Ultrasonography最新文献

Purpose: This study aimed to develop and evaluate a deep learning model that directly analyzes three-dimensional automated breast ultrasound videos (DL-3DABUV) to assist breast cancer diagnosis, and to examine the optimal reading mode for clinical implementation.

Methods: This retrospective study included 547 patients (285 benign, 262 malignant), who were randomly assigned to a training set (n=437) and a test set (n=110). The DL-3DABUV model, built using ResNet50 and multi-instance learning, was trained by directly analyzing videos without image selection or manual annotation. Six radiologists (three experienced and three novice) evaluated the test set under three modes: independent-reading (without DL-3DABUV), second-reading (without prior knowledge of DL-3DABUV results), and concurrent-reading (after viewing DL-3DABUV results). The diagnostic performance of DL-3DABUV, experienced radiologists, and novice radiologists was compared. Reading times across the three modes were also assessed.

Results: Compared to experienced radiologists in independent reading, DL-3DABUV showed no significant differences in area under the receiver operating characteristic curve (AUC) (0.82 vs. 0.83), sensitivity (82.1% vs. 81.6%), or specificity (81.5% vs. 88.3%) (all P>0.05). DL-3DABUV exhibited higher AUC and specificity than novice radiologists in independent-reading (0.82 vs. 0.68, P<0.001; 81.5% vs. 57.4%, P<0.001). However, novice performance reached parity with DL-3DABUV in both second-reading and concurrent-reading. No significant differences in diagnostic performance were observed between second-reading and concurrent-reading. Concurrent reading significantly reduced reading time by 33.6 seconds compared with second-reading (P<0.001).

Conclusion: DL-3DABUV achieves diagnostic performance comparable to experienced radiologists and enhances diagnostic accuracy for novices. Concurrent reading provides a more efficient workflow by reducing reading time while maintaining diagnostic performance.

Purpose: This study evaluated the clinical utility of a fully automated two-dimensional shear wave elastography (2D-SWE) method incorporating automated frame selection and region-of-interest (ROI) placement algorithms in participants with diffuse liver disease. The analysis assessed examination time and reproducibility compared with semi-automated and manual methods.

Methods: This prospective study included 40 participants who underwent liver stiffness measurements with a Samsung Medison ultrasound system using three.

Methods: fully automated (automated frame selection via a reliability indicator derived from elasticity uniformity and the reliability measurement index, with automated ROI placement); semi-automated (automated frame selection with manual ROI placement); and manual (fully operator-controlled). Examination time and inter-method (n=40), intra-observer (n=21), and inter-observer (n=15) variabilities were analyzed using analysis of variance, paired t-tests, and intraclass correlation coefficients (ICCs). All participants underwent comparative measurements with a Canon i800 system for cross-platform analysis.

Results: One technical failure each occurred during the intra-observer and cross-platform measurement sessions. Fully automated measurement significantly reduced total examination time (14.38±5.01 seconds) compared to semi-automated (30.46±7.11 seconds, P<0.001) and manual (33.05±7.10 seconds, P<0.001) measurements. No significant differences in liver stiffness were observed among methods (P=0.556). Inter-method agreement was excellent (ICC, 0.997). Intra-observer and inter-observer agreements were also excellent with the fully automated method (ICCs, 0.971 and 0.984, respectively). Good cross-platform agreement with the Canon system was observed (ICC, 0.833).

Conclusion: Fully automated 2D-SWE markedly reduces examination time while maintaining high reproducibility in liver stiffness assessment.

Purpose: This study was performed to evaluate the diagnostic performance of ultrasound-derived fat fraction (UDFF) for detecting hepatic steatosis in patients with steatotic liver disease (SLD), using liver biopsy and histopathology as the gold standard.

Methods: In this prospective study, patients referred for evaluation of suspected SLD were enrolled. All patients underwent UDFF and controlled attenuation parameter (CAP) measurements. Histopathological findings were assessed using the steatosis-activity-fibrosis scoring system as the gold standard. Pearson or Spearman correlation analyses and multivariate linear regression were performed. Areas under the receiver operating characteristic curves (AUCs) were utilized to assess the diagnostic performance of UDFF and CAP in identifying hepatic steatosis. The DeLong test was used to compare AUCs.

