Curvelet based automatic segmentation of supraspinatus tendon from ultrasound image: a focused assistive diagnostic method

Beyond pixel: Superpixel-based MRI segmentation through traditional machine learning and graph convolutional network

BACKGROUND AND OBJECTIVE

Tendon segmentation is crucial for studying tendon-related pathologies like tendinopathy, tendinosis, etc. This step further enables detailed analysis of specific tendon regions using automated or semi-automated methods. This study specifically aims at the segmentation of Achilles tendon, the largest tendon in the human body.

METHODS

This study proposes a comprehensive end-to-end tendon segmentation module composed of a preliminary superpixel-based coarse segmentation preceding the final segmentation task. The final segmentation results are obtained through two distinct approaches. In the first approach, the coarsely generated superpixels are subjected to classification using Random Forest (RF) and Support Vector Machine (SVM) classifiers to classify whether each superpixel belongs to a tendon class or not (resulting in tendon segmentation). In the second approach, the arrangements of superpixels are converted to graphs instead of being treated as conventional image grids. This classification process uses a graph-based convolutional network (GCN) to determine whether each superpixel corresponds to a tendon class or not.

RESULTS

All experiments are conducted on a custom-made ankle MRI dataset. The dataset comprises 76 subjects and is divided into two sets: one for training (Dataset 1, trained and evaluated using leave-one-group-out cross-validation) and the other as unseen test data (Dataset 2). Using our first approach, the final test AUC (Area Under the ROC Curve) scores using RF and SVM classifiers on the test data (Dataset 2) are 0.992 and 0.987, respectively, with sensitivities of 0.904 and 0.966. On the other hand, using our second approach (GCN-based node classification), the AUC score for the test set is 0.933 with a sensitivity of 0.899.

CONCLUSIONS

Our proposed pipeline demonstrates the efficacy of employing superpixel generation as a coarse segmentation technique for the final tendon segmentation. Whether utilizing RF, SVM-based superpixel classification, or GCN-based classification for tendon segmentation, our system consistently achieves commendable AUC scores, especially the non-graph-based approach. Given the limited dataset, our graph-based method did not perform as well as non-graph-based superpixel classifications; however, the results obtained provide valuable insights into how well the models can distinguish between tendons and non-tendons. This opens up opportunities for further exploration and improvement.

An automated method for tendon image segmentation on ultrasound using grey-level co-occurrence matrix features and hidden Gaussian Markov random fields

BACKGROUND

Despite knowledge of qualitative changes that occur on ultrasound in tendinopathy, there is currently no objective and reliable means to quantify the severity or prognosis of tendinopathy on ultrasound.

OBJECTIVE

The primary objective of this study is to produce a quantitative and automated means of inferring potential structural changes in tendinopathy by developing and implementing an algorithm which performs a texture based segmentation of tendon ultrasound (US) images.

METHOD

A model-based segmentation approach is used which combines Gaussian mixture models, Markov random field theory and grey-level co-occurrence (GLCM) features. The algorithm is trained and tested on 49 longitudinal B-mode ultrasound images of the Achilles tendons which are labelled as tendinopathic (24) or healthy (25). Hyperparameters are tuned, using a training set of 25 images, to optimise a decision tree based classification of the images from texture class proportions. We segment and classify the remaining test images using the decision tree.

RESULTS

Our approach successfully detects a difference in the texture profiles of tendinopathic and healthy tendons, with 22/24 of the test images accurately classified based on a simple texture proportion cut-off threshold. Results for the tendinopathic images are also collated to gain insight into the topology of structural changes that occur with tendinopathy. It is evident that distinct textures, which are predominantly present in tendinopathic tendons, appear most commonly near the transverse boundary of the tendon, though there was a large variability among diseased tendons.

CONCLUSION

The GLCM based segmentation of tendons under ultrasound resulted in distinct segmentations between healthy and