Degenerative spinal disease is a leading cause of lower back pain worldwide, impairing mobility and quality of life. Early degeneration is increasingly seen in younger adults due to sedentary lifestyles and occupational stress. Conventional MRI, particularly T2-weighted imaging, is limited in detecting early microstructural changes. Diffusion kurtosis imaging (DKI), by quantifying non-Gaussian water diffusion, offers enhanced sensitivity for identifying subtle alterations preceding overt degeneration. To evaluate the ability of quantitative DKI parameters to detect early intervertebral disc degeneration by correlating them with the Pfirrmann grading system in the thoracolumbar spine. This prospective study included 76 participants, 54 with degenerative spine disease and 22 healthy controls. MRI was performed on a 3 T scanner using sagittal T1-, T2-, and DKI sequences. Parameter maps included mean kurtosis (MK), axial kurtosis (AK), radial kurtosis (RK), fractional anisotropy (FA), and diffusivity indices (MD, AD, RD, KA). Regions of interest were placed in the nucleus pulposus and annulus fibrosus, with subregional analysis of the anterior and posterior annulus. Grade V discs were excluded, yielding (1,287) discs (Grades I-IV). Statistical analyses included group comparisons and Spearman's correlation. MK, AK, and RK were significantly higher in the degenerative group (p < 0.001), while MD, AD, and RD were substantially lower (p < 0.001). Correlations were strongest in the mid-to-lower lumbar levels, particularly within the nucleus pulposus and anterior annulus. DKI enables quantitative characterisation of disc microstructure and demonstrates diagnostic potential for differentiating degenerative from non-degenerative discs, supporting its role as an emerging imaging biomarker for early degeneration assessment.

Background: Artificial intelligence (AI) techniques are increasingly applied to magnetic resonance imaging (MRI) for detecting temporomandibular joint (TMJ) anomalies; however, their overall diagnostic accuracy and generalizability remain uncertain. Objectives: To systematically review and meta-analyse the diagnostic performance of AI models for TMJ anomaly detection on MRI and to identify factors influencing model performance. Methods: A comprehensive search of PubMed, Scopus, Embase, and Web of Science was conducted for studies published between January 2015 and September 2025. Two reviewers independently screened and extracted data. Eligible studies developed and tested AI, machine learning, or deep learning models on human TMJ MRI and reported quantitative performance metrics. Risk of bias was assessed using the QUADAS-2 tool. Pooled sensitivity and specificity were estimated using a bivariate random-effects model, while pooled accuracy was derived using logit transformation. Heterogeneity ( I ² ) was explored through subgroup analyses by model architecture and validation strategy. Results: Fourteen studies were included in the systematic review, of which six met the criteria for meta-analysis. Across these six studies, 18 models were analyzed for accuracy, 29 for sensitivity, and 24 for specificity. The pooled diagnostic accuracy was 0.487 (95% CI 0.403–0.571), with pooled sensitivity and specificity of 0.399 (95% CI 0.348–0.450) and 0.399 (95% CI 0.343–0.456), respectively, all showing substantial heterogeneity ( I ² > 90%). Subgroup analyses indicated that advanced architectures such as ResNet-18, Inception v3, and EfficientNet-b4 achieved higher and more consistent diagnostic performance. Conclusions: Advanced deep learning architectures such as ResNet-18, Inception v3, and EfficientNet-b4 demonstrated superior diagnostic performance for detecting temporomandibular joint anomalies on MRI. These findings highlight the potential of AI-assisted MRI interpretation to improve diagnostic consistency, efficiency, and early detection of TMJ pathology. However, substantial heterogeneity and limited external validation currently limit clinical translation. Standardized multicenter studies and transparent model validation are essential to ensure reliable integration of AI tools into clinical TMJ imaging workflows.