Results: From February 2023 to May 2025, 66 patients were included. The median body mass index was 24 kg/m2. UDFF was positively correlated with hepatic steatosis grade (r=0.66). Using UDFF thresholds of 9.5%, 14.5%, and 16.3% (determined by the Youden index), the AUCs for detecting steatosis grades ≥S1, ≥S2, and S3 were 0.82, 0.90, and 0.99, respectively. UDFF exhibited a significantly higher AUC than CAP for detecting steatosis grade S3 (P=0.037). UDFF was correlated with hepatic steatosis grade, activity grade, CAP, aspartate aminotransferase level, and albumin level (all P<0.05).

Conclusion: UDFF demonstrated good diagnostic performance for assessing hepatic steatosis, using histopathology as the reference. UDFF may serve as a quantitative imaging approach to measure liver fat content in patients with SLD.

Purpose: Hepatic steatosis, which is associated with liver diseases and adverse clinical outcomes, requires accurate, noninvasive diagnostic methods because of the limitations of liver biopsy. This study aimed to evaluate the effectiveness of the ultrasound-guided attenuation parameter (UGAP), a novel technique for assessing hepatic steatosis, and to compare its diagnostic performance with that of the controlled attenuation parameter (CAP), using magnetic resonance imaging (MRI)-derived proton density fat fraction (PDFF) as the reference standard.

Methods: A total of 255 patients with chronic liver disease who underwent CAP, UGAP, and MRI-PDFF were prospectively enrolled from three liver centers in Japan. Using the area under the receiver operating characteristic curve (AUROC) analysis, cutoff values for UGAP were determined according to steatosis grades based on MRI-PDFF. To minimize overfitting, diagnostic performance was validated using five-fold cross-validation.

Results: CAP and UGAP values followed normal distributions, whereas PDFF values deviated from normality and were therefore log-transformed, yielding the variable MRI-logPDFF. CAP and UGAP demonstrated significant correlations with MRI-logPDFF, with intraclass correlation coefficients of 0.696 and 0.797, respectively. For MRI-PDFF-based grading, the AUROC (95% confidence interval) values of CAP and UGAP were 0.878 (0.813-0.923) versus 0.926 (0.867-0.960) for S0 versus S1-S3 (P=0.041), 0.820 (0.763-0.865) versus 0.908 (0.861-0.940) for S0-S1 versus S2-S3 (P<0.001), and 0.863 (0.811-0.902) versus 0.897 (0.852-0.930) for S0-S2 versus S3 (P=0.128), respectively. The validation analysis produced results consistent with those of the primary cohort.

Conclusion: UGAP showed greater diagnostic accuracy than CAP for hepatic steatosis, notably in detecting grades ≥S1 and ≥S2.

Purpose: Barker-coded excitation, which is well suited for portable ultrasound imaging because it preserves frame rate and uses binary encoding, exhibits a high range sidelobe level of 1/N after pulse compression. This limitation can be mitigated by employing a mismatched filter, albeit at the cost of increased computational complexity. This study proposes a pulse compression technique that applies a mismatched filter for Barker-coded excitation with reduced computational complexity.

Methods: In the proposed method, pulse compression is performed using complex baseband in-phase and quadrature (IQ) data after decimation rather than beamformed radio-frequency (RF) data. By decimating both the IQ data and the coefficients of the compression filter, hardware complexity can be reduced by a factor of L² (where L is the decimation factor) through the use of time-shared multipliers. The proposed approach was implemented in a custom-built portable ultrasound imaging system, and its performance was evaluated through simulations as well as phantom and in vivo experiments.

Results: From the simulation and phantom experiments, the proposed method achieved an identical -6 dB axial resolution compared to the conventional approach, i.e., pulse compression using RF data. The range sidelobes were comparable between the conventional and proposed methods, and consistent results were also obtained in the in vivo experiment.

Conclusion: These findings demonstrate that the proposed method substantially reduces computational complexity while maintaining pulse compression performance.

Purpose: This study evaluated the performance of supplemental automated breast ultrasound (ABUS) in women with negative screening mammographic findings and assessed its utility across clinical subgroups to inform personalized implementation strategies.

Methods: We retrospectively identified 3,417 ABUS examinations performed concurrently with mammography between January 2022 and April 2024. Examinations with negative mammographic findings were included, while those with positive mammographic findings, no workup for ABUS-positive results, or less than 12 months of follow-up were excluded. The reference standards were histopathology and 12-month follow-up outcomes. ABUS performance was evaluated overall and within subgroups stratified by age (<50 vs. ≥50 years), mammographic density (non-dense vs. dense), body mass index (<25 vs. ≥25 kg/m²), and prior ultrasound history (prevalence vs. incidence).

