Academic Radiology最新文献

Rationale and objectives: As of September 2024, the FDA requires that breast imaging practices inform women about their breast density. This study aimed to evaluate the impact of breast density on the performance metrics of screening digital breast tomosynthesis (DBT) examinations.

Materials and methods: We retrospectively reviewed screening DBT examinations performed from 2013 to 2019 at a single academic medical center. Performance metrics were calculated according to the 5th Edition of the BI-RADS Atlas. Associations between breast density and screening performance were examined using multivariable logistic regression with generalized estimating equations.

Results: The cohort included 111,143 women (mean age, 59 ± 11 years) with 301,400 DBT examinations. Breast density was almost entirely fatty (category A) in 8.8%, scattered areas of fibroglandular density (B) in 50.5%, heterogeneously dense (C) in 36.9%, and extremely dense (D) in 3.8%. Cancer detection rates (CDR, per 1000 exams) were 3.4, 5.6, 5.2, and 3.7 for categories A-D, respectively. Sensitivities were 92.8%, 90.1%, 81.0%, and 61.8%. Specificities were 96.7%, 94.4%, 92.5%, and 93.3%. Category D was associated with significantly lower sensitivity than each of the other categories (adjusted odds ratios [aOR] 0.19-0.43, p<0.01 for all). It was associated with significantly lower specificity than almost entirely fatty tissue (aOR 0.64, p<0.001) but not the other two density categories.

Conclusion: Dense breast tissue significantly decreases the sensitivity of screening DBT. These findings highlight the need to report and consider breast density in screening recommendations and necessitate further research on more effective screening regimens for women with dense breast tissue.

Rationale and objectives: To evaluate the feasibility of dual-energy cardiac computed tomography (DE-CCT) in detecting left atrial fibrosis (LAF), atrial fibrillation (AF) classifications, and postablation recurrence, using left atrial cardiac magnetic resonance (LA-CMR) as reference.

Materials and methods: This retrospective study included 81 pre-ablation AF patients without concomitant other arrhythmia or prior cardiac surgery. DE-CCT (utilizing dark-blood images with 5-minute post-injection iodine quantification) was quantitatively compared to concurrent LA-CMR to assess LAF, AF types, and recurrence risk.

Results: For LAF detection, dark-blood images manifested high specificity (100%, 253/253 segments), moderate sensitivity (77.07%, 242/314 segments), and satisfactory accuracy (87.30%, 495/567 segments) compared with LA-CMR. Increased iodine ratio of LAF to blood pool (OR = 1.214, p = 0.041), incremental LAF segments in DE-CCT (OR = 1.739, p = 0.050), elevated brain natriuretic peptide (BNP), higher body mass index, and worse degree of left atrial volume index (LAVI) were significant risk indicators for PeAF (AUC=0.965, p<0.001). Concerning LA-CMR, independent risk factors for PeAF were also elevated LAF segments in CMR (OR = 2.108, p = 0.008), BNP levels, and LAVI grades (AUC=0.954, p<0.001). Cumulatively, 59 patients completed 1-year follow-up postablation, with 30.5% experiencing recrudescence. Significant predictors for 1-year relapse included increased LAF segments in DE-CCT (OR = 2.130, p = 0.009) or LA-CMR (OR = 1.740, p = 0.039), PeAF, thyrotoxicosis, and larger CHA2DS2-VASc scores, yielding AUCs of 0.895 (p<0.001) and 0.869 (p<0.001), respectively. No significant differences were found between DE-CCT and LA-CMR in discerning PeAF (p = 0.451) and early relapse (p = 0.114).

Conclusions: DE-CCT is a viable alternative for evaluating LAF, AF subtypes, and high-risk recurrence, comparable to LA-CMR, potentially enabling personalized therapy and guiding further studies on higher-resolution multi-energy CT.

Rationale and objectives: The metabolic predominance of hepatocellular carcinoma (HCC) on dual-tracer PET/CT with C-11 acetate and F-18 fluorodeoxyglucose (FDG) reflects tumor differentiation, but a practical MRI-based biomarker for predicting this phenotype remains needed. This study aimed to assess whether enhancement patterns on gadoxetic acid-enhanced liver MRI can differentiate the metabolic predominance of HCC.

Materials and methods: This exploratory retrospective study included patients with HCC who underwent both dual-tracer PET/CT and liver MRI between January 2015 and December 2021. HCC lesions were categorized as acetate- or FDG-dominant based on tracer avidity. Signal intensity (SI) in each MRI phase was quantified using the percentage signal ratio (PSR), defined as the ratio of SI in adjacent normal liver parenchyma to that in the lesion. The area under the receiver operating characteristic curve (AUC) was measured to determine the MRI phase that best distinguished the metabolic predominance of HCC lesions.