Artificial intelligence (AI)-driven image reconstruction has shown potential for enhancing image quality in CT angiography while reducing radiation exposure. However, reproducibility and methodological transparency are essential for real-world adoption. To compare image quality and radiation dose of cerebral CT angiography reconstructed with conventional iterative reconstruction (IR) and AI-based Precise Imaging (PI) at 100 kVp using a reduced Dose Right Index (DRI-15). In this prospective single-center study, 68 patients underwent 100 kVp cerebral CT angiography with a 50% reduced Dose Right Index (DRI-15 vs. DRI-36). Identical raw datasets were reconstructed using IR (iDose⁴) and PI. Quantitative analysis included attenuation, image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR), while two blinded radiologists independently performed qualitative assessment using a 5-point Likert scale. Radiation dose metrics (CTDIvol, DLP, effective dose) were recorded. Standardized acquisition protocols, fixed contrast injection, and consistent ROI placement ensured reproducibility. PI reconstructions yielded significantly lower noise and higher SNR and CNR than IR (all p < 0.001; r = 0.61-0.71), with no difference in vascular attenuation (p = 0.118). Qualitative scores were superior for PI across all criteria (p < 0.01), while inter-rater agreement was substantial for both methods, slightly higher with IR. The median effective dose was 0.785 mSv (IQR 0.71-0.90). AI-based PI reconstruction enhances image quality without increasing radiation dose or altering vascular attenuation, consistently outperforming IR and supporting its use in low-dose cerebral CT angiography, especially in dose-sensitive patients.

Degenerative cervical spine disorders are common and may cause progressive neurological impairment if not detected early. Conventional magnetic resonance imaging has limitations in identifying subtle microstructural changes. Diffusion kurtosis imaging (DKI) is sensitive to non-Gaussian water diffusion and may improve diagnostic and prognostic evaluation, but its utility in cervical spine degeneration remains unclear. The aim of this study is to systematically assess the diagnostic and prognostic utility of DKI-derived metrics in cervical spine degeneration. A systematic review and meta-analysis were conducted following Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 guidelines. PubMed, Scopus, Web of Science, and Embase were searched (2005-2025). Eligible studies evaluated mean kurtosis (MK), fractional anisotropy (FA), and mean diffusivity (MD) in grey matter (GM) and white matter (WM). Methodological quality was assessed using QUADAS-2 (University of Bristol, Bristol, UK). Random-effects models were applied to calculate pooled standardised mean differences, and heterogeneity was quantified using Q and I² statistics. Eight studies (n=273) were included; three underwent meta-analysis. In GM, MK and FA showed significant differences, with higher values in patients compared to controls. White matter findings were nonsignificant with moderate heterogeneity. MD showed no significant differences, with substantial heterogeneity in WM. Several studies reported associations between DKI metrics and clinical severity scores, supporting potential prognostic value. DKI shows promise as a sensitive tool for detecting microstructural alterations in the cervical spinal cord. Significant changes in MK and FA in GM suggest consistent involvement, while findings in WM and MD remain variable. Further standardised, large-scale studies are needed to confirm clinical utility.

Background Cerebral computed tomography angiography (CTA) is widely used to assess neurovascular disorders, but venous contamination often obscures arteries. Optimizing bolus-tracking thresholds is crucial, yet patient factors influencing contrast dynamics and the value of radiation dose indices in head CTA remain unclear. Objectives To optimize bolus-tracking thresholds in cerebral CTA by examining patient-related influences on enhancement and radiation metrics. Methods 126 adults undergoing cerebral CTA were evaluated in this prospective study. Demographics, physiologic parameters, peak enhancement time (PET), peak enhancement attenuation (PEA), and dose indices (CTDIvol, SSDE) were recorded. Linear regression identified predictors of enhancement. Two blinded radiologists graded venous contamination. ROC analysis, including age subgroups, determined the optimal HU threshold. Results Median age was 55.5 years; 70% were male. PET rose with age (+0.086 s/year, p < 0.001) and was shorter in females (–2.39 s, p = 0.003). PEA increased with threshold (+1.03 HU/unit, p < 0.001). Arterial enhancement was higher in females (+40.7 HU, p < 0.001) and patients ≥60 years (+70 HU, p < 0.001). Venous enhancement correlated with PET (p = 0.023) and systolic pressure (p = 0.002). ROC analysis showed an optimal threshold of 105 ± 5 HU (AUC = 0.634; sensitivity 88.4%, specificity 77.1%). CTDIvol, but not SSDE, correlated with weight (p = 0.015). Conclusion Intrinsic (age, gender) and extrinsic (threshold) factors shape CTA enhancement. A 105 ± 5 HU threshold reduces venous contamination, especially in younger patients. CTDIvol remains the preferred dose index. Findings support individualized, resource-efficient CTA protocols aligned with UN SDGs 3, 9, and 12.