Ultrasound Kidney Images Research Articles

Renal ultrasound to evaluate for blunt renal trauma in children: A retrospective comparison to contrast enhanced CT imaging

The standard imaging modality for hemodynamically stable blunt abdominal trauma patients is a contrast enhanced CT scan, which is reflected in the current AUA urotrauma guidelines. This comes, however, with radiation exposure and the potential sequalae of IV contrast administration in the pediatric patient. We hypothesize that ultrasound imaging would be able to diagnose and rule out clinically significant renal injuries when compared to the gold standard of CT scan in the setting of pediatric blunt abdominal trauma. All children <18 years of age who were evaluated for blunt abdominal trauma who had a CT scan and ultrasound imaging of kidneys were identified. The ultrasound images were reviewed by four reviewers who were blinded to CT results and all clinical information. The ability of ultrasound to diagnose and rule out clinically significant renal injury was evaluated by diagnostic test performance characteristics including sensitivity, specificity, negative predictive value and positive predictive value. There were 76 patients identified, 24 of which had a renal injury (1 bilateral) diagnosed by CT scan for a total of 25 injuries in 152 renal units. There were six grade I-II injuries and 19 grade III-V injuries. The sensitivity of the four blinded reviewers by ultrasound alone to detect the 19 grade III-V injuries ranged from 79 to 100% with NPV between 97 and 100%. Three of the four reviewers identified all 19 grade III-V injuries by ultrasound. When combined with significant hematuria, all 19 grade III-IV injuries were identified. Of note, all patients with a grade III-V injury of the kidney had significant hematuria. Of the grade I-II renal injuries, all reviewers identified 1/5 or 2/5 by ultrasound alone. The limitations of this study include: its retrospective nature, limited number of patients and reviewers, quality of the ultrasound machine. and experience of technologist, radiologist and urologist. A major limitation is the inability to assess other solid organ injuries during this initial study. When compared to a CT scan as the gold standard, kidney ultrasound images had a sensitivity of 79-100% to detect grade III-V injuries and NPV of 97-100% by four blinded reviewers. All grade III-V injuries had either an episode of gross hematuria or microscopic hematuria >50 RBC/hpf. A prospective study that includes full abdominal imaging is needed to confirm that ultrasound can safely be used in place of CT scan for evaluation of hemodynamically stable blunt trauma patients.

Automated detection of kidney abnormalities using multi-feature fusion convolutional neural networks

Kidney abnormalities have a high incidence in people of all ages. The requisite manual examinations to detect these abnormalities are costly and time consuming. The rapid and accurate detection of kidney abnormalities has emerged as a focus of computer-aided medical research. Traditional methods work on images of kidneys and rely on hand-crafted features to identify abnormal symptoms as independent classes that lack generalization to different kidney diseases. In this study, an automated architecture to detect various kidney abnormalities is proposed that works on abdominal ultrasound images using convolutional neural networks. Our model consists of three components: the first selects appropriate ultrasound images of kidneys, the second is a detection model used to locate the area occupied by the kidney in the given image, and the third component using the multi-feature fusion neural network (Mf-Net) discriminates normal and abnormal kidneys, by determining whether there are abnormal symptoms in images. The detection model is combined with a weighted ensemble method to improve performance. A multi-feature fusion layer is also designed in the Mf-Net to extract distinctive features for multiple views of images. The three components work together to automatically recognize abnormalities associated with kidneys. A large dataset containing 3,722 abdominal ultrasound images with classification and localization annotations is established to train and evaluate the model. Experimental results show that the proposed ensemble detection model performs best with an average TPF of 98.0%, and the Mf-Net achieves an average classification accuracy of 94.67%. The obtained high classification and detection accuracy demonstrate the effectiveness of the proposed method for recognizing kidney abnormalities.

