A fuzzy deep learning approach for liver lesions detection and classification in big data context

Improving liver lesions classification on CT/MRI images based on Hounsfield Units attenuation and deep learning

Deep learning-based diagnosis systems are useful to identify abnormalities in medical images with the greatly increased workload of doctors. Specifically, the rate of new cases and deaths from malignancies is rising for liver diseases. Early detection of liver lesions plays an extremely important role in effective treatment and gives a higher chance of survival for patients. Therefore, automatic detection and classification of common liver lesions are essential for doctors. In fact, radiologists mainly rely on Hounsfield Units to locate liver lesions but previous studies often pay little attention to this factor. In this paper, we propose an improved method for the automatic classification of common liver lesions based on deep learning techniques and the variation of Hounsfield Unit densities on CT images with and without contrast. Hounsfield Unit is used to locate liver lesions accurately and support data labeling for classification. We construct a multi-phase classification model developed on the deep neural networks of Faster R-CNN, R-FCN, SSD, and Mask R-CNN with the transfer learning approach. The experiments are conducted on six scenarios with multi-phase CT images of common liver lesions. Experimental results show that the proposed method improves the detection and classification of liver lesions compared with recent methods because its accuracy achieves up to 97.4%. The proposed models are very useful to assist doctors in the automatic segmentation and classification of liver lesions to solve the problem of depending on the clinician's experience in the diagnosis and treatment of liver lesions.

We developed a method for automated classification and detection of liver lesions in CT images based on image patch representation and bag-of-visual-words (BoVW). BoVW analysis has been extensively used in the computer vision domain to analyze scenery images. In the current work we discuss how it can be used for liver lesion classification and detection. The methodology includes building a dictionary for a training set using local descriptors and representing a region in the image using a visual word histogram. Two tasks are described: a classification task, for lesion characterization, and a detection task in which a scan window moves across the image and is determined to be normal liver tissue or a lesion. Data: In the classification task 73 CT images of liver lesions were used, 25 images having cysts, 24 having metastasis and 24 having hemangiomas. A radiologist circumscribed the lesions, creating a region of interest (ROI), in each of the images. He then provided the diagnosis, which was established either by biopsy or clinical follow-up. Thus our data set comprises 73 images and 73 ROIs. In the detection task, a radiologist drew ROIs around each liver lesion and two regions of normal liver, for a total of 159 liver lesion ROIs and 146 normal liver ROIs. The radiologist also demarcated the liver boundary. Results: Classification results of more than 95% were obtained. In the detection task, F1 results obtained is 0.76. Recall is 84%, with precision of 73%. Results show the ability to detect lesions, regardless of shape.

Detection and classification of focal liver lesions in patients with colorectal cancer: Retrospective comparison of diffusion-weighted MR imaging and multi-slice CT

The number of training data is the key bottleneck in achieving good results for medical image analysis and especially in deep learning. Due to small medical training data, deep learning models often fail to mine useful features and have serious over-fitting problems. In this paper, we propose a clean and effective feature fusion adversarial learning network to mine useful features and relieve over-fitting problems. Firstly, we train a fully convolution autoencoder network with unsupervised learning to mine useful feature maps from our liver lesion data. Secondly, these feature maps will be transferred to our adversarial SENet network for liver lesion classification. Our experiments on liver lesion classification in CT show an average accuracy as 85.47% compared with the baseline training scheme, which demonstrate our proposed method can mime useful features and relieve over-fitting problem. It can assist physicians in the early detection and treatment of liver lesions.

Gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced magnetic resonance imaging (MRI) has high diagnostic accuracy in the detection of liver lesions. There is a demand for computer-aided detection/diagnosis software for Gd-EOB-DTPA-enhanced MRI. We propose a deep learning-based method using one three-dimensional fully convolutional residual network (3D FC-ResNet) for liver segmentation and another 3D FC-ResNet for simultaneous detection and classification of a focal liver lesion in Gd-EOB-DTPA-enhanced MRI. We prepared a five-phase (unenhanced, arterial, portal venous, equilibrium, and hepatobiliary phases) series as the input image sets and labeled focal liver lesion (hepatocellular carcinoma, metastasis, hemangiomas, cysts, and scars) images as the output image sets. We used 100 cases to train our model, 42 cases to determine the hyperparameters of our model, and 42 cases to evaluate our model. We evaluated our model by free-response receiver operating characteristic curve analysis and using a confusion matrix. Our model simultaneously detected and classified focal liver lesions. In the test cases, the detection accuracy for whole focal liver lesions had a true-positive ratio of 0.6 at an average of 25 false positives per case. The classification accuracy was 0.790. We proposed the simultaneous detection and classification of a focal liver lesion in Gd-EOB-DTPA-enhanced MRI using multichannel 3D FC-ResNet. Our results indicated simultaneous detection and classification are possible using a single network. It is necessary to further improve detection sensitivity to help radiologists.

