Acta radiologica最新文献

BackgroundManual data curation was necessary to extract radiology reports due to the ambiguities of natural language.PurposeTo develop a fine-tuned large language model that classifies computed tomography (CT)-guided interventional radiology reports into technique categories and to compare its performance with that of the readers.Material and MethodsThis retrospective study included patients who underwent CT-guided interventional radiology between August 2008 and November 2024. Patients were chronologically assigned to the training (n = 1142; 646 men; mean age = 64.1 ± 15.7 years), validation (n = 131; 83 men; mean age = 66.1 ± 16.1 years), and test (n = 332; 196 men; mean age = 66.1 ± 14.8 years) datasets. In establishing a reference standard, reports were manually classified into categories 1 (drainage), 2 (lesion biopsy within fat or soft tissue density tissues), 3 (lung biopsy), and 4 (bone biopsy). The bi-directional encoder representation from the transformers model was fine-tuned with the training dataset, and the model with the best performance in the validation dataset was selected. The performance and required time for classification in the test dataset were compared between the best-performing model and the two readers.ResultsCategories 1/2/3/4 included 309/367/270/196, 30/42/40/19, and 75/124/78/55 patients for the training, validation, and test datasets, respectively. The model demonstrated an accuracy of 0.979 in the test dataset, which was significantly better than that of the readers (0.922-0.940) (P ≤0.012). The model classified reports within a 49.8-53.5-fold shorter time compared to readers.ConclusionThe fine-tuned large language model classified CT-guided interventional radiology reports into four categories demonstrating high accuracy within a remarkably short time.

BackgroundThe use of artificial intelligence (AI) in screen-reading of mammograms has shown promising results for cancer detection. However, less attention has been paid to the false positives generated by AI.PurposeTo investigate mammographic features in screening mammograms with high AI scores but a true-negative screening result.Material and MethodsIn this retrospective study, 54,662 screening examinations from BreastScreen Norway 2010-2022 were analyzed with a commercially available AI system (Transpara v. 2.0.0). An AI score of 1-10 indicated the suspiciousness of malignancy. We selected examinations with an AI score of 10, with a true-negative screening result, followed by two consecutive true-negative screening examinations. Of the 2,124 examinations matching these criteria, 382 random examinations underwent blinded consensus review by three experienced breast radiologists. The examinations were classified according to mammographic features, radiologist interpretation score (1-5), and mammographic breast density (BI-RADS 5th ed. a-d).ResultsThe reviews classified 91.1% (348/382) of the examinations as negative (interpretation score 1). All examinations (26/26) categorized as BI-RADS d were given an interpretation score of 1. Classification of mammographic features: asymmetry = 30.6% (117/382); calcifications = 30.1% (115/382); asymmetry with calcifications = 29.3% (112/382); mass = 8.9% (34/382); distortion = 0.8% (3/382); spiculated mass = 0.3% (1/382). For examinations with calcifications, 79.1% (91/115) were classified with benign morphology.ConclusionThe majority of false-positive screening examinations generated by AI were classified as non-suspicious in a retrospective blinded consensus review and would likely not have been recalled for further assessment in a real screening setting using AI as a decision support.

