Search Results (10)

The term “Liver disease” refers to a broad category of disorders affecting the liver. There are a variety of common liver ailments, such as hepatitis, cirrhosis, and liver cancer. Accurate and early diagnosis is an emergent demand for the prediction and diagnosis of liver disease. Conventional diagnostic techniques, such as radiological, CT scan, and liver function tests, are often time-consuming and prone to inaccuracies in several cases. An application of machine learning (ML) and deep learning (DL) techniques is an efficient approach to diagnosing diseases in a wide range of medical fields. This type of machine-related learning can handle various tasks, such as image recognition, analysis, and classification, because it helps train large datasets and learns to identify patterns that might not be perceived by humans. This paper is presented here with an evaluation of the performance of various DL models on the estimation and subtyping of liver ailment and prognosis. In this manuscript, we propose a novel approach, termed CNN+LSTM, which is an integration of convolutional neural network (CNN) and long short-term memory (LSTM) networks. The results of the study prove that ML and DL can be used to improve the diagnosis and prognosis of liver disease. The CNN+LSTM model achieves a better accuracy of 98.73% compared to other models such as CNN, Recurrent Neural Network (RNN), and LSTM. The incorporation of the proposed CNN+LSTM model has better results in terms of accuracy (98.73%), precision (99%), recall (98%), F1 score (98%), and AUC (Area Under the Curve)-ROC (Receiver Operating Characteristic) (99%), respectively. The use of the CNN+LSTM model shows robustness in predicting the liver ailment with an accurate diagnosis and prognosis. Full article

An automated segmentation method for computed tomography (CT) images of liver tumors is an urgent clinical need. Tumor areas within liver cancer images are easily missed as they are small and have unclear borders. To address these issues, an improved liver tumor segmentation method based on U-Net is proposed. This involves incorporating attention mechanisms into the U-Net’s skip connections, giving higher weights to important regions. Through dynamically adjusting the attention recognition characteristics, the method achieves accurate localization that is focused on and discriminates target regions. Testing using the LiTS (liver tumor segmentation) public dataset resulted in a Dice similarity coefficient of 0.69. The experiments demonstrated that this method can accurately segment liver tumors. Full article

Segmenting the liver and liver tumors in computed tomography (CT) images is an important step toward quantifiable biomarkers for a computer-aided decision-making system and precise medical diagnosis. Radiologists and specialized physicians use CT images to diagnose and classify liver organs and tumors. Because these organs have similar characteristics in form, texture, and light intensity values, other internal organs such as the heart, spleen, stomach, and kidneys confuse visual recognition of the liver and tumor division. Furthermore, visual identification of liver tumors is time-consuming, complicated, and error-prone, and incorrect diagnosis and segmentation can hurt the patient’s life. Many automatic and semi-automatic methods based on machine learning algorithms have recently been suggested for liver organ recognition and tumor segmentation. However, there are still difficulties due to poor recognition precision and speed and a lack of dependability. This paper presents a novel deep learning-based technique for segmenting liver tumors and identifying liver organs in computed tomography maps. Based on the LiTS17 database, the suggested technique comprises four Chebyshev graph convolution layers and a fully connected layer that can accurately segment the liver and liver tumors. Thus, the accuracy, Dice coefficient, mean IoU, sensitivity, precision, and recall obtained based on the proposed method according to the LiTS17 dataset are around 99.1%, 91.1%, 90.8%, 99.4%, 99.4%, and 91.2%, respectively. In addition, the effectiveness of the proposed method was evaluated in a noisy environment, and the proposed network could withstand a wide range of environmental signal-to-noise ratios (SNRs). Thus, at SNR = −4 dB, the accuracy of the proposed method for liver organ segmentation remained around 90%. The proposed model has obtained satisfactory and favorable results compared to previous research. According to the positive results, the proposed model is expected to be used to assist radiologists and specialist doctors in the near future. Full article

Transdiaphragmatic fistulae are rare conditions characterized by pathological communication between two epithelium-lined surfaces. Hepato-thoracic fistula consists of abnormal communication between the liver and/or the biliary system and the thorax; while the pancreaticopleural fistula consists of abnormal communication between the pancreas and the thorax, the pleuro-biliary fistula represents the more common type. Clinical symptoms and laboratory findings are generally non-specific (e.g., thoracic and abdominal pain, dyspnea, cough, neutrophilia, elevated CPR, and bilirubin values) and initially, first-level investigations, such as chest RX and abdominal ultrasound, are generally inconclusive for the diagnosis. Contrast-enhanced CT represents the first two-level radiological imaging technique, usually performed to identify and evaluate the underlying pathology sustained by transdiaphragmatic fistulae, their complications, and the evaluation of the fistulous tract. When the CT remains inconclusive, other techniques such as MRI and MRCP can be performed. A prompt and accurate diagnosis is crucial because the recognition of fistulae and the precise definition of the fistulous tract have a major impact on the management acquisition process. Full article

