1 March 2022 in Radiology
Quantitative US methods for liver fat content are promising but analysis of optimal acquisition protocols and confounding factors is incomplete. Nonalcoholic fatty liver disease (NAFLD) is the most common chronic liver disease worldwide, with an estimated prevalence of up to 30% in the general population and higher in people with type 2 dia...
23 May 2019 in Journal of Ultrasound in Medicine
1 September 2021 in Academic Radiology
18 April 2018 in Biomedical Physics & Engineering Express
Quantitative 2D and 3D contrast-enhanced ultrasound (CEUS) was assessed to evaluate early transarterial chemoembolization (TACE) treatment response. Seventeen patients scheduled for TACE for the treatment of hepatocellular carcinoma participated in the study. 2D and 3D CEUS were performed for each patient at three time points: prior to TACE, 1–2 weeks post TACE, and 1 month post TACE. Peak-intensities of the tumor and surrounding liver tissue were calculated from 2D and 3D data before and after TACE and used to evaluate tumor treatment response. Residual tumor percentages were calculated from 2D and 3D CEUS acquired 1–2 weeks and 1 month post TACE and compared with results from MRI 1 month post TACE. Nine subjects had complete response while 8 had incomplete response. Peak-intensities of the tumor from 3D CEUS prior to TACE were similar between the complete and incomplete treatment groups (p = 0.70), while 1–2 weeks (p < 0.01) and 1 month post treatment (p < 0.01) were significantly lower in the complete treatment group than in the incomplete treatment group. For 2D CEUS, only the peak-intensity values of the tumor from 1 month post TACE were significantly different (p < 0.01). The correlation coefficients between 2D and 3D residual tumor estimates 1–2 weeks post TACE and the estimates from MRI were 0.73 and 0.94, respectively, while those from 2D and 3D CEUS 1 month post TACE were 0.66 and 0.91, respectively. Quantitative analysis on 2D and 3D CEUS shows potential to differentiate patients with complete versus incomplete response to TACE as early as 1–2 weeks post treatment.
23 October 2020 in Journal of Medical Imaging
<!-- *** Custom HTML *** --> Purpose: While mammography has excellent sensitivity for the detection of breast lesions, its specificity is limited. Adjunct screening with ultrasound may partially alleviate this issue but also increases false positives, resulting in unnecessary biopsies. Our study investigated the use of Google AutoML Vision (Mountain View, California), a commercially available machine learning service, to both identify and characterize indeterminate breast lesions on ultrasound. Approach: B-mode images from 253 independent cases of indeterminate breast lesions scheduled for core biopsy were used for model creation and validation. The performances of two sub-models from AutoML Vision, the image classification model and object detection model, were evaluated, while also investigating training strategies to enhance model performances. Pathology from the patient’s biopsy was used as a reference standard. Results: The image classification models trained under different conditions demonstrated areas under the precision–recall curve (AUC) ranging from 0.85 to 0.96 during internal validation. Once deployed, the model with highest internal performance demonstrated a sensitivity of 100% [95% confidence interval (CI) of 73.5% to 100%], specificity of 83.3% (CI = 51.6 % to 97.9%), positive predictive value (PPV) of 85.7% (CI = 62.9 % to 95.5%), and negative predictive value (NPV) of 100% (CI non-evaluable) in an independent dataset. The object detection model demonstrated lower performance internally during development (AUC = 0.67) and during prediction in the independent dataset [sensitivity = 75 % (CI = 42.8 to 94.5), specificity = 80 % (CI = 51.9 to 95.7), PPV = 75 % (CI = 50.8 to 90.0), and NPV = 80 % (CI = 59.3 % to 91.7%)], but was able to demonstrate the location of the lesion within the image. Conclusions: Two models appear to be useful tools for identifying and classifying suspicious areas on B-mode images of indeterminate breast lesions.
14 February 2023 in RMD Open
Objective: To assess the finger vascularity of systemic sclerosis patients with Raynaud’s phenomenon (RP-SSc) using various ultrasound techniques. Methods: All fingers (except thumbs) of 18 RP-SSc patients and 18 controls were imaged at room temperature using four ultrasound vascular imaging techniques. The percent vascular area was quantified by counting blood flow pixels in a 25 mm2 square centred at the nail fold for the dorsal side and in 25 mm2 and 100 mm2 square from the fingertip for the ventral side. The mean vascular intensity was calculated from the corresponding areas for dorsal and ventral sides. Results: The percent vascular areas and mean vascular intensities in RP-SSc were significantly lower than those in controls for both dorsal and ventral sides (p<0.01). The mean vascular intensities showed slightly higher area under the curve (AUC) than the percent vascular areas (0.53–0.91 vs 0.53–0.90) regardless of imaging technique and assessment side. For each imaging technique, the ventral side vascularity showed a higher AUC (0.74–0.91) compared with the dorsal side (0.53–0.81). Moreover, ventral side abnormalities were associated with a history of digital ulcers. Conclusions: Ultrasound demonstrated potential to quantify finger vascularity of RP-SSc. The ventral side of the fingers showed a higher accuracy in detecting RP-SSc than the dorsal side.