Results: In 38 patients (mean age, 62.0±8.0 years) with 48 HCC lesions, median PSR values were lower in acetate-dominant group (28 lesions) than in FDG-dominant group (20 lesions) during the early arterial (64.7 vs. 113.9; P<0.001), late arterial (103.8 vs. 135.4; P<0.001), and portal venous (114.5 vs. 142.3; P = 0.001) phases, but not in the hepatobiliary phase (171.6 vs. 161.9; P = 0.390). The AUC was highest in the early arterial phase (0.99; 95% CI: 0.96, 1.00; P<0.001), where a PSR cutoff of 95.2 effectively distinguished the metabolic predominance of HCC.

Conclusion: PSR on liver MRI, especially in the early arterial phase, differentiates acetate- from FDG-dominant HCCs, indicating metabolic predominance.

Rationale and objectives: Contrast-enhanced ultrasound (CEUS) is an effective method for assessing varicocele (VC) hemodynamics; however, its value in predicting improvements in semen parameters after varicocelectomy remains unclear. This study investigated whether CEUS-based hemodynamic patterns could predict surgical outcomes and developed and validated a novel predictive model.

Materials and methods: In this prospective cohort study, 187 patients with VC undergoing microscopic varicocelectomy were included. A total of 23 clinical, grayscale ultrasound, and CEUS-based hemodynamic parameters were collected. The primary outcome was semen parameter improvement (≥25% increase in total motile sperm count) six months postoperatively. Elastic net regression (ENR) selected key predictors, which were then included in a multivariate logistic regression model. The resulting nomogram's performance was evaluated using receiver operating characteristic curves, calibration, and decision curve analysis, and was internally validated using 1000 bootstrap resamples.

Results: A total of 118 patients achieved post-operative semen parameter improvement. From 23 candidate variables, ENR identified three independent predictors as follows: left and right CEUS hemodynamic patterns and follicle-stimulating hormone. The nomogram demonstrated excellent discrimination, with an area under the curve of 0.822 and good calibration (Hosmer-Lemeshow test, P = 0.405). Decision curve analysis confirmed the clinical utility of the model across a wide range of risk thresholds.

Conclusion: CEUS-based pampiniform plexus hemodynamic patterns are independent predictors of post-operative semen parameter improvement in patients with VC. We successfully developed and validated a novel predictive nomogram. This study provides an effective, non-invasive tool for surgical decision-making and expands the clinical application of CEUS.

Background: Lung cancer remains the leading cause of cancer-related death worldwide. Low-dose computed tomography (LDCT) is the current gold standard for screening, but radiation exposure remains a concern. Cone-Beam Computed Tomography (CBCT) offers ultra-low-dose acquisition but has not yet been tested for thoracic screening.

Purpose: To evaluate the feasibility of using a latest-generation CBCT system for thoracic imaging in a lung cancer screening context.

Materials and methods: 88 patients, current or former smokers, were retrospectively included. They underwent thoracic CBCT (7 G Dual Energy system, NewTom, Italy) between July 2024 and April 2025 at a preventive health center, with either a standard-dose protocol (n = 48) or a low-dose protocol (n = 40). Two board-certified radiologists independently assessed pulmonary findings and image quality. Radiation doses were compared. Statistical analysis included mixed-effects models and interobserver agreement.

Results: The radiation exposure was significantly lower in the low-dose group, with a median cumulative DLP of 42.8 mGy·cm, compared to 107.7 mGy·cm, representing a 2.5-fold reduction in radiation (p<0.001). Nodule detection occurred in 41% of patients overall. Overall image quality was rated good/excellent in 85% (low-dose) vs 83% (standard-dose); Risk Difference (RD) = +2% [IC95% -12; +16], Risk Ratio (RR) = 1.03. Diagnostic confidence was high in 82% vs 83%; RD = -1% [-14; +12], RR = 0.99. Spatial resolution was slightly lower in the low-dose group (OR = 0.51, p = 0.04). Interobserver agreement for nodule detection was substantial across both groups (κ = 0.73-0.83).

Conclusion: Low-dose CBCT offers substantial radiation dose reduction without compromising diagnostic confidence or interobserver reliability. CBCT may represent a promising, cost-effective alternative or complement to low-dose CT in preventive lung imaging programs, but prospective studies comparing CBCT to LDCT are warranted.

Rationale and objectives: Lung cancer screening (LCS) with low-dose computed tomography (LDCT) reduces lung cancer mortality by 20% and all-cause mortality by 6.7%. In 2013, the United States Preventive Services Task Force (USPSTF) recommended LCS with LDCT for adults aged 55-80 with a ≥30 pack-year smoking history who currently smoke or quit within the past 15 years. In 2021, these recommendations grew to include more at-risk populations by lowering the screening age to 50 years and reducing the smoking history threshold to 20 pack-years. We assessed the feasibility of a brief, multilingual smoking-history questionnaire in radiology waiting areas to identify LCS eligibility and standardize notification to primary care providers (PCPs) in a safety-net hospital.

