Background: The ultrasound is one of the most used medical imaging investigations worldwide. It is non-invasive and effective in assessing liver tumors or other types of parenchymal changes. Methods: The aim of the study was to build a deep learning model for image segmentation in ultrasound video investigations. The dataset used in the study was provided by the University of Medicine and Pharmacy Craiova, Romania and contained 50 video examinations from 49 patients. The mean age of the patients in the cohort was 69.57. Regarding presence of a subjacent liver disease, 36.73% had liver cirrhosis and 16.32% had chronic viral hepatitis (5 patients: chronic hepatitis C and 3 patients: chronic hepatitis B). Frames were extracted and cropped from each examination and an expert gastroenterologist labelled the lesions in each frame. After labelling, the labels were exported as binary images. A deep learning segmentation model (U-Net) was trained with focal Tversky loss as a loss function. Two models were obtained with two different sets of parameters for the loss function. The performance metrics observed were intersection over union and recall and precision. Results: Analyzing the intersection over union metric, the first segmentation model obtained performed better compared to the second model: 0.8392 (model 1) vs. 0.7990 (model 2). The inference time for both models was between 32.15 milliseconds and 77.59 milliseconds. Conclusions: Two segmentation models were obtained in the study. The models performed similarly during training and validation. However, one model was trained to focus on hard-to-predict labels. The proposed segmentation models can represent a first step in automatically extracting time-intensity curves from CEUS examinations.
Hepatic hemangioma is usually detected on a routine ultrasound examination because of silent clinical behaviour. The typical ultrasound appearance of hemangioma is easily recognizable and quickly guides the diagnosis without the need for further investigation. But there is also an entire spectrum of atypical and uncommon ultrasound features and our review comes to detail these particular aspects. An atypical aspect in standard ultrasound leads to the continuation of explorations with an imaging investigation with contrast substance [ultrasound/ computed tomography/or magnetic resonance imaging (MRI)]. For a clinician who practices ultrasound and has an ultrasound system in the room, the easiest, fastest, non-invasive and cost-effective method is contrast enhanced ultrasound (CEUS). Approximately 85% of patients are correctly diagnosed with this method and the patient has the correct diagnosis in about 30 min without fear of malignancy and without waiting for a computer tomography (CT)/MRI appointment. In less than 15% of patients CEUS does not provide a conclusive appearance; thus, CT scan or MRI becomes mandatory and liver biopsy is rarely required. The aim of this updated review is to synthesize the typical and atypical ultrasound aspects of hepatic hemangioma in the adult patient and to propose a fast, non-invasive and cost-effective clinical-ultrasound algorithm for the diagnosis of hepatic hemangioma.
Clinical utility of ancillary features (AFs) in contrast-enhanced ultrasound (CEUS) Liver Imaging Reporting and Data System (LI-RADS®) is yet to be established. In this study, we assessed the diagnostic yield of CEUS LI-RADS and AFs in hepatocellular carcinoma (HCC). We retrospectively included patients with risk factors for HCC and newly diagnosed focal liver lesions (FLL). All lesions have been categorized according to the CEUS LI-RADS v2017 by an experienced sonographer blinded to clinical data and to the final diagnosis. From a total of 143 patients with 191 FLL, AFs favoring HCC were observed in 19.8% cases as hypoechoic rim and in 16.7% cases as nodule-in nodule architecture. From the total of 141 HCC cases, 83.6% were correctly classified: 57.4%- LR-5 and 26.2%- LR-4. In 9.21% cases, CEUS indicated LR-M; 2.12% cases- LR-3. The LR-5 category was 96.2% predictive (PPV) of HCC. LR-5 had 60.4% sensitivity and 93.6% specificity. PPV for primitive malignancy (LR-4 + LR-5) was 95.7%, with 88% sensitivity, 89.3% specificity and 88.4% accuracy for HCC. LR-4 category had 94.8% PPV and 26.2% sensitivity. CEUS LR4 + LR5 had 81,8% sensitivity for HCCs over 2 cm and 78.57% sensitivity for smaller HCCs. CEUS LR-5 remains an excellent diagnostic tool for HCC, despite the size of the lesion. The use of AFs might improve the overarching goal of LR-5 + LR-4 diagnosis of high specificity for HCC and exclusion of non-HCC malignancy.
Background: Contrast-enhanced ultrasound (CEUS) is an important imaging modality in the diagnosis of liver tumors. By using contrast agent, a more detailed image is obtained. Time-intensity curves (TIC) can be extracted using a specialized software, and then the signal can be analyzed for further investigations. Methods: The purpose of the study was to build an automated method for extracting TICs and classifying liver lesions in CEUS liver investigations. The cohort contained 50 anonymized video investigations from 49 patients. Besides the CEUS investigations, clinical data from the patients were provided. A method comprising three modules was proposed. The first module, a lesion segmentation deep learning (DL) model, handled the prediction of masks frame-by-frame (region of interest). The second module performed dilation on the mask, and after applying colormap to the image, it extracted the TIC and the parameters from the TIC (area under the curve, time to peak, mean transit time, and maximum intensity). The third module, a feed-forward neural network, predicted the final diagnosis. It was trained on the TIC parameters extracted by the second model, together with other data: gender, age, hepatitis history, and cirrhosis history. Results: For the feed-forward classifier, five classes were chosen: hepatocarcinoma, metastasis, other malignant lesions, hemangioma, and other benign lesions. Being a multiclass classifier, appropriate performance metrics were observed: categorical accuracy, F1 micro, F1 macro, and Matthews correlation coefficient. The results showed that due to class imbalance, in some cases, the classifier was not able to predict with high accuracy a specific lesion from the minority classes. However, on the majority classes, the classifier can predict the lesion type with high accuracy. Conclusions: The main goal of the study was to develop an automated method of classifying liver lesions in CEUS video investigations. Being modular, the system can be a useful tool for gastroenterologists or medical students: either as a second opinion system or a tool to automatically extract TICs.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.