The automatic recognition of CT (Computed Tomography) images of liver cancer is important for the diagnosis and treatment of early liver cancer. However, there are problems such as single model structure and loss of pooling layer information when using a traditional convolutional neural network to recognize CT images of liver cancer. Therefore, this paper proposes an efficient method for liver CT image recognition based on the capsule network (CapsNet). Firstly, the liver CT images are preprocessed, and in the process of image denoising, the traditional non-local mean (NLM) denoising algorithm is optimized with a superpixel segmentation algorithm to better protect the information of image edges. After that, CapsNet was used for image recognition for liver CT images. The experimental results show that the average recognition rate of liver CT images reaches 92.9% when CapsNet is used, which is 5.3% higher than the traditional CNN model, indicating that CapsNet has better recognition accuracy for liver CT images. Full article

Segmentation of images is a common task within medical image analysis and a necessary component of medical image segmentation. The segmentation of the liver and liver tumors is an important but challenging stage in screening and diagnosing liver diseases. Although many automated techniques have been developed for liver and tumor segmentation; however, segmentation of the liver is still challenging due to the fuzzy & complex background of the liver position with other organs. As a result, creating a considerable automated liver and tumour division from CT scans is critical for identifying liver cancer. In this article, deeply dense-network ResU-Net architecture is implemented on CT scan using the 3D-IRCADb01 dataset. An essential feature of ResU-Net is the residual block and U-Net architecture, which extract additional information from the input data compared to the traditional U-Net network. Before being fed to the deep neural network, image pre-processing techniques are applied, including data augmentation, Hounsfield windowing unit, and histogram equalization. The ResU-Net network performance is evaluated using the dice similarity coefficient (DSC) metric. The ResU-Net system with residual connections outperformed state-of-the-art approaches for liver tumour identification, with a DSC value of 0.97% for organ recognition and 0.83% for segmentation methods. Full article

The research was based on the image recognition technology of artificial intelligence, which is expected to assist physicians in making correct decisions through deep learning. The liver dataset used in this study was derived from the open source website (LiTS) and the data provided by the Kaohsiung Chang Gung Memorial Hospital. CT images were used for organ recognition and lesion segmentation; the proposed Residual-Dense-Attention (RDA) U-Net can achieve high accuracy without the use of contrast. In this study, U-Net neural network was used to combine ResBlock in ResNet with Dense Block in DenseNet in the coder part, allowing the training to maintain the parameters while reducing the overall recognition computation time. The decoder was equipped with Attention Gates to suppress the irrelevant areas of the image while focusing on the significant features. The RDA model was used to identify and segment liver organs and lesions from CT images of the abdominal cavity, and excellent segmentation was achieved for the liver located on the left side, right side, near the heart, and near the lower abdomen with other organs. Better recognition was also achieved for large, small, and single and multiple lesions. The study was able to reduce the overall computation time by about 28% compared to other convolutions, and the accuracy of liver and lesion segmentation reached 96% and 94.8%, with IoU values of 89.5% and 87%, and AVGDIST of 0.28 and 0.80, respectively. Full article

In the recent years, the number of liver resections has seen an impressive growth. Usually, hepatic resections remain the treatment of various liver diseases, such as malignant tumors, benign tumors, hydatid disease, and abscesses. Despite technical advancements and tremendous experience in the field of liver resection of specialized centers, there are moderately high rates of postoperative morbidity and mortality, especially in high-risk and older patient populations. Although ultrasonography is usually the first-line imaging examination for postoperative complications, Computed Tomography (CT) is the imaging tool of choice in emergency settings due to its capability to assess the whole body in a few seconds and detect all possible complications. Magnetic resonance cholangiopancreatography (MRCP) is the imaging modality of choice for delineating early postoperative bile duct injuries and ischemic cholangitis that may arise in the late postoperative phase. Moreover, both MDCT and MRCP can precisely detect tumor recurrence. Consequently, radiologists should have knowledge of these surgical procedures for better comprehension of postoperative changes and recognition of the radiological features of various postoperative complications. Full article