Detection and classification of different liver lesions: Comparison of Gd-EOB-DTPA-enhanced MRI versus multiphasic spiral CT in a clinical single centre investigation

The purpose of this study was to determine the prevalence of benign liver lesions in patients with breast cancer who are referred to magnetic resonance (MR) imaging for suspected breast cancer metastases at initial presentation. The original MR imaging reports of consecutive patients with breast cancer were reviewed; these patients had undergone MR imaging at our institution to investigate for suspected breast cancer liver metastases, at initial presentation between April 1993 and May 1998. Determination of the presence of benign and malignant liver lesions in each patient was made, as well as their relative frequencies. Diagnostic accuracy of MR imaging was evaluated by correlation with histologic specimens (5 patients) and imaging follow-up (27 patients). Thirty-four patients with newly diagnosed breast carcinoma were evaluated with MR imaging. A total of 11 (32%) of these patients had benign lesions only. Of 21 (62%) total patients who had malignant liver lesions, 19 had breast cancer metastases (2 had coexistent benign lesions), 1 had metastatic carcinoid, and 1 had hepatocellular carcinoma. No liver lesions were detected in two patients (6%). In one patient with biopsy-proven subcentimeter breast metastases, no focal lesions were shown on MR imaging. No other diagnostic errors in classification of liver lesions by MR imaging occurred, as shown by clinical correlation and imaging follow-up in all patients. True positive detection of malignant liver lesion was 20/21, true negative was 13/13, false positive was 0/13, and false negative was 1/21, for a sensitivity of 95% and a specificity of 100% for the detection of malignant liver lesions. Benign liver lesions are common in breast cancer patients suspected clinically of having liver metastases. Benign lesions alone were observed in one-third of our patients. The high diagnostic accuracy of MR imaging in the evaluation of hepatic lesions underscores the value of this technique for baseline investigation of breast cancer patients with clinically suspected liver metastases, particularly patients in whom treatment approaches are dramatically affected by the presence of liver metastases. J. Magn. Reson. Imaging 1999;10:165–169. © 1999 Wiley-Liss, Inc.

The purpose of this study was to determine the prevalence of benign liver lesions in patients with breast cancer who are referred to magnetic resonance (MR) imaging for suspected breast cancer metastases at initial presentation. The original MR imaging reports of consecutive patients with breast cancer were reviewed; these patients had undergone MR imaging at our institution to investigate for suspected breast cancer liver metastases, at initial presentation between April 1993 and May 1998. Determination of the presence of benign and malignant liver lesions in each patient was made, as well as their relative frequencies. Diagnostic accuracy of MR imaging was evaluated by correlation with histologic specimens (5 patients) and imaging follow-up (27 patients). Thirty-four patients with newly diagnosed breast carcinoma were evaluated with MR imaging. A total of 11 (32%) of these patients had benign lesions only. Of 21 (62%) total patients who had malignant liver lesions, 19 had breast cancer metastases (2 had coexistent benign lesions), 1 had metastatic carcinoid, and 1 had hepatocellular carcinoma. No liver lesions were detected in two patients (6%). In one patient with biopsy-proven subcentimeter breast metastases, no focal lesions were shown on MR imaging. No other diagnostic errors in classification of liver lesions by MR imaging occurred, as shown by clinical correlation and imaging follow-up in all patients. True positive detection of malignant liver lesion was 20/21, true negative was 13/13, false positive was 0/13, and false negative was 1/21, for a sensitivity of 95% and a specificity of 100% for the detection of malignant liver lesions. Benign liver lesions are common in breast cancer patients suspected clinically of having liver metastases. Benign lesions alone were observed in one-third of our patients. The high diagnostic accuracy of MR imaging in the evaluation of hepatic lesions underscores the value of this technique for baseline investigation of breast cancer patients with clinically suspected liver metastases, particularly patients in whom treatment approaches are dramatically affected by the presence of liver metastases. J. Magn. Reson. Imaging 1999;10:165-169.

