Background: Accurate preoperative assessment of lymphovascular invasion (LVI) remains challenging due to the high heterogeneity of gastric cancer (GC).
Purpose: To evaluate the feasibility of a subregion-based radiomics model using multiparametric MRI (mpMRI) for preoperative evaluation of LVI and to further assess its prognostic value.
Study type: Retrospective.
Subjects: A total of 878 GC patients from four centers: 313 training, 133 internal test, and 432 external validation cases.
Field strength/sequence: 1.5 T and 3 T/mpMRI including T2-weighted imaging (FSE/TSE), diffusion-weighted imaging (SS-EPI), and contrast-enhanced T1-weighted imaging (FFE/VIBE).
Assessment: The fuzzy c-means clustering was applied to subregion generation after manual segmentation. The subregional radiomics model was established using LVI-related features from a four-step extracted pipeline, with logistic regression, random forest, and support vector machine algorithms. The corresponding intra-tumoral subregion (ITS) index for each patient was obtained from the optimal subregional model. Subsequently, a combined model incorporating the ITS index and independent clinical characteristics was developed. Performance was further validated in test and validation cohorts. Additionally, the prognostic utility for overall survival (OS) and disease-free survival (DFS) was assessed in the follow-up cohort.
Statistical tests: Model area under the receiver operating characteristic curves (AUCs) was compared using net reclassification improvement (NRI) and integrated discrimination improvement (IDI). Kaplan-Meier survival analyses were conducted for prognostic evaluation. p < 0.05 was considered statistically significant.
Results: Pathological LVI-positive was detected in 448 (51.0%) patients. The combined model demonstrated satisfactory discrimination of LVI, achieving AUCs of 0.814 (training), 0.769 (test), and 0.758-0.783 (validation), outperforming the optimal subregional model with positive NRI and IDI across all cohorts. Furthermore, the ITS index maintained a significant association with OS (HR 33.50) and DFS (HR 30.00).
Data conclusion: The combined model, which integrated the ITS index derived from subregional radiomics with clinical factors, demonstrated robust performance in evaluating both LVI and survival outcomes in GC patients.
Evidence level: 3.
Technical efficacy: Stage 3.
Retained ballistic projectiles are a common consequence of firearm injuries in the United States. As the role of magnetic resonance imaging (MRI) in medicine continues to expand, clinicians increasingly encounter patients with retained bullets who would benefit from an MRI. However, the interaction between the strong magnetic field of an MRI and ferromagnetic implants can cause dangerous projectile movement and significant susceptibility artifacts, creating both safety and diagnostic concerns. Most bullets encountered in the United States are composed of copper alloy jackets and lead cores, which are diamagnetic and generally safe for MRI. Yet, some bullets are manufactured with steel components and are ferromagnetic. Radiologists have identified several approaches to determine the MRI compatibility of retained bullets, but few definitive guidelines exist to decide which of these patients can safely undergo MRI. In this article, we review the interactions of ferromagnetic bullets with the magnetic field of an MRI, list the material construction of various ammunition, discuss various proposed methods for determining the MR safety of retained projectiles, and synthesize literature recommendations into simplified algorithms that can be utilized to effectively triage patients with retained ballistic projectiles. Standardizing this screening process enables clinicians to stratify patient risk and avoid unnecessary MRI exclusions.
5.
Stage 5—Improvements in Patient Care.
Background: Subamniotic or subchorionic hematoma (SAH/SCH) is associated with diverse pregnancy outcomes. The clinical implications of accompanying oligohydramnios and hemorrhagic amniotic fluid on MRI remain unclear.
Purpose: To investigate the importance of oligohydramnios and hemorrhagic amniotic fluid on placental MRI for SAH/SCH in risk stratification.
Study type: Retrospective.
Population: Seventy-one singleton pregnancies with SAH/SCH identified on placental MRI performed during the second or third trimesters, from 2016 to 2023.
Field strength/sequence: 1.5 T, Fat-saturated T1-weighted gradient echo and half-Fourier-acquired single-shot turbo spin echo sequences.
Assessment: Cases were classified into three groups: Groups A (oligohydramnios and hemorrhagic amniotic fluid), B (either oligohydramnios or hemorrhagic amniotic fluid), and C (SAH or SCH only). Groups B and C were subclassified as B-1 (oligohydramnios), B-2 (hemorrhagic amniotic fluid), C-1 (detected hematoma on ultrasound before MRI), and C-2 (incidentally detected hematoma on MRI). Unfavorable obstetric outcome (abortion or birth before 34 gestational weeks) and neonatal outcome (duration of neonatal intensive care unit [NICU] stay) were compared.
Statistical tests: Fisher's exact test, Kruskal-Wallis test, Mann-Whitney U test, and Kaplan-Meier analysis with Log-rank test. Significance was determined at p < 0.05.