Anatomical Kidney Model Fabrication For Ultrasound Simulation Using Direct 3D Printing

With the advance of 3D printing technology showing improvements in increasing number of medical applications, there are more and more 3D kidney model printings emerging in the field. Material selection is an important aspect in 3D printing technology. Having compared many materials, Agilus30 was chosen as our final option due to its high printing accuracy, feasibility to be printed directly and desired ultrasound compatibility. ITK-SNAP 3.6.0, Seg3D and Netfabb2019 were used to segment structures in 3D medical images which were downloaded from embodi website, to smooth the model and to export the model to STL format file respectively. Eventually we printed a whole kidney in Agilus30 using Object500 Connex3 3D printer and two halves kidney in rigid resin using Anycubic 3D printer. Then we used X3-1 3Mhz probe image to get ultrasound result from the Agilus30 model in water to get outer structure and the rigid resin model in paraffin to get inner structure. After validating the result with that of a real kidney ultrasound image, we found the results are in good agreement.

Feature Extraction and Selection of kidney Ultrasound Images Using GLCM and PCA

Medical imaging applications in hospitals and laboratories have shown benefits in visualizing patient’s body for diagnosis and treatment of disease. Ultrasound (US) is considered as safest medical imaging technique and is therefore used extensively in medical and healthcare using computer aided system. However presence of some artifacts due to patient mobility and equipment limitations makes diagnosis of these US images difficult. There is need for some pre-processing methods to improve quality of images for the purpose of classification and segmentation while preserving pixels of interest. These pixels contain information about images known as image features which forms the data model for classification. So, feature extraction and selection is important phase in classification step of diagnostic system. Keeping this in mind, this study focuses on preprocessing and feature extraction and selection phase of ultrasound images of kidney for making a classification model. Four operations cropping, interpolation, rotation and background removal are applied as preprocessing methods to enhance the quality of images and for making diagnosis easy and effective. Afterwards, a number of second order statistical texture features including energy, entropy, homogeneity, correlation, contrast, dissimilarity are generated using GLCM. Finally obtained features are reduced to optimal subset using principal component analysis (PCA). The results show that GLCM in combination with PCA for feature reduction gives high classification accuracy when classifying images using Artificial Neural Network (ANN).

Artificial intelligence approaches to improve kidney care.

Artificial intelligence is increasingly being used to improve diagnosis and prognostication for acute and chronic kidney diseases. Studies with this objective published in 2019 relied on a variety of available data sources, including electronic health records, intraoperative physiological signals, kidney ultrasound imaging, and digitized biopsy specimens.

Speckle noise suppression in 2D ultrasound kidney images using local pattern based topological derivative

One of the affordable and least harmful modalities used to efficiently detect and diagnose the kidney diseases is the Ultrasound scans. The main drawback of the ultrasound images is the presence of speckle noise that reduces the efficiency of image processing and hinders the interpretation. This paper proposes a novel technique named Local Binary Pattern based Discrete Topological Derivative and its variants to address speckle noise reduction problem in 2D ultrasound kidney images. In typical Discrete Topological Derivative, the execution time is higher and as a result an optimizer is incorporated based on local pattern and gradient tolerance value resulting in 20 times reduction in execution time with improved results. The experimentation is carried out on 100 clinical 2D ultrasound images and moreover, proposed methods are compared with the competing Discrete Topological Derivative and some commonly used speckle noise removal filters. The results are promising and thereby confirms that the proposed Local Binary Pattern based Discrete Topological Derivative can produce a better speckle noise reduction that enables the doctors to detect and diagnose kidney diseases in 2D ultrasound images with lesser strain.