Introduction: In the last decade one of the main reasons for people mortality and disability is liver diseases. Early detection of these diseases can help adopt appropriate treatment methods. Ultrasound imaging is a non-invasive method for visualizing tissue specification and liver lesions detection which its resolution is lower than CT and MRI images. Precise determination of liver tissue lesions and progression degree of disease is possible with advanced computer techniques such as artificial neural networks (ANN) from medical images. In this paper, a classification-based method is presented to identify and diagnose liver lesions using the Gabor wavelet features and edge detection. In this method, the vector of features from healthy and damaged tissues is trained to the network based on Gabor filters. Then the suspected cases of tissue lesions in various liver diseases are identified by features extraction of entry images. After that, the edge detection technique is implemented and the internal points of the edge are tested as an inputs of a neural network which determine the healthy and unhealthy liver tissues.
 Methods: Image features are extracted and processed by Gabor wavelet. Also the ANN is used to liver disease classification based on the images features. The forward multilayer perceptron neural network is organized with three layers of input, hidden and output. The training of this network is done with back propagation method and all of the data include "healthy tissues" and "damaged tissues" of the liver are collected in a large cellular array. Furthermore, an edge detection technique is used to indicate the points where the intensity of the light changes sharply. The sharp changes in image characteristics are usually representative of important events and changes in environments characteristics.
 Results: The results of the implementation indicate a significant reduction in processing time of liver ultrasound images and also increase the precision and accuracy of liver lesions detection (approximately 5%) among different classified groups of hepatic patients compared with the similar image processing methods. In the proposed method, the total time of operations include feature extraction, image processing, lesions detection and diagnosis of the disease has been decreased by reduction of the number of examined points. In addition, an edge detection technique had been used to diagnose the size of damaged tissues in various liver diseases, which helps improve the early detection of tissue lesions because of reduction of the checking domain of points.
 Conclusion: In this paper, a new method was presented to identify liver tissue lesions. Gabor wavelet method is employed to extract the features of the liver ultrasound images. These wavelets provide the context to understand the images frequency and their analysis in the area of the space, and given their great advantage, which is slow changes in the frequency domain, it is an appropriate filter to extract the image features. Then, the extracted features of the ultrasound images of various liver patients are stored to train a neural network, and finally the image processing method is performed to identify the healthy and damaged tissues and also to diagnose the type of disease. The search scope of problem is minimized as the input of the neural network to find the liver damaged tissue by the edge detection technique which is lead to errors reduction in identifying the tissue damages, increasing the detection speed of these lesions, and diagnosing the disease as well as determining the damage degree of liver.

This paper discusses about a method adopted to develop a computer-aided diagnostic system to achieve automatic detection and classification of liver lesions. The procedure followed consists of first segmenting the CT scan image so as to accurately extract out the lesion region alone from the rest of the abdominal details. This Region Of Interest(ROI) is now used up for extracting out first order and second order statistical feature values, which aids in the correct classification of lesions. The lesions can be classified into five types: normal liver, cysts, abscesses, benign growth (hemangioma, focal nodular hyperplasia, hepatocellular adenoma etc) and malignant growth (Hepatocellular Carcinoma, metastases etc), and this paper discusses a robust method for correctly identifying and classifying these lesions of the liver.

BackgroundTo assess the feasibility of dual-contrast spectral photon-counting computed tomography (SPCCT) for liver imaging.MethodsWe present an SPCCT in-silico study for simultaneous mapping of the complementary distribution in the liver of two contrast agents (CAs) subsequently intravenously injected: a gadolinium-based contrast agent and an iodine-based contrast agent. Four types of simulated liver lesions with a characteristic arterial and portal venous pattern (haemangioma, hepatocellular carcinoma, cyst, and metastasis) are presented. A material decomposition was performed to reconstruct quantitative iodine and gadolinium maps. Finally, a multi-dimensional classification algorithm for automatic lesion detection is presented.ResultsOur simulations showed that with a single-scan SPCCT and an adapted contrast injection protocol, it was possible to reconstruct contrast-enhanced images of the liver with arterial distribution of the iodine-based CA and portal venous phase of the gadolinium-based CA. The characteristic patterns of contrast enhancement were visible in all liver lesions. The approach allowed for an automatic detection and classification of liver lesions using a multi-dimensional analysis.ConclusionsDual-contrast SPCCT should be able to visualise the characteristic arterial and portal venous enhancement with a single scan, allowing for an automatic lesion detection and characterisation, with a reduced radiation exposure.

