Acta radiologica最新文献

BackgroundMagnetic resonance imaging (MRI) scores have been validated for monitoring Crohn's disease (CD) but are not widely used in clinical practice.PurposeTo evaluate treatment response in patients with active CD using simple imaging markers compared to complex multifactorial scores.Material and MethodsThis was a post-hoc analysis of MRI-enterocolonographies performed in a prospective, blinded, multicenter study. Patients with endoscopically active CD completed ileocolonoscopy and MRI with intravenous contrast and diffusion-weighted sequences before and after medical treatment. Response was defined as ≥50% reduction of the Simple Endoscopic Score for Crohn's Disease. Changes in bowel wall thickness (BWT) and apparent diffusion coefficient (ADC) were compared to the magnetic resonance index of activity (MaRIA), simplified MaRIA, and Clermont score.ResultsA total of 42 patients entered the analysis, and endoscopic response was achieved in 19 (45.2%). All activity scores improved in patients with endoscopic response compared to non-responders (P <0.05). The relative reduction of BWT for the most severely affected bowel segment (area under the curve [AUC]=0.76, 95% confidence interval [CI]=0.61-0.91) trended towards a higher diagnostic accuracy compared to the global MaRIA score (AUC=0.63, 95% CI=0.45-0.81; P = 0.1). The per-segment response was more accurately evaluated with BWT compared to the segmental MaRIA score (AUC=0.82 and 0.67, respectively; P = 0.05). A 17% decrease in BWT resulted in an optimal sensitivity and specificity of 78.9% (95% CI=54.4-93.9) and 85.2% (95% CI=66.3-95.8), respectively.ConclusionBWT is not inferior to MaRIA when determining treatment response in CD. This measure is simple to perform and does not require intravenous contrast.

BackgroundAccurate preoperative assessment of endometrial cancer (EC) is crucial in young women who may be eligible for fertility-preserving therapy, which is generally limited to patients with grade 1, endometrioid-type tumors without myometrial invasion (MI).PurposeTo evaluate the utility of quantitative parameters derived from intravoxel incoherent motion (IVIM) and diffusion kurtosis imaging (DKI) for improving the diagnostic performance of magnetic resonance imaging (MRI).Material and MethodsThis retrospective study included 107 patients diagnosed with EC (mean age = 59 years; age range = 25-89 years) who underwent preoperative MRI, including multiple b-value (0-2000 s/mm²) diffusion-weighted imaging, between January 2022 and March 2024. Quantitative parameters were extracted from the mono-exponential (ADC), IVIM (Di, D*, f), and DKI (Dk, K) models and compared across clinical and pathological features.ResultsADC, Di, and Dk values were significantly higher in patients without MI (P = 0.015, 0.035, and 0.005, respectively). Di and Dk were significantly higher (P = 0.003 and 0.016), and K was significantly lower (P = 0.013) in the G1 group. Patients eligible for fertility preservation had significantly higher ADC, Di, and Dk values (P = 0.002, 0.002, and 0.001) and significantly lower K values (P = 0.044). The overall diagnostic performance of these parameters was moderate (area under the curve < 0.70).ConclusionIVIM and DKI-derived metrics may enhance preoperative assessment of tumor grade and MI, supporting decisions regarding fertility-preserving treatment.

