflammatory joint diseases. Recognizing these changes early may enable informed treatment decisions, thereby facilitating personalized medicine approaches. The evolving landscape of cartilage MRI underscores its increasing importance in clinical practice, offering valuable insights for patient management and therapeutic interventions. This review aims to discuss the old evidence and new insights about the evaluation of articular cartilage through MRI, with an update on the most recent literature published on novel quantitative sequences.

CT protocols that diagnose COVID-19 vary in regard to the associated radiation exposure and the desired image quality (IQ). This study aims to evaluate CT protocols of hospitals participating in the RACOON (Radiological Cooperative Network) project, consolidating CT protocols to provide recommendations and strategies for future pandemics. In this retrospective study, CT acquisitions of COVID-19 patients scanned between March 2020 and October 2020 (RACOON phase 1) were included, and all non-contrast protocols were evaluated. For this purpose, CT protocol parameters, IQ ratings, radiation exposure (CTDIvol), and central patient diameters were sampled. Eventually, the data from 14 sites and 534 CT acquisitions were analyzed. IQ was rated good for 81% of the evaluated examinations. Motion, beam-hardening artefacts, or image noise were reasons for a suboptimal IQ. The tube potential ranged between 80 and 140 kVp, with the majority between 100 and 120 kVp. CTDIvol was 3.7 ± 3.4 mGy. Most healthcare facilities included did not have a specific non-contrast CT protocol. Furthermore, CT protocols for chest imaging varied in their settings and radiation exposure. In future, it will be necessary to make recommendations regarding the required IQ and protocol parameters for the majority of CT scanners to enable comparable IQ as well as radiation exposure for different sites but identical diagnostic questions.

Based on in silico, in vitro, in situ, and in vivo evaluations, this study aims to establish and optimize the chemical exchange saturation transfer (CEST) imaging of lactate (Lactate-CEST—LATEST). To this end, we optimized LATEST sequences using Bloch–McConnell simulations for optimal detection of lactate with a clinical 3 T MRI scanner. The optimized sequences were used to image variable lactate concentrations in vitro (using phantom measurements), in situ (using nine human cadaveric lower leg specimens), and in vivo (using four healthy volunteers after exertional exercise) that were then statistically analyzed using the non-parametric Friedman test and Kendall Tau-b rank correlation. Within the simulated Bloch–McConnell equations framework, the magnetization transfer ratio asymmetry (MTRasym) value was quantified as 0.4% in the lactate-specific range of 0.5–1 ppm, both in vitro and in situ, and served as the imaging surrogate of the lactate level. In situ, significant differences (p < 0.001) and strong correlations (τ = 0.67) were observed between the MTRasym values and standardized intra-muscular lactate concentrations. In vivo, a temporary increase in the MTRasym values was detected after exertional exercise. In this bench-to-bedside comprehensive feasibility study, different lactate concentrations were detected using an optimized LATEST imaging protocol in vitro, in situ, and in vivo at 3 T, which prospectively paves the way towards non-invasive quantification and monitoring of lactate levels across a broad spectrum of diseases.

Chemical Exchange Saturation Transfer (CEST) magnetic resonance imaging (MRI) provides a novel method for analyzing biomolecule concentrations in tissues without exogenous contrast agents. Despite its potential, achieving a high signal-to-noise ratio (SNR) is imperative for detecting small CEST effects. Traditional metrics such as Magnetization Transfer Ratio Asymmetry (MTRasym) and Lorentzian analyses are vulnerable to image noise, hampering their precision in quantitative concentration estimations. Recent noise-reduction algorithms like principal component analysis (PCA), nonlocal mean filtering (NLM), and block matching combined with 3D filtering (BM3D) have shown promise, as there is a burgeoning interest in the utilization of neural networks (NNs), particularly autoencoders, for imaging denoising. This study uses the Bloch–McConnell equations, which allow for the synthetic generation of CEST images and explores NNs efficacy in denoising these images. Using synthetically generated phantoms, autoencoders were created, and their performance was compared with traditional denoising methods using various datasets. The results underscored the superior performance of NNs, notably the ResUNet architectures, in noise identification and abatement compared to analytical approaches across a wide noise gamut. This superiority was particularly pronounced at elevated noise intensities in the in vitro data. Notably, the neural architectures significantly improved the PSNR values, achieving up to 35.0, while some traditional methods struggled, especially in low-noise reduction scenarios. However, the application to the in vivo data presented challenges due to varying noise profiles. This study accentuates the potential of NNs as robust denoising tools, but their translation to clinical settings warrants further investigation.