PELVIS Syndrome: A Case Report

Objective − PELVIS syndrome is an association between large perineal hemangiomas and congenital anomalies, including anorectal, urinary tract, spine, and external genitalia malformations. We present a 7-year-old boy with PELVIS syndrome. Case report − . The boy had been monitored since early intrauterine age by ultrasound (US) due to left kidney cysts. Clinical examination at birth found large segmental hemangiomas in the perineal, gluteal region, and the left leg. At the age of 4 weeks US showed a multicystic, dysplastic left kidney with compensatory hypertrophy of the right kidney. Voiding cystography (VCUG) did not show vesicoureteral reflux and morphology of the bladder, and the urethra was normal. PELVIS syndrome was suspected and the child underwent further diagnostic tests which showed: undescended left testis with a hypoplastic left scrotum and micropenis, right-sided angulation of the sacrum and closed spinal dysraphism from S1 to S5, with lipoma of the filum terminale, without tethering. The urodynamics investigation was normal. During the seven-year follow-up there was spontaneous regression of the left kidney cysts; the growth and function of the right kidney was normal, as was also sphincter control. There was significant spontaneous regression of the hemangiomas. Follow-up magnetic resonance imaging (MRI) of the lumbosacral spine showed no progression of the lipoma. Conclusion − Large perineal hemangiomas should be recognized as indicators of underlying congenital anomalies, including urogenital, anorectal and spinal. Therefore, additional diagnostic evaluation, including kidney ultrasound and spine imaging, are necessary.

Detection of Renal Calculi in Ultrasound Image Using Meta-Heuristic Support Vector Machine.

Typically, the acquired renal ultrasound image includes a course of speckle noises. This paper primarily investigates an approach for the detection of renal calculi by processing those raw US images with the help of a meta-heuristic SVM classifier. One of the major downsides of involving Ultrasound images in medical analysis is the prevalence of Speckle Noises. An Adaptive Mean Median Filter approach has been introduced in the work to get rid of the speckle noises to the maximum extent ever in the literature. Segmentation is performed by employing conventional K-Means and GLCM features were extracted for classification using a meta-heuristic SVM classifier. The proposed methodology investigates with a Real-time Acquired Dataset of Mithra Scans, Tamilnadu, India comprises of 250 clinical Ultra-Sound Kidney Images of which 150 are having Calculi and the rest are Healthy. With the experimental results, the proposed meta-heuristic SVM classifier have performed better in noisy images while comparing with other conventional methods considered in the literature. It exhibits an Accuracy of 98.8% with a FAR rate of 1.8 for FRR as high as 3.3. The results clearly proposed that the novel AMM-PSO-SVM could be a promising technique for object or foreign body detection in a medical imaging application that uses ultrasound imaging.

SAT-180 KIDNEY FUNCTION CLASSIFICATION AND PREDICTION THROUGH ULTRASOUND-BASED KIDNEY IMAGING: FROM DEEP LEARNING TO MASS SCREENING OF CHRONIC KIDNEY DISEASE

Prediction of kidney function and chronic kidney disease (CKD) by kidney ultrasound imaging has long been desirable in the clinical practice for its safety, convenience, and affordability. However, this application is beyond the reach of human eyes. In the renaissance era of artificial intelligence, we proposed a deep learning approach to automatically determine the level of estimated glomerular filtration rate (GFR) and CKD status.

Design, Evaluation and Prototyping of a New Robotic Mechanism for Ultrasound Imaging

This paper presents a new robotic mechanism for ultrasound imaging. The device is placed on a patient's body by an operator, and an ultrasound expert controls the motions of the device to obtain ultrasound images. The paper focuses on the robotic mechanism that performs ultrasound imaging. The design of the mechanism is based on two approaches to produce center of motion for an ultrasound probe. This center of motion which is located on the tip of the ultrasound probe helps to create clear ultrasound images. Detailed designs, kinematic relationships, prototyping and ultrasound imaging tests are presented. A novel cabling mechanism is developed to create the center of motion required for ultrasound imaging. The mechanism provides all four necessary motions for ultrasound imaging by using two actuators which significantly reduces the weight of the device to make it suitable for portable ultrasound applications. The device has been successfully used for ultrasound imaging of kidney, gallbladder, liver, ovary and uterus of volunteer patients.