BackgroundLimited data exist comparing the detection of microscopic fat in adrenal adenomas on two-dimensional chemical shift dual-echo (2D CSI) magnetic resonance imaging (MRI) and three-dimensional two-point Dixon techniques (3D Dixon).PurposeTo compare the sensitivity of 2D CSI versus 3D Dixon techniques for the diagnosis of adrenal adenomas.Material and MethodsA retrospective analysis was conducted of 33 patients with adrenal masses who underwent both 2D CSI and 3D Dixon sequences on a 1.5-T scanner. Two blinded radiologists measured and calculated signal intensity (SI) index (SII) (100×(SI in phase - SI out of phase)/SI in phase) of nodules on each technique. Reference standard diagnosis of 30 adrenal adenomas was established. Sensitivity for adrenal adenoma diagnosis was determined using a SII >16.5%.ResultsIn total, 33 nodules were investigated (mean size=22 mm, range=11-55 mm). Of the 30 adenomas, the mean SII on 2D CSI was 48% for reader 1 and 44% for reader 2, compared to 34% on 3D Dixon for both readers (P < 0.001). Sensitivity for the diagnosis of adenoma with 2D CSI was 90% (95% confidence interval [CI]=82-98) for both readers, while 3D Dixon demonstrated a sensitivity of 73% (95% CI=65-82) for reader 1 and 63% (95% CI=55-72) for reader 2.Conclusion2D dual gradient-echo CSI demonstrated a higher sensitivity for the diagnosis of adrenal adenoma than the 3D Dixon technique. Adrenal MRI evaluation of the adrenal glands at 1.5 T should include 2D dual gradient-echo CSI and not rely solely on 3D two-point Dixon techniques for the diagnosis of adrenal adenomas.

BackgroundA timely assessment of local recurrence (LoR) risk in extremity high-grade osteosarcoma is crucial for optimizing treatment strategies and improving patient outcomes.PurposeTo explore the potential of machine-learning algorithms in predicting LoR in patients with osteosarcoma.Material and MethodsData from patients with high-grade osteosarcoma who underwent preoperative radiograph and multiparametric magnetic resonance imaging (MRI) were collected. Machine-learning models were developed and trained on this dataset to predict LoR. The study involved selecting relevant features, training the models, and evaluating their performance using the receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC). DeLong's test was utilized for comparing the AUCs.ResultsThe performance (AUC, sensitivity, specificity, and accuracy) of four classifiers (random forest [RF], support vector machine, logistic regression, and extreme gradient boosting) using radiograph-MRI as image inputs were stable (all Hosmer-Lemeshow index >0.05) with the fair to good prognosis efficacy. The RF classifier using radiograph-MRI features as training inputs exhibited better performance (AUC = 0.806, 0.868) than that using MRI only (AUC = 0.774, 0.771) and radiograph only (AUC = 0.613 and 0.627) in the training and testing sets (P <0.05) while the other three classifiers showed no difference between MRI-only and radiograph-MRI models.ConclusionThis study provides valuable insights into the use of machine learning for predicting LoR in osteosarcoma patients. These findings emphasize the potential of integrating radiomics data with algorithms to improve prognostic assessments.

BackgroundArtificial intelligence (AI)-based systems have the potential to increase the efficiency and effectiveness of breast cancer screening programs but need to be carefully validated before clinical implementation.PurposeTo retrospectively evaluate an AI system to safely reduce the workload of a double-reading breast cancer screening program.Material and MethodsAll digital mammography (DM) screening examinations of women aged 40-74 years between August 2021 and January 2022 in Östergötland, Sweden were included. Analysis of the interval cancers (ICs) was performed in 2024. Each examination was double-read by two breast radiologists and processed by the AI system, which assigned a score of 1-10 to each examination based on increasing likelihood of cancer. In a retrospective simulation, the AI system was used for triaging; low-risk examinations (score 1-7) were selected for single reading and high-risk examinations (score 8-10) for double reading.ResultsA total of 15,468 DMs were included. Using an AI triaging strategy, 10,473 (67.7%) examinations received scores of 1-7, resulting in a 34% workload reduction. Overall, 52/53 screen-detected cancers were assigned a score of 8-10 by the AI system. One cancer was missed by the AI system (score 4) but was detected by the radiologists. In total, 11 cases of IC were found in the 2024 analysis.ConclusionReplacing one reader in breast cancer screening with an AI system for low-risk cases could safely reduce workload by 34%. In total, 11 cases of IC were found in the 2024 analysis; of them, three were identified correctly by the AI system at the 2021-2022 examination.