Sodium magnetic resonance imaging (MRI) can be used to evaluate the change in the proteoglycan content in Achilles tendons (ATs) of patients with different AT pathologies by measuring the 23Na signal-to-noise ratio (SNR). As 23Na SNR alone is difficult to compare between different studies, because of the high influence of hardware configurations and sequence settings on the SNR, we further set out to measure the apparent tissue sodium content (aTSC) in the AT as a better comparable parameter. Ten healthy controls and one patient with tendinopathy in the AT were examined using a clinical 3 Tesla (T) MRI scanner in conjunction with a dual tuned 1H/23Na surface coil to measure 23Na SNR and aTSC in their ATs. 23Na T1 and T2* of the AT were also measured for three controls to correct for different relaxation behavior. The results were as follows: 23Na SNR = 11.7 ± 2.2, aTSC = 82.2 ± 13.9 mM, 23Na T1 = 20.4 ± 2.4 ms, 23Na T2s* = 1.4 ± 0.4 ms, and 23Na T2l* = 13.9 ± 0.8 ms for the whole AT of healthy controls with significant regional differences. These are the first reported aTSCs and 23Na relaxation times for the AT using sodium MRI and may serve for future comparability in different studies regarding examinations of diseased ATs with sodium MRI.

Based on in silico, in situ, and in vivo studies, this study aims to develop a new method for the quantitative chemical exchange saturation transfer (qCEST) technique considering multi-pool systems. To this end, we extended the state-of-the-art apparent exchange-dependent relaxation (AREX) method with a Lorentzian correction (LAREX). We then validated this new method with in situ and in vivo experiments on human intervertebral discs (IVDs) using the Kendall-Tau correlation coefficient. In the in silico experiments, we observed significant deviations of the AREX method as a function of the underlying exchange rate (kba) and fractional concentration (fb) compared to the ground truth due to the influence of other exchange pools. In comparison to AREX, the LAREX-based Ω-plot approach yielded a substantial improvement. In the subsequent in situ and in vivo experiments on human IVDs, no correlation to the histological reference standard or Pfirrmann classification could be found for the fb (in situ: τ = −0.17 p = 0.51; in vivo: τ = 0.13 p = 0.30) and kba (in situ: τ = 0.042 p = 0.87; in vivo: τ = −0.26 p = 0.04) of Glycosaminoglycan (GAG) with AREX. In contrast, the influence of interfering pools could be corrected by LAREX, and a moderate to strong correlation was observed for the fractional concentration of GAG for both in situ (τ = −0.71 p = 0.005) and in vivo (τ = −0.49 p < 0.001) experiments. The study presented here is the first to introduce a new qCEST method that enables qCEST imaging in systems with multiple proton pools.

Sodium MRI has the potential to depict cartilage health accurately, but synovial fluid can influence the estimation of sodium parameters of cartilage. Therefore, this study aimed to reduce the impact of synovial fluid to render the quantitative compositional analyses of cartilage tissue technically more robust. Two dedicated protocols were applied for determining sodium T1 and T2* relaxation times. For each protocol, data were acquired from 10 healthy volunteers and one patient with patellar cartilage damage. Data recorded with multiple repetition times for T1 measurement and multi-echo data acquired with an additional inversion recovery pulse for T2* measurement were analysed using biexponential models to differentiate longitudinal relaxation components of cartilage (T1,car) and synovial fluid (T1,syn), and short (T2s*) from long (T2l*) transversal relaxation components. Sodium relaxation times and concentration estimates in patellar cartilage were successfully determined: T1,car = 14.5 ± 0.7 ms; T1,syn = 37.9 ± 2.9 ms; c(T1-protocol) = 200 ± 48 mmol/L; T2s* = 0.4 ± 0.1 ms; T2l* = 12.6 ± 0.7 ms; c(T2*-protocol) = 215 ± 44 mmol/L for healthy volunteers. In conclusion, a robust determination of sodium relaxation times is possible at a clinical field strength of 3T to quantify sodium concentrations, which might be a valuable tool to determine cartilage health.