Automation of the kidney function prediction and classification through ultrasound-based kidney imaging using deep learning

Prediction of kidney function and chronic kidney disease (CKD) through kidney ultrasound imaging has long been considered desirable in clinical practice because of its safety, convenience, and affordability. However, this highly desirable approach is beyond the capability of human vision. We developed a deep learning approach for automatically determining the estimated glomerular filtration rate (eGFR) and CKD status. We exploited the transfer learning technique, integrating the powerful ResNet model pretrained on an ImageNet dataset in our neural network architecture, to predict kidney function based on 4,505 kidney ultrasound images labeled using eGFRs derived from serum creatinine concentrations. To further extract the information from ultrasound images, we leveraged kidney length annotations to remove the peripheral region of the kidneys and applied various data augmentation schemes to produce additional data with variations. Bootstrap aggregation was also applied to avoid overfitting and improve the model’s generalization. Moreover, the kidney function features obtained by our deep neural network were used to identify the CKD status defined by an eGFR of <60 ml/min/1.73 m2. A Pearson correlation coefficient of 0.741 indicated the strong relationship between artificial intelligence (AI)- and creatinine-based GFR estimations. Overall CKD status classification accuracy of our model was 85.6% —higher than that of experienced nephrologists (60.3%–80.1%). Our model is the first fundamental step toward realizing the potential of transforming kidney ultrasound imaging into an effective, real-time, distant screening tool. AI-GFR estimation offers the possibility of noninvasive assessment of kidney function, a key goal of AI-powered functional automation in clinical practice.

Novel approaches to diagnostic imaging of locally advanced cervical cancer

Cervical cancer is the second most common cancer after breast cancer and the third most common cause of cancer-related death followed by breast and lung cancers among women worldwide. advances in diagnostic imaging techniques provide better assessment of regional and distant cervical cancer metastasis. the use of contrast-enhanced ultrasound is a revolutionary imaging modality; it has several advantages over ct: no radiation exposure, nephrotoxicity, obtaining real-time information, relatively low cost and ease of use. currently, the contrast agent sonoVue is widely used in ultrasound imaging of liver, kidneys and pancreas lesions, as well as for closed abdominal injuries, multiple organ failure, breast and prostate cancers, etc. However, the role of contrast-enhanced ultrasound in gynecology is not clearly established. one of the most effective tools for the detection of locally advanced cervical cancer is mRi, which is used mainly to determine the local extent of the tumor. However, the use of functional mRitechniques has not yet been included in the standards. cervical cancer tissue has been found to show significantly lower diffusion-weighted imaging (dWi) values than normal cervical tissue, thus facilitating the detection of tumor and its spread. dWiis also used for differentiating changes after biopsy from residual tumor and for identifying small lymph nodes. the pEt/cttechnique combines the metabolic images of pEtwith anatomical images of ctand is more accurate than high resolution ctalone, especially in determining the involvement of regional lymph nodes and distant organs. 18-Fdg-pEt/cthas been successfully used for accurate staging of the disease (especially late stage), assessment of treatment response, radiotherapy planning, and detection of disease progression. in patients with advanced stages of cervical cancer (iiBiV stage), the 18-Fdg-pEt/ ctfindings can determine the treatment strategy in most cases, primarily due to high sensitivity (75–100 %) and specificity (87–100 %) in the detection of lymph node metastases.

Optimal GLCM combined FCM segmentation algorithm for detection of kidney cysts and tumor

In this document, we employed an efficient Optimal GLCM attribute related FCM segmentation algorithm which is used to categorize the kidney cysts and tumor from the ultrasound kidney images. The FCM is exploiting some appropriate attributes of GLCM texture feature extractor and optimally attach the cluster centroids of FCM by the help of Whale optimization algorithm. The proposed approach is executed in the working platform of Matlab. The findings demonstrate that the proposed model have better performance in recognizing the detection of kidney cysts and tumor in patients by examining US kidney images. Also, we have shown the comparison of our proposed method FB-FCM-WOA with the existing methodologies like FB-FCM, FB-K-means, IB-FCM and IB-K-means. Hence, we would suggest that our proposed method is much better for detecting kidney cysts and tumor.