Results: Unfavorable obstetric outcomes were significantly higher in Group A (11/12, 91.7%) than groups B (6/17, 35.3%) and C (9/42, 21.4%). Significant differences were found among the five subclassified groups, most notably between B-1 and B-2. The median duration of NICU stay was 87, 30.5, 0, 25, and 8 days in Groups A (n = 12), B-1 (n = 5), B-2 (n = 12), C-1 (n = 11), and C-2 (n = 31), respectively. Group A showed the worst neonatal outcomes.
Data conclusion: MRI findings of oligohydramnios and/or hemorrhagic amniotic fluid in pregnancies with SAH/SCH are associated with adverse obstetric and neonatal outcomes, supporting risk stratification.
Evidence level: 4.
Technical efficacy: Stage 5.
Background: Accurate assessment of 90-day functional outcomes after anterior circulation large vessel occlusion (LVO) stroke remains challenging. Conventional models relying on a single data dimension have limited assessment power, suggesting that a multidimensional integration strategy could enhance evaluations.
Purpose: To develop and validate an interpretable machine learning model that integrates radiomics, infarct location, brain frailty, and clinical variables for assessing 90-day functional outcomes in LVO stroke.
Study type: Retrospective.
Population: 1051 patients with anterior circulation LVO stroke (mean age 63 ± 13 years; 722 males) from five centers (2018-2023). Eight hundred and seventy-five patients from four centers formed the training (n = 612) and internal validation (n = 263) cohorts, while 176 from the fifth center comprised the external validation cohort.
Field strength/sequence: T1-weighted spin-echo imaging (T1WI), T2-weighted spin-echo imaging (T2WI), T2-weighted fluid-attenuated inversion recovery (FLAIR) imaging, and diffusion-weighted echo-planar imaging (DWI).
Assessment: Infarct volume and radiomic features were extracted from DWI. Infarct location was assessed using the Alberta Stroke Program Early CT Score. Brain frailty was evaluated using cortical/subcortical atrophy, white matter hyperintensity (WMH), and old infarcts. Least Absolute Shrinkage and Selection Operator (LASSO) regression was used for feature selection.
Statistical tests: Chi-square, Fisher's exact, t-test, Mann-Whitney U, area under the receiver operating characteristic curve (AUC), DeLong test, decision curve analysis, calibration curves, sensitivity, specificity, positive predictive value, negative predictive value, F1 score. Significance level p < 0.05.
Results: The fused model outperformed all single-dimension models (ΔAUC = 0.12-0.22), achieving AUCs of 0.87 (training), 0.84 (internal validation), and 0.86 (external validation). The fused model achieved a sensitivity and a specificity of 0.80 in the external validation cohort. Features with the highest mean absolute Shapley Additive Explanations (SHAP) values included lentiform nucleus lesion burden (SHAP = 0.083), WMH (SHAP = 0.080), and lesion burden in the M6 region (posterior middle cerebral artery territory; SHAP = 0.061).
Data conclusion: Integration of infarct location, brain frailty, radiomics, and clinical features improved the 90-day outcome assessment in anterior circulation LVO stroke, providing an interpretable tool for personalized prognosis.
Level of evidence: 3:
Technical efficacy stage: 2.
Background: Multi-center imaging studies create large-scale data that are useful for identifying pathological patterns and robust training of deep learning models. However, variation due to site and scanner differences can confound analyses, emphasizing the need for harmonization.
Purpose: To evaluate scanner-related differences in T1w and T2-FLAIR images in the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset and assess the performance of publicly available image-level harmonization tools.
Study type: Retrospective.
Population: Scanner group analysis: 1143 ADNI3 subjects (233 GE, 173 Philips, 250 Siemens, with 487 Siemens subjects used as an independent reference group). Within-subject comparison: paired multi-vendor scan sessions from 8 subjects.
Field strength/sequence: 3.0T, T1w, and T2-FLAIR MRI sequences.
Assessment: Gray/white matter contrast ratio (G/W ratio), white matter hyperintensity (WMH) volume, and image feature similarity metrics (Fréchet Inception Distance [FID], Learned Perceptual Image Patch Similarity [LPIPS]) were compared across scanner vendors before and after harmonization with statistical (ComBat) and deep learning (HACA3) algorithms.
Statistical tests: One-way ANOVA and post hoc Games-Howell tests were conducted to assess differences between scanner groups across image pipelines (baseline, post-harmonization). Repeated-measures ANOVA and post hoc paired t-tests with Bonferroni correction were used to evaluate similarity metric changes pre- and post-harmonization for multi-vendor subjects. We defined statistical significance as p < 0.05.