BackgroundComputed tomography (CT)-derived acetabular angles are commonly used in the diagnosis of hip dysplasia, but the measurements are labor-intensive, with higher inter- and intra-operator variation, necessitating an automated method.PurposeTo develop and validate an automatic method for segmenting the acetabular lunate surface and measure diagnostic angles using CT images to improve diagnosis and preoperative planning for patients with hip dysplasia.Material and MethodsWe developed a method to segment the acetabular lunate surface, automatically identify five landmark points (center, anterior, posterior, lateral, and medial) and calculate diagnostic angles for center-edge (CE), anterior-sector (AASA), posterior-sector (PASA), acetabular anteversion (AcAV), and acetabular-index (AI). The method was validated against repeated manual measurements by three raters on a dataset of 18 patients (36 hips).ResultsNo differences between raters and the automatic method for the center (P = 0.18), anterior (P = 0.55), posterior (P = 0.18), lateral (P = 0.13), and medial (P = 0.12) landmarks. No statistically significant differences were observed between raters and the automatic method for the AASA (P = 0.01) and PASA (P = 0.08) angles. Statistically significant differences were found between the automatic method and rater 3 for the CE and AI angles, and between the automatic method and rater 2 for the AcAV angle. The ICC for all angle measurements by raters and the automated method was in the range of 0.90-0.99.ConclusionWith similar agreement between manual and automatic measurements, the automatic method provides important information that may be used for both diagnosis and surgical planning, with the potential to greatly reduce the time used for analysis per patient.

BackgroundThis study examines the correlation between magnetic resonance imaging (MRI)-derived volume parameters, surgical outcomes, and renal function in adults undergoing ureteropelvic junction (UPJ) obstruction surgery. Understanding these relationships can improve surgical planning, patient selection, and postoperative prognosis.PurposeTo assess the correlation between anatomical parameters and surgical outcomes in adult patients with UPJ obstruction.Material and MethodsA prospective cross-sectional study was conducted on 60 patients with UPJ obstruction, selected via simple random sampling. The hydronephrosis volume (HV) to renal volume (RV) ratio was calculated using MR urography (MRU). Preoperative diethylene triamine pentaacetic acid (DTPA) differential renal function (DRF) and creatinine levels were also recorded. Patients requiring surgery were followed up after 6 months, measuring pelvis/RV ratio, creatinine, and DTPA DRF. Statistical analyses were performed to find correlations.ResultsOperated patients had a higher preoperative HV/RV ratio (AUC=0.914, 95% confidence interval [CI]=0.829-1.000; P <0.001) and higher DTPA DRF values (AUC=0.936, 95% CI=0.860-1.000; P <0.001). Patients with greater preoperative HV/RV ratios were less likely to achieve anatomical normalization. Significant correlations were found between HV/RV ratios with DTPA DRF and creatinine (P <0.05). DeLong's test showed no significant differences between HV/RV ratios and DTPA DRF in predicting surgical need.ConclusionQuantitative volumetric analysis using MRU can effectively predict the need for surgery and renal function deterioration in patients with UPJ obstruction. The HV/RV ratio plays a crucial role in guiding surgical decisions and predicting outcomes. This study emphasizes and tests the hypothesis that higher degree of hydronephrosis correlates with higher degree of deterioration of renal function and need for surgical intervention.

BackgroundEmbolization plays a significant role in interventional radiology and modern medicine, intersecting with several specialties. The technological advancement of embolic agents has contributed to successful outcomes in the treatment of a wide range of diseases.PurposeTo compare the histomorphological effects of Embosoft microspheres and Embosphere microspheres in uterine artery embolization (UAE) in sheep.Material and MethodsSuperselective and bilateral UAE was performed with Embosoft and Embosphere microspheres in 10 adult non-pregnant sheep. Embosoft microspheres with a diameter of 500-700 μm were compared with Embosphere microspheres of a similar diameter in two groups of five sheep each. One sheep was embolized only with non-ionic iodinated contrast solution and saline (SF0.9%). The evaluation was based on histopathological examination of the uterus, performed 7 days after embolization. Necrosis scores, the diameter of occluded arteries, and the number of particles were assessed. ANOVA test and Student's t-test were used to determine the differences between the study groups.ResultsThe mean volume of embolic agents was 18.35 mL in the Embosphere group and 19.1 mL in the Embosoft group, with no statistically significant difference (P = 0.62). No significant difference was observed between the corresponding sides in the two groups. In addition, there were no significant differences in the thickness of the surface epithelium (Embosphere 21.26 μm vs. 19.72 μm Embosoft; P = 0.56) and glandular area between the groups (Embosphere 12.20% vs. 17.77% Embosoft; P = 0.18).ConclusionEmbosoft micropheres were associated with a greater inflammatory response and a smaller area of degeneration compared to Embosphere microspheres.