Materials and methods: Quality improvement initiative, exempt from formal IRB review and the requirement for informed consent. Over an 18-month period between 2021 and 2024, we administered a voluntary smoking history questionnaire assessing demographics, lung cancer risk, LCS eligibility, and relevant medical and family history to all patients arriving for imaging appointments.

Results: From an estimated total of 54,000 surveys distributed, 6160 questionnaires were collected (11.4% response rate), and 373 patients (6.0%) self-identified as eligible for LCS based on either 2013 or 2021 USPSTF criteria. Among these patients, 202 (54.2%) were not currently undergoing LCS. Following PCP notification of their patients' LCS eligibility, only 19 of the 202 patients (9.4%) subsequently had baseline LCS exams ordered. These proportions reflect feasibility/process and are not evidence of effectiveness.

Conclusion: A brief, multilingual smoking-history questionnaire in radiology waiting areas in a safety-net setting was feasible to implement. LCS rates remain low despite patient self-identification of LCS eligibility and PCP notification. This low uptake highlights the challenges of LCS and may reflect patient, healthcare provider, and systems-level barriers faced by patients in safety-net hospitals, such as financial constraints and limited healthcare access.

Objective: Perineural invasion (PNI) and lymphovascular invasion (LVI) are critical predictors of aggressive behavior and poor prognosis in prostate cancer (PCa), yet their diagnosis relies on postoperative histopathology. This study aims to develop a noninvasive radiomic model based on biparametric magnetic resonance imaging (bpMRI) for preoperative prediction of PNI and LVI.

Methods: A total of 256 patients with pathologically confirmed PCa who underwent radical prostatectomy were retrospectively enrolled. Patients from Center 1 (n = 179) constituted the training set, while those from Center 2 (n = 77) formed the external test set. A rigorous imaging-pathology correlation protocol was applied to ensure accurate lesion matching. Inter-observer variability in segmentation was assessed (ICC > 0.75 for 85% of features), with final ROIs determined by consensus. Radiomic features were extracted from T2-weighted and diffusion-weighted imaging. Feature selection was performed using Spearman's correlation and LASSO algorithm. Multiple machine learning classifiers were constructed and interpreted with SHAP.

Results: The best-performing model for PNI prediction was Multilayer Perceptron (MLP), with an AUC of 0.805 (95% CI: 0.741-0.869) in the training set and 0.795 (95% CI: 0.698-0.896) in the test set. For LVI prediction, Logistic Regression achieved the highest performance, with an AUC of 0.859 (95% CI: 0.804-0.914) in the training set and 0.810 (95% CI: 0.714-0.906) in the test set. Calibration curves and decision curve analysis indicated good model accuracy and clinical utility.

Conclusion: Radiomic models derived from bpMRI can noninvasively and robustly predict PNI and LVI in PCa, demonstrating good generalizability across independent cohorts.

Rationale and objectives: Multiple approaches are available for defining the tibial anatomical axis when measuring the posterior tibial slope (PTS) with MRI. This study aims to evaluate the reliability of PTS measurements on MRI when the anatomical axis is defined at different tibial levels.

Materials and methods: This study included 103 patients who underwent two distinct MRI examinations of the same knee between 2018 and 2023, with each pair of scans performed within a one-year interval and without significant morphological changes. Two anatomical axes were defined: one below the tibial plateau and another below the tibial tuberosity. The medial and lateral posterior tibial slopes (MPTS and LPTS) were measured relative to each axis. Reliability was evaluated by assessing inter-scan (test-retest), intra-rater, and inter-rater agreement. Variability between scans was further examined using Bland-Altman limits of agreement (LOA).

Results: Defining the anatomical axis below the tibial plateau resulted in only moderate inter-scan agreement (MPTS: ICC = 0.651; LPTS: ICC = 0.618), whereas defining it below the tibial tuberosity yielded good agreement (MPTS: ICC = 0.864; LPTS: ICC = 0.852). The 95% LOA between scans for MPTS were -6.8° to 6.4° with the plateau-based axis and -4.4° to 4.8° with the tuberosity-based axis, while those for LPTS were -6.4° to 6.5° and -4.6° to 4.5°, respectively. Both definitions of the axis demonstrated good to excellent intra- and inter-rater reliability for MPTS and LPTS measurements.

Conclusion: Defining the anatomical axis below the tibial tuberosity yields more reliable PTS measurements, with better inter-scan agreement and good intra- and inter-rater agreement.