BackgroundAutomatic exposure control (AEC) devices are necessary to reduce the radiation dose and enhance image quality in radiography. However, AEC devices are not being used in portable X-ray systems due to technical issues.PurposeTo evaluate the radiation dose and image quality of a new AEC for a wireless portable X-ray system compared to the standard fixed radiation dose of manual controls in chest radiography.Material and MethodsThis retrospective analysis evaluated the performance of a portable X-ray system equipped with software-based AEC on anteroposterior chest radiographs. A quantitative evaluation comparing the manual and AEC groups was conducted to ascertain the exposure index (EI) and dose-area product (DAP). We performed observer-based analysis of image quality for lowest exposure and average exposure images in the manual and AEC groups.ResultsOverall, 2093 examinations were conducted in 467 patients. AEC yielded a statistically significant reduction in the DAP (AEC: 1.64 dGycm²; manual: 1.99 dGycm²) and EI (AEC: 266; manual: 393) compared to manual control (P <0.001). Evaluation of the average and low exposure images revealed that the DAP and EI were consistently lower in the AEC group (P <0.01). Observer-based analysis of the lowest exposure images revealed that AEC yielded a statistically significant higher score compared to manual control (12.5 manual, 13.4 AEC; P <0.001).ConclusionAEC within a wireless portable X-ray system improves subjective image quality while reducing radiation exposure, as substantiated by quantitative and qualitative metrics.

BackgroundHuman epidermal growth factor receptor 2 (HER-2) is a key biomarker in breast cancer, guiding therapeutic decisions and prognosis. Conventional assessment relies on tissue biopsy, an invasive procedure that may impose both physical and financial burdens on patients.PurposeTo develop an interpretable deep learning-based imaging framework capable of non-invasively predicting preoperative HER-2 expression.Material and MethodsWe retrospectively analyzed magnetic resonance imaging data and clinical records from 450 patients with pathologically confirmed HER-2 status across four medical centers. Several conventional machine learning algorithms were compared with a deep neural network model. A ResNet-based architecture was used to generate a probability score (D-score) reflecting the likelihood of HER-2 positivity. Independent clinical predictors were identified through logistic regression and integrated with the D-score to construct a combined predictive framework. Model performance was evaluated using receiver operating characteristic analysis, and interpretability techniques were applied to visualize the contribution of individual features.ResultsThe combined deep learning model achieved an area under the curve of 0.809 in the external validation cohort, outperforming the clinical model. Interpretability analysis identified the D-score, rim enhancement, and diameter of the largest axillary lymph node as the most influential predictors, consistent with established clinical knowledge.ConclusionThe proposed model enables accurate, non-invasive, and interpretable prediction of HER-2 expression in breast cancer. It may serve as a preoperative stratification tool, support individualized treatment planning, and reduce reliance on invasive diagnostic procedures.

BackgroundInterreader variability in ureteral stone size measurements affect the predicted probability of spontaneous stone passage (SSP), especially in proximal ureteral stones. Window settings have been shown to influence interreader variability.PurposeTo investigate interreader variability of ureteral stone size measurements in four different window settings.Material and MethodsPatients with a unilateral proximal ureteral stone ≥2.0 mm detected during emergency computed tomography (CT) were included in this single-center study. Five observers measured each stone in three dimensions in a soft-tissue window, bone window, and two half-value windows (based on the mean [half-value MEAN] or maximum attenuation of the stone [half-value MAX]). Limits of agreement of the mean (LOAM) for stone size in each window setting were assessed. Logistic regression curves were created for predicted probability of SSP.ResultsIn total, 124 patients (87 men, 37 women; mean age = 52 years; age range = 22-82 years) were retrospectively evaluated. LOAM: bone window (±1.6 mm, 95% confidence interval [CI]=1.24-4.90), soft-tissue window (±0.4 mm, 95% CI=0.37-0.82), half-value MEAN window (±0.3 mm, 95% CI=0.24-0.40), half-value MAX window (±0.2 mm, 95% CI=0.14-0.30). Prediction curves aligned and shifted to the left as mean stone size decreased in the half-value window settings.ConclusionThe bone window is unsatisfactory for ureteral stone size measurements. The interreader variability in soft-tissue and half-value windows is on a sub-mm magnitude, with no expected impact on clinical decision-making. The half-value MAX window had the smallest interreader variability and should be considered for reproducible and semiautomated ureteral stone size measurements.