Using glycosaminoglycan Chemical Exchange Saturation Transfer (gagCEST) magnetic resonance imaging (MRI), this study comparatively evaluated the GAG contents of lumbar intervertebral disks (IVDs) of patients with non-specific low back pain (nsLBP), radiculopathy, and asymptomatic volunteers to elucidate the association of clinical manifestation and compositional correlate. A total of 18 patients (mean age 57.5 ± 22.5 years) with radiculopathy, 16 age-matched patients with chronic nsLBP and 20 age-matched volunteers underwent standard morphologic and compositional gagCEST MRI on a 3T scanner. In all cohorts, GAG contents of lumbar IVDs were determined using gagCEST MRI. An assessment of morphologic IVD degeneration based on the Pfirrmann classification and T2-weighted sequences served as a reference. A linear mixed model adjusted for multiple confounders was used for statistical evaluation. IVDs of patients with nsLBP showed lower gagCEST values than those of volunteers (nsLBP: 1.3% [99% confidence intervals (CI): 1.0; 1.6] vs. volunteers: 1.9% [99% CI: 1.6; 2.2]). Yet, IVDs of patients with radiculopathy (1.8% [99% CI: 1.4; 2.1]) were not different from patients with nsLBP or volunteers. In patients with radiculopathy, IVDs directly adjacent to IVD extrusions demonstrated lower gagCEST values than distant IVDs (adjacent: 0.9% [99% CI: 0.3; 1.5], distant: 2.1% [99% CI: 1.7; 2.5]). Advanced GAG depletion in nsLBP and directly adjacent to IVD extrusions in radiculopathy indicates close interrelatedness of clinical pathology and compositional degeneration.

Abstract Background Written medical examinations consist of multiple-choice questions and/or free-text answers. The latter require manual evaluation and rating, which is time-consuming and potentially error-prone. We tested whether natural language processing (NLP) can be used to automatically analyze free-text answers to support the review process. Methods The European Board of Radiology of the European Society of Radiology provided representative datasets comprising sample questions, answer keys, participant answers, and reviewer markings from European Diploma in Radiology examinations. Three free-text questions with the highest number of corresponding answers were selected: Questions 1 and 2 were “unstructured” and required a typical free-text answer whereas question 3 was “structured” and offered a selection of predefined wordings/phrases for participants to use in their free-text answer. The NLP engine was designed using word lists, rule-based synonyms, and decision tree learning based on the answer keys and its performance tested against the gold standard of reviewer markings. Results After implementing the NLP approach in Python, F1 scores were calculated as a measure of NLP performance: 0.26 (unstructured question 1, n = 96), 0.33 (unstructured question 2, n = 327), and 0.5 (more structured question, n = 111). The respective precision/recall values were 0.26/0.27, 0.4/0.32, and 0.62/0.55. Conclusion This study showed the successful design of an NLP-based approach for automatic evaluation of free-text answers in the EDiR examination. Thus, as a future field of application, NLP could work as a decision-support system for reviewers and support the design of examinations being adjusted to the requirements of an automated, NLP-based review process. Clinical relevance statement Natural language processing can be successfully used to automatically evaluate free-text answers, performing better with more structured question-answer formats. Furthermore, this study provides a baseline for further work applying, e.g., more elaborated NLP approaches/large language models. Key points • Free-text answers require manual evaluation, which is time-consuming and potentially error-prone. • We developed a simple NLP-based approach — requiring only minimal effort/modeling — to automatically analyze and mark free-text answers. • Our NLP engine has the potential to support the manual evaluation process. • NLP performance is better on a more structured question-answer format. Graphical Abstract

Key points 1. We developed an NLP-based reporting tool that converts free text into structured reports. 2. The reporting tool sends visual feedback if relevant findings are missed. 3. Testing with text samples and original free-text reports revealed excellent NLP recognition. 4. The reporting tool successfully enables the integration of speech into structured reporting.

Abstract Computer-assisted reporting (CAR) tools were suggested to improve radiology report quality by context-sensitively recommending key imaging biomarkers. However, studies evaluating machine learning (ML) algorithms on cross-lingual ontological (RadLex) mappings for developing embedded CAR algorithms are lacking. Therefore, we compared ML algorithms developed on human expert-annotated features against those developed on fully automated cross-lingual (German to English) RadLex mappings using 206 CT reports of suspected stroke. Target label was whether the Alberta Stroke Programme Early CT Score (ASPECTS) should have been provided (yes/no:154/52). We focused on probabilistic outputs of ML-algorithms including tree-based methods, elastic net, support vector machines (SVMs) and fastText (linear classifier), which were evaluated in the same 5 × fivefold nested cross-validation framework. This allowed for model stacking and classifier rankings. Performance was evaluated using calibration metrics (AUC, brier score, log loss) and -plots. Contextual ML-based assistance recommending ASPECTS was feasible. SVMs showed the highest accuracies both on human-extracted- (87%) and RadLex features (findings:82.5%; impressions:85.4%). FastText achieved the highest accuracy (89.3%) and AUC (92%) on impressions. Boosted trees fitted on findings had the best calibration profile. Our approach provides guidance for choosing ML classifiers for CAR tools in fully automated and language-agnostic fashion using bag-of-RadLex terms on limited expert-labelled training data.