BackgroundHigher-resolution magnetic resonance imaging sequences are needed for the early detection of pancreatic cancer.PurposeTo compare the quality of our novel T2-weighted, high-contrast, thin-slice imaging sequence, with an improved spatial resolution and deep learning-based reconstruction (three-shot turbo spin-echo with deep learning-based reconstruction [3S-TSE-DLR]), for imaging the pancreas with imaging using three conventional sequences (half-Fourier acquisition single-shot turbo spin-echo [HASTE], fat-suppressed 3D T1-weighted [FS-3D-T1W] imaging, and magnetic resonance cholangiopancreatography [MRCP]).Material and MethodsPancreatic images of 50 healthy volunteers acquired with 3S-TSE-DLR, HASTE, FS-3D-T1W imaging, and MRCP were compared by two diagnostic radiologists. A 5-point scale was used for assessing motion artifacts, pancreatic margin sharpness, and the ability to identify the main pancreatic duct (MPD) on 3S-TSE-DLR, HASTE, and FS-3D-T1W imaging, respectively. The ability to identify MPD via MRCP was also evaluated.ResultsArtifact scores (the higher the score, the fewer the artifacts) were significantly higher for 3S-TSE-DLR than for HASTE, and significantly lower for 3S-TSE-DLR than for FS-3D-T1W imaging, for both radiologists. Sharpness scores were significantly higher for 3S-TSE-DLR than for HASTE and FS-3D-T1W imaging, for both radiologists. The rate of identification of MPD was significantly higher for 3S-TSE-DLR than for FS-3D-T1W imaging, for both radiologists, and significantly higher for 3S-TSE-DLR than for HASTE for one radiologist. The rate of identification of MPD was not significantly different between 3S-TSE-DLR and MRCP.Conclusion3S-TSE-DLR provides better image sharpness than conventional sequences, can identify MPD equally as well or better than HASTE, and shows identification performance comparable to that of MRCP.

BackgroundRadiofrequency ablation (RFA) is a first-line therapy for early-stage, single, small (≤3 cm) hepatocellular carcinoma (HCC) tumors; however, adequate control of subcapsular HCC by RFA remains challenging due to the higher risk of major complications and local tumor recurrence than non-subcapsular HCC.PurposeTo compare safety and efficacy of conventional transarterial chemoembolization (cTACE) and RFA as treatments for single, small (≤3 cm) HCC with a subcapsular location.Material and MethodsBetween 2008 and 2017, 717 treatment-naïve patients who underwent cTACE (n = 362) or RFA (n = 355) as a first-line treatment for single, small (≤3 cm), subcapsular HCC were enrolled. Propensity score analysis using inverse probability weighting (IPW) was applied to reduce the effect of potential confounding factors.ResultsThe median follow-up time was 87 months. After propensity score analysis using IPW, the 15-year overall survival rates in the cTACE and RFA groups were 47% and 45%, respectively (P = 0.89). The 15-year time to local tumor recurrence rates were 55% and 71%, respectively (P <0.001), and the 15-year time to recurrence rates were 29% and 30%, respectively (P = 0.18). The rates of major complication associated with cTACE and RFA after IPW were 1% and 4%, respectively (P = 0.01).ConclusioncTACE is a viable alternative to RFA for treating subcapsular HCCs measuring ≤3 cm, with a comparable overall survival rate and fewer major complications.