Pancreatic neuroendocrine neoplasms (panNENs) are heterogeneous neoplasms with neuroendocrine differentiation that show peculiar clinical and histomorphological features, with variable prognosis. In recent years, advances in knowledge regarding the pathophysiology and heterogeneous clinical presentation, as well as the availability of different diagnostic procedures for panNEN diagnosis and novel therapeutic options for patient clinical management, has led to the recognition of the need for an active multidisciplinary discussion for optimal patient care. Molecular imaging with positron emission tomography/computed tomography (PET/CT) has become indispensable for the management of panNENs. Several PET radiopharmaceuticals can be used to characterize either panNEN receptor expression or metabolism. The aim of this review is to offer an overview of all the currently used radiopharmaceuticals and of the new upcoming tracers for pancreatic neuroendocrine tumors (panNETs), and their clinical impact on therapy management. [⁶⁸Ga]Ga-DOTA-peptide PET/CT (SSA-PET/CT) has high sensitivity, specificity, and accuracy and is recommended for the staging and restaging of any non-insulinoma well-differentiated panNEN cases to carry out detection of unknown primary tumor sites or early relapse and for evaluation of in vivo somatostatin receptors expression (SRE) to select patient candidates for peptide receptor radiometabolic treatment (PRRT) with ⁹⁰Y or ¹⁷⁷Lu and/or cold analogs. SSA-PET/CT also has a strong impact on clinical management, leading to a change in treatment in approximately a third of the cases. Its role for treatment response assessment is still under debate due to the lack of standardized criteria, even though some semiquantitative parameters seem to be able to predict response. [¹⁸F]FDG PET/CT generally shows low sensitivity in small growing and well-differentiated neuroendocrine tumors (NET; G1 and G2), while it is of utmost importance in the evaluation and management of high-grade NENs and also provides important prognostic information. When positive, [¹⁸F]FDG PET/CT impacts therapeutical management, indicating the need for a more aggressive treatment regime. Although FDG positivity does not exclude the patient from PRRT, several studies have demonstrated that it is certainly useful to predict response, even in this setting. The role of [¹⁸F]FDOPA for the study of panNET is limited by physiological uptake in the pancreas and is therefore not recommended. Moreover, it provides no information on SRE that has crucial clinical management relevance. Early acquisition of the abdomen and premedication with carbidopa may be useful to increase the accuracy, but further studies are needed to clarify its utility. GLP-1R agonists, such as exendin-4, are particularly useful for benign insulinoma detection, but their accuracy decreases in the case of malignant insulinomas. Being a whole-body imaging technique, exendin-PET/CT gives important preoperative information on tumor size and localization, which is fundamental for surgical planning as resection (enucleation of the lesion or partial pancreatic resection) is the only curative treatment. New upcoming tracers are under study, such as promising SSTR antagonists, which show a favorable biodistribution and higher tumor-to-background ratio that increases tumor detection, especially in the liver. [⁶⁸Ga]pentixafor, an in vivo marker of CXCR4 expression associated with the behavior of more aggressive tumors, seems to only play a limited role in detecting well-differentiated NET since there is an inverse expression of SSTR2 and CXCR4 in G1 to G3 NETs with an elevation in CXCR4 and a decrease in SSTR2 expression with increasing grade. Other tracers, such as [⁶⁸Ga]Ga-PSMA, [⁶⁸Ga]Ga-DATA-TOC, [¹⁸F]SiTATE, and [¹⁸F]AlF-OC, are also under investigation. Full article

Introduction: The aim of this study was to analyze the diagnostic potential of contrast enhanced ultrasound (CEUS) for the recognition of focal lesions of the spleen and liver in patients suffering from sarcoidosis. Material and methods: We analyzed the outcome of diagnostic imaging in a group of 21 patients treated for pulmonary sarcoidosis, searching for the systemic infiltration of the liver and/or spleen. All the participants are patients with inactive disease, who are monitored every 6 months at the Pulmonology Clinic. Apart from the check-up high-resolution computed tomography (HR-CT)—every 2 years, patients underwent an initial ultrasound examination (US) and if there was a suspicion of systemic infiltration, abdominal CT and/or magnetic resonance imaging (MRI) and CEUS were performed. Results: In 18 patients suffering from pulmonary sarcoidosis diagnostic imaging revealed no systemic infiltration. In three patients, the use of CEUS exposed the presence of lesions in the parenchymal organs. In all cases, the images from CEUS were consistent with those from CT/MRI. Conclusions: CEUS has the potential to become a reliable and safe screening tool for systemic infiltration in patients with sarcoidosis. It may also be an important method of monitoring the effects of therapy. Full article