Results: At baseline, significant image differences in G/W ratio and WMH volumes between vendors were identified. Both ComBat and HACA3 harmonization improved G/W ratio consistency for T1w and T2-FLAIR imaging across vendors, particularly for GE T2-FLAIRs. HACA3 led to the best similarity between scanner datasets: mean FID T1w/T2-FLAIR: 10.45/14.62 (Baseline); 7.45/11.71 (ComBat); 5.60/8.91 (HACA3). Only HACA3 harmonization resulted in non-significant differences between vendors for WMH volume.
Data conclusion: HACA3 deep learning harmonization outperformed a statistical method, ComBat, improving MR contrast consistency and feature similarity across vendors. However, difficulties in harmonizing T2-FLAIRs highlight limitations in current multi-contrast MR harmonization tools.
Evidence level: 3.
Technical efficacy: Stage 1.
Background: Measurement error in imaging reduces statistical power and potentially biases parameter estimation, compromising study reliability.
Purpose: To introduce a dual data collection design (reliability and main datasets) to quantify measurement error and apply regression calibration to correct error-prone imaging markers, thereby improving biomarker-outcome estimation, statistical power, and sample size planning.
Study type: Prospective (reliability) and retrospective (regression calibration).
Population: 65 healthy volunteers (mean age: 23.2), 60 age and sex matched with 34 epilepsy patients (mean age: 28.7).
Field strength/sequence: 3.0 T, MR fingerprinting (MRF) and T1-weighted (T1w) MPRAGE.
Assessment: Three-dimensional brain scan-rescan data were acquired in 5 volunteers (6 identical acquisitions per volunteer across 3 scanners) to estimate reliability coefficients ( ) for MRF T1 and T1w signal intensity (SI) mean and standard deviation (SD). These coefficients were applied in regression calibration to correct imaging markers in the epilepsy cohort. Effect sizes for distinguishing lesional from control were compared before and after correction. Simulations evaluated the impact of additive and proportional bias on sample size, statistical power, and association estimates under single and multi-scanner scenarios.
Statistical test: Reliability coefficient, Cohen's d, regression calibration, generalized estimation equations.
Results: MRF T1 markers exhibited higher reliability ( = 0.887-0.941) than T1w SI markers with site effects ( = 0.246-0.554). Regression calibration increased effect size more for T1w SI mean (333.22% increase) than for MRF T1 mean (12.57% increase). In multi-site simulations, regression calibration alone achieved unbiased estimate under small site effects (additive and proportional SD ≤ 0.2), whereas under larger site effects (additive SD ≥ 0.5) only the combined regression calibration and Combat produced near-zero bias (-0.024), outperforming naïve analysis (-0.423).
Data conclusion: The dual data acquisition design with regression calibration restores attenuated imaging biomarker associations, improves statistical power, and informs sampling requirements, thus enhancing reliability and generalizability in multi-site imaging studies.
Evidence level: 3.
Technical efficacy: 2.
Background: Early detection of knee osteoarthritis (OA) is important. Spin-lattice relaxation in the rotating frame (T1ρ) mapping is sensitive to early cartilage changes, but the mono-exponential (ME) model may be limited. Multi-component models can capture more tissue complexity, but their diagnostic advantage has not been validated.
Purpose: To evaluate if stretched- (SE) and bi-exponential (BE) T1ρ models can improve early knee OA detection over the ME model.
Study type: Case-control study.
Population: Twenty-six healthy subjects (mean age 51.5) and 26 early knee OA patients (mean age 61.8).
Field strength/sequence: T1ρ-prepared Turbo FLASH sequence at 3 T field strength.
Assessment: T1ρ parameters from three exponential models were adjusted for age. To maximize group separability, the parameters were combined into single discriminators for both global knee cartilage and six anatomical sub-regions. Diagnostic performance was assessed based on the ability of these combined models to distinguish early OA.
Statistical tests: Parameters were adjusted for age. Mann-Whitney U-test (group comparisons), linear discriminant analysis (LDA), and area under the receiver operating characteristic (ROC) curve (AUC) with bootstrapped 95% confidence intervals (CI). Significance level set at p < 0.05, using the false discovery rate (FDR) to correct for multiple comparisons.
Results: In the global analysis, no model demonstrated significant diagnostic performance (p-values of 0.63, 0.96, 0.63 for ME, SE, and BE). Multi-regional SE model (AUC = 0.83, CI: 0.72, 0.93) significantly distinguished OA and healthy groups. Calibration analysis showed the SE model had the lowest Brier score (0.17), significantly better than the ME model (0.26).
Data conclusion: Sub-regional analysis of T1ρ parameter maps suggests an improvement in diagnostic performance for early knee OA compared to globally averaged measurements. The stretched-exponential model showed the most promise. However, small sample size and wide confidence intervals highlight the need for further validation with a larger cohort before clinical utility claims can be made.
Evidence level: 4.
Technical efficacy: Stage 2.
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：