Automatic Detection of Renal Abnormalities by Off-the-shelf CNN Features

ABSTRACTThe chronic kidney disease (CKD) and end-stage renal disease are the leading causes of kidney (renal) failure which may be fatal. Thus, it is important to diagnose the kidney diseases at an earlier stage before it cause fatality. This paper proposes an algorithm for automatic kidney abnormalities detection and classification at an early stage. Kidney abnormalities detection involves two stages: feature extraction and detection. The input images used in this work are kidney ultrasound (US) images which are classified into three categories such as normal kidney, kidney with cyst and kidney with stone. The pre-trained convolutional neural networks (CNN) is used to extract features from the kidney US images then, the extracted features from CNN are fed into the support vector machine (SVM) classifier for classifying the kidney abnormalities. The training and testing phase of CNN requires large number of labelled images. Due to unavailability of larger datasets in the kidney US images, off-the-shelf CNN features are used which perform well for smaller datasets. Performance evaluation is done for classification and 91.8% of the accuracy is obtained. Further, the sensitivity and specificity achieved for the SVM classifier in detecting normal kidney images is 93.75% and 92.5%, respectively.

FPGA based Abnormality Classification in Kidney Ultrasound Images using KNN

Ultrasound imaging has been widely used for diagnosing the internal abnormalities. Tele-radiographers are lacking nowadays to provide accurate diagnosis of the presence of abnormalities. To overcome this issue, Teleradiology paved a new way for the online access of the doctors around all parts of the world. Frequent access through online is a bottleneck problem. So, CAD based technique is implemented in VIRTEX-6 FPGA platform in order to clearly classify the presence of the abnormality. Manual intervention is reduced in this case. The proposed algorithm involves five steps: pre-processing segmentation, feature extraction, feature selection and classification. Based on selected features, the classification is made using KNN Classifier. Good range of specificity is achieved which is comparable with the existing algorithms. The specificity for normal, cystic and stone kidneys is 95, 80 and 75 respectively. The proposed method produces far higher accuracy when compared with SVM using linear, RBF and polynomial kernel. But it shows a bit lesser value when compared with SVM using MLP kernel.

A modified level set algorithm based on point distance shape constraint for lesion and organ segmentation

PurposeThe segmentation of organs and lesions from medical images is a challenging task due to the presents of noise, intensity inhomogeneity, blurry/weak boundaries. In this paper, a point distance shape constraint is proposed and incorporated in the level set framework for the segmentation of objects with various shapes. MethodsThe proposed shape constraint is a linear combination of the Euclidean distance of user selected points. By selecting different numbers of points, it can generate different shape constraints and therefore is more flexible in dealing with different shapes. Then this shape constraint is incorporated into the variational level set framework. A convex relaxation is applied to get a convex model which can be efficiently solved by a primal-dual hybrid gradient algorithm. ResultsThe proposed algorithm is tested on 60 CT images with the nodular type of hepatic cellular cancer (HCC), 100 ultrasound kidney images, 20 prostate MR images, 20 lumbar CT images and 100 transrectal ultrasound prostate images. The algorithms performance is evaluated using a number of metrics by comparison with expert delineations. The validation results show that, for five datasets mentioned previously, the average DSCs of the proposed algorithm are 95.6% ± 1.4%, 94.3% ± 3.1%, 91.3% ± 3.8%, 92.7% ± 1.5% and 94.4% ± 2.2% respectively. Both quantitative and qualitative evaluation confirm that the proposed method can provide more accurate segmentation than four state-of-the-art methods. ConclusionThe proposed point distance shape constraint segmentation model can accurately segment organs and lesions with a number of shapes in medical images.