BackgroundMeningiomas are the most common extra-axial tumors of the central nervous system, and accurate preoperative assessment of their histological grade is essential for effective treatment planning.PurposeTo investigate the relationship between the apparent diffusion coefficient (ADC) sequence, histopathological grade, and Ki-67 proliferation index for radiologically identifying meningiomas with poor prognosis.Material and MethodsThe study included 90 patients with histopathologically confirmed meningioma between March 2019 and February 2021. The Ki-67 proliferation index was assessed using an image analysis program. Retrospectively, ADC maps and diffusion-weighted imaging (DWI) were reviewed. An oval-shaped region of interest was placed over the lesion's solid component and the normal-appearing white matter in the opposite hemisphere. Each patient's ADC ratio (ADC meningioma/ADC normal-appearing white matter) was calculated. The relationship between ADC and Ki-67 proliferation index was investigated, and ADC values of benign and atypical meningiomas were compared. Independent sample t-test, Mann-Whitney U test, and receiver operating characteristic were used for statistical assessment.ResultsThe mean ADC value was 844.11 ± 123.55 mm²/s for low-grade and 743.75 ± 92.64 mm²/s for high-grade meningiomas. The mean ADC ratio was 1.11 ± 0.19 for low-grade and 1.00 ± 0.15 for high-grade meningiomas. Both ADC values and ADC ratio significantly distinguished histopathologic grades (P = 0.003, P = 0.030, respectively). No significant correlation was found between ADC values or ADC ratio and the Ki-67 proliferation index (r = -0.123, P = 0.248; r = 0.033, P = 0.755).ConclusionA statistically significant difference was found between ADC values and ADC ratio of low- and high-grade meningiomas. There was no correlation between either ADC values or ADC ratio and Ki-67 proliferation index.

BackgroundTo study the effects of direct oral anticoagulant (DOAC) treatment in patients with acute pulmonary embolism (PE), it is important to analyze iodine density perfusion maps by dual-layer spectral detector computed tomography (DLCT).PurposeTo investigate whether the total lung iodine value (TLIV) obtained from CT pulmonary angiography (CTPA) using DLCT provides valuable insights for assessing treatment response in acute PE.Material and MethodsWe conducted a retrospective study enrolling individuals receiving DOAC therapy for acute PE. Using DLCT, lung CT imaging before contrast was performed, followed by two contrast phases (when the region of interest in the pulmonary artery exceeded 150 HU [pulmonary arterial phase (PAP)] and 60 s after the start of contrast administration). Changes in TLIV and TLIV/LV were assessed from pre-treatment to post-DOAC treatment in both greater clot resolution group (group 1) and lesser clot resolution group (group 2). In addition, a comparison of the iodine level ratio between PAP and 60 s (60s/PAP ratio) for TLIV and TLIV/LV before and after treatment was performed between the two groups.ResultsIn total, 24 patients with acute PE were analyzed using DLCT before and after DOAC therapy. The TLIV (60s/PAP ratio; median 1.34, interquartile range [IQR]=1.18-1.72) of group 1 (n = 16) was significantly higher (P = 0.002) than the TLIV (60s/PAP ratio; median = 0.91, IQR = 0.79-0.99) of group 2 (n = 8).ConclusionMeasuring the iodine maps of all lungs showed promise as the level of lung perfusion after DOAC treatment appeared to reflect the treatment effect in acute PE.