Results: We analyzed 1,932 ABUS examinations from 1,597 women (mean age, 56±9 years). ABUS detected 13 cancers, yielding a detection rate of 6.7 per 1,000. Of these, 11 (84.6%) were invasive, including nine (81.8%) node-negative lesions, with a median size of 1.2 cm (range, 0.1 to 2.6 cm). One interval cancer was identified as a palpable mass 8 months after a negative ABUS examination. The abnormal interpretation rate, biopsy rate, sensitivity, and specificity were 28.0% (542/1,932), 4.0% (78/1,932), 92.9% (13/14), and 72.4% (1,389/1,918), respectively. Higher abnormal interpretation rates and lower specificity were observed among women aged <50 years, those with dense breasts, and during prevalence examinations. No cancers were detected in women with non-dense breasts.

Conclusion: ABUS identified small, node-negative invasive cancers with likely favorable prognoses but demonstrated limited value in women with non-dense breasts, supporting its personalized use based on breast density and patient preference.

Purpose: This study investigated whether Doppler and microvascular imaging help distinguish malignant from benign thyroid nodules in relation to the Thyroid Imaging Reporting and Data System (TI-RADS) and the pathologic subtype of thyroid malignancy.

Methods: This study included 113 consecutive thyroid nodules from 103 patients with confirmed final diagnoses from February 2022 to September 2022. Two radiologists conducted a retrospective review of ultrasonographic (US) findings and assigned vascular scores by consensus using color Doppler imaging (CDI) and microflow imaging (MFI) on a 4‑point visual scale. Vascular scores were compared between malignant and benign nodules and analyzed by US pattern according to the Korean TI‑RADS (K‑TIRADS).

Results: Of the 113 nodules, 72 were benign and 41 were malignant. All nodules were categorized using the K‑TIRADS lexicon, and each lexicon feature (composition, echogenicity, orientation, margin, and calcification) significantly differentiated malignant from benign lesions (P≤0.008). The CDI score also differed significantly (P<0.001). In subgroup analysis by K‑TIRADS pattern, differences in CDI and MFI scores were not observed for low‑ and intermediate‑suspicion nodules, whereas in high‑suspicion nodules benign lesions had higher CDI and MFI scores than malignant ones (P=0.018 and P=0.051, respectively). Classic papillary thyroid carcinoma showed lower vascular scores on both CDI and MFI, while the follicular variant showed higher scores (P=0.075 and P=0.007, respectively).

Conclusion: CDI and MFI may assist in distinguishing between malignant and benign nodules, and evaluation of nodular vascularity may be tailored to the TI‑RADS category.

Purpose: This study evaluated the 3-year safety and efficacy of microwave ablation (MWA) as an alternative to radiofrequency ablation in Korean patients with hepatocellular carcinoma (HCC).

Methods: In this prospective single-center study, 33 adults with a single HCC lesion (2-5 cm; Barcelona Clinic Liver Cancer stage A) were enrolled between December 2017 and May 2020. All underwent ultrasound-guided MWA using the NEUWAVE system. One patient did not complete post-treatment evaluation and two had protocol deviations, yielding a full analysis set of 30 patients and a safety analysis set of 33 patients. Study endpoints included technical success (TS), technique efficacy (TE), local tumor progression (LTP), overall survival (OS), and adverse events (AEs) over a 3-year follow-up period.

Results: All 33 patients were included in the safety analysis, while 30 were included in the efficacy analysis. Most patients were male (83.3%), with a median age of 64 years and a mean tumor size of 2.5±0.6 cm. Median follow-up was 36.2 months. TS and TE were achieved in all cases. The 3-year cumulative LTP rate was 7.3%, and the 3-year OS rate was 90.0%. Minor AEs occurred in 54.5% of patients, most commonly post-ablation pain and fever. Two patients (6.1%) experienced major AEs, including pyrexia and duodenal thermal injury. No procedure- or device-related deaths occurred.

Conclusion: MWA appears to be a safe and effective treatment for early-stage HCC, demonstrating durable tumor control and a low incidence of major complications over 3 years.