Feature Analysis of Kidney Ultrasound Image in Four Different Ultrasound using Gray Level Co-occurrence Matrix (GLCM) and Intensity Histogram (IH)

Misinterpretation analysis of ultrasound images has been huge issues in the world nowadays. Lack of skills and knowledge, as well as unclear ultrasound image due to the presence of speckle noise in ultrasound, are some factors lead to this issue. In this research, we compare 188 kidney ultrasound images from four different types of ultrasound machines, named as ultrasound A, B, C and D. Image pre-processing of images which involve cropping, enhancement, and filtering are performed before manual segmentation and texture analysis process to indicates the wanted region and improve contrast in each image. Texture analysis is performed using gray level co-occurrence matrix (GLCM) and intensity histogram (IH) to find differences and similarities in kidney image texture between all ultrasounds. Four GLCM parameters, contrast, correlation, energy and homogeneity and four parameters from IH (mean, standard deviation, variance, and skewness) used to indicate the most significant features between all ultrasound machines. Results show that contrast in GLCM is the most significant features that can be extracted from all four ultrasound machines and will be used in the classification process.

Computer-aided diagnosis of congenital abnormalities of the kidney and urinary tract in children based on ultrasound imaging data by integrating texture image features and deep transfer learning image features

Anatomic characteristics of kidneys derived from ultrasound images are potential biomarkers of children with congenital abnormalities of the kidney and urinary tract (CAKUT), but current methods are limited by the lack of automated processes that accurately classify diseased and normal kidneys. The objective of the study was to evaluate the diagnostic performance of deep transfer learning techniques to classify kidneys of normal children and those with CAKUT. A transfer learning method was developed to extract features of kidneys from ultrasound images obtained during routine clinical care of 50 children with CAKUT and 50 controls. To classify diseased and normal kidneys, support vector machine classifiers were built on the extracted features using (1) transfer learning imaging features from a pretrained deep learning model, (2) conventional imaging features, and (3) their combination. These classifiers were compared, and their diagnosis performance was measured using area under the receiver operating characteristic curve (AUC), accuracy, specificity, and sensitivity. The AUC for classifiers built on the combination features were 0.92, 0.88, and 0.92 for discriminating the left, right, and bilateral abnormal kidney scans from controls with classification rates of 84%, 81%, and 87%; specificity of 84%, 74%, and 88%; and sensitivity of 85%, 88%, and 86%, respectively. These classifiers performed better than classifiers built on either the transfer learning features or the conventional features alone (p<0.001). The present study validated transfer learning techniques for imaging feature extraction of ultrasound images to build classifiers for distinguishing children with CAKUT from controls. The experiments have demonstrated that the classifiers built on the transfer learning features and conventional image features could distinguish abnormal kidney images from controls with AUCs greater than 0.88, indicating that classification of ultrasound kidney scans has a great potential to aid kidney disease diagnosis. A limitation of the present study is the moderate number of patients that contributed data to the transfer learning approach. The combination of transfer learning and conventional imaging features yielded the best classification performance for distinguishing children with CAKUT from controls based on ultrasound images of kidneys.

Review on Classifications of Medical Ultrasound Images of Kidney

Review on Classifications of Medical Ultrasound Images of Kidney

TRANSFER LEARNING FOR DIAGNOSIS OF CONGENITAL ABNORMALITIES OF THE KIDNEY AND URINARY TRACT IN CHILDREN BASED ON ULTRASOUND IMAGING DATA.

Classification of ultrasound (US) kidney images for diagnosis of congenital abnormalities of the kidney and urinary tract (CAKUT) in children is a challenging task. It is desirable to improve existing pattern classification models that are built upon conventional image features. In this study, we propose a transfer learning-based method to extract imaging features from US kidney images in order to improve the CAKUT diagnosis in children. Particularly, a pre-trained deep learning model (imagenet-caffe-alex) is adopted for transfer learning-based feature extraction from 3-channel feature maps computed from US images, including original images, gradient features, and distanced transform features. Support vector machine classifiers are then built upon different sets of features, including the transfer learning features, conventional imaging features, and their combination. Experimental results have demonstrated that the combination of transfer learning features and conventional imaging features yielded the best classification performance for distinguishing CAKUT patients from normal controls based on their US kidney images.