BackgroundTreatment response assessment can be challenging in bone-metastasizing breast cancer. Whole-body magnetic resonance imaging (WB-MRI) is superior to contrast-enhanced computed tomography (CE-CT) for detecting progressive disease (PD) in bone-only metastatic breast cancer; however, the ability to detect PD in mixed visceral- and bone-metastasizing breast cancer remains to be assessed in prospective studies.PurposeTo prospectively compare WB-MRI and CE-CT for detecting PD in mixed visceral- and bone-metastasizing breast cancer and determine if WB-MRI could detect PD earlier than CE-CT.Material and MethodsA total of 43 women with mixed visceral- and bone-metastasizing breast cancer were prospectively followed with WB-MRI and CE-CT from the initiation of new chemotherapy. Kaplan-Meier analysis of time to progression was performed for both WB-MRI and CE-CT, and the time difference between the two was compared using a one-sample t-test. The ability to detect PD earliest was compared using McNemar's test.ResultsOf the 43 patients enrolled, 21 completed the study. Treatment was changed due to PD in 18 patients and due to side effects in three patients. Median time until change of treatment was 8 months (range=4-21 months). WB-MRI correctly detected PD in 16/18 patients compared to 13/18 for CE-CT. A total of 73 pairs of examinations were analyzed, of which WB-MRI detected PD earliest in 7/18 patients compared to 1/18 for CE-CT (P = 0.035).. There were no statistically significant differences in time to progression between WB-MRI and CE-CT (P = 0.41).ConclusionWB-MRI can detect PD in women with mixed visceral- and bone-metastasizing breast cancer earlier than CE-CT.

BackgroundOptimizing operational protocols in medical imaging is essential to ensure the quality of radiological diagnoses. However, a quantitative method for evaluating the image quality of actual patients and detectability of lesions within these clinical images has not yet been established.PurposeTo quantitatively assess the difficulty in detecting nodules on chest radiographs using a pixel value (PV)-based receiver operating characteristic (ROC) analysis approach.Material and MethodsA chest radiograph database from the Japanese Society of Radiological Technology-containing lung nodule images classified into five levels of detection difficulty-was used for analysis. Multiple regions of interest (ROIs) were defined to encompass both nodules and surrounding anatomical structures. The mean PV and standard deviation values were calculated for each region. Assuming normal PV distributions for both nodules and backgrounds, the PV-based area under the ROC curve (AUC) was computed using a theoretical formula. The method's validity was verified by analyzing correlations with the subtlety classification, which reflects detection difficulty.ResultsAnalysis of 154 nodule images demonstrated a strong correlation with nodule subtlety (r = 0.998), and with observer-derived AUC values (r = 0.955), confirming the effectiveness of the proposed metric.ConclusionThe proposed method enables quantitative evaluation of lesion detectability in clinical images. This novel index may offer valuable clinical feedback for optimizing imaging conditions and can serve as a practical tool for training in diagnostic radiology.

BackgroundHepatic artery infusion chemotherapy (HAIC) combined with tyrosine kinase inhibitors (TKIs) shows promise for treating advanced hepatocellular carcinoma (HCC).PurposeTo compare the efficacy and safety of HAIC versus transarterial chemoembolization (TACE), each combined with TKIs, in patients with advanced HCC.Material and MethodsThis retrospective study analyzed 86 patients with unresectable HCC, predominantly Barcelona Clinic Liver Cancer (BCLC) stage B/C and with well-preserved liver function, treated at a tertiary medical center between January 2019 and December 2022. Patients were assigned to either the HAIC + TKI group (n = 40) or the TACE + TKI group (n = 46) based on multidisciplinary team recommendations, considering tumor burden, vascular invasion, and patient preference. Overall survival (OS) was evaluated using Kaplan-Meier analysis, and treatment-related adverse events (AEs) were recorded.ResultsThe median OS was 13.7 months in the HAIC group and 8.3 months in the TACE group (hazard ratio = 0.5489, 95% confidence interval = 0.3557-0.8472; P = 0.00661). The most frequent grade 3/4 AEs were elevated ALT (15.0% vs. 39.1%), elevated AST (17.5% vs. 34.8%), and abdominal pain (25.0% vs. 28.3%), with only ALT showing significant between-group difference. All AEs were manageable, with no treatment-related deaths.ConclusionIn a cohort of patients with unresectable HCC, primarily BCLC stage B/C, and preserved liver function, HAIC combined with TKIs offers superior survival and tumor control compared to TACE + TKIs in unresectable HCC, with acceptable safety. Prospective multicenter trials are needed to validate these findings and further optimize treatment strategies.