Purpose: To evaluate the benefit of the additional available information present in spectral CT datasets, as compared to conventional CT datasets, when utilizing convolutional neural networks for fully automatic localisation and classification of liver lesions in CT images. Materials and Methods: Conventional and spectral CT images (iodine maps, virtual monochromatic images (VMI)) were obtained from a spectral dual-layer CT system. Patient diagnosis were known from the clinical reports and classified into healthy, cyst and hypodense metastasis. In order to compare the value of spectral versus conventional datasets when being passed as input to machine learning algorithms, we implemented a weakly-supervised convolutional neural network (CNN) that learns liver lesion localisation without pixel-level ground truth annotations. Regions-of-interest are selected automatically based on the localisation results and are used to train a second CNN for liver lesion classification (healthy, cyst, hypodense metastasis). The accuracy of lesion localisation was evaluated using the Euclidian distances between the ground truth centres of mass and the predicted centres of mass. Lesion classification was evaluated by precision, recall, accuracy and F1-Score. Results: Lesion localisation showed the best results for spectral information with distances of 8.22 ± 10.72 mm, 8.78 ± 15.21 mm and 8.29 ± 12.97 mm for iodine maps, 40 keV and 70 keV VMIs, respectively. With conventional data distances of 10.58 ± 17.65 mm were measured. For lesion classification, the 40 keV VMIs achieved the highest overall accuracy of 0.899 compared to 0.854 for conventional data. Conclusion: An enhanced localisation and classification is reported for spectral CT data, which demonstrates that combining machine-learning technology with spectral CT information may in the future improve the clinical workflow as well as the diagnostic accuracy.

Automatic liver lesion classification on computed tomography images is of great importance to early cancer diagnosis and remains a challenging task. State-of-the-art liver lesion classification algorithms are currently based on manually selected regions of interest (ROIs) or automatically detected ROIs. However, liver lesions usually vary in size and shape, which makes the ROI selection process labor-intensive and also poses an obstacle to automatic lesion detection. In this paper, we propose a dual-attention dilated residual network (DADRN) as a potential solution to lesion classification task without manual ROI selection or automatic lesion detection. We incorporated a novel dual-attention module in order to capture the non-local feature dependencies and help the deep neural network focus on the lesion area by enlarging the difference between the lesion area and nonlesion area. To the best of our knowledge, we are the first to employ the self-attention mechanism to address liver lesion classification task. In addition, the well-trained DADRN can be used for weakly-supervised lesion localization without any architectural change or retraining. Experiment results show that DADRN could achieve a lesion classification accuracy comparable to that of the state-of-the-art ROI-based method and outperformed state-of-the-art attention-based approaches in both liver lesion classification and localization tasks.

Aim. To develop an artificial intelligence-based system for the diagnosis of focal liver lesions aimed at supporting clinical decision-making in surgical hepatology.Materials and methods. An artificial intelligence-based technological service was developed for the automatic segmentation and classification of contrast-enhanced computed tomography (CT) images of four types of liver lesions: focal nodular hyperplasia, carcinoma, hemangioma, and simple cyst. The service was trained and tested on datasets comprising 725 CT images using the nnU-Net architecture. Diagnostic performance was evaluated by calculating the AUC-ROC, sensitivity, specificity, and accuracy.Results. The service achieved high performance metrics. The AUC-ROC ranged from 0.847 to 0.928, with a maximum sensitivity of 0.940 for carcinoma and a specificity of 0.900 for focal nodular hyperplasia. Accuracy ranged from 0.883 to 0.922, which demonstrates the algorithm's ability to clearly differentiate between malignant and benign lesions.Conclusion. The machine learning-based service demonstrated high diagnostic performance and shows promise for integration into clinical practice, offering improved detection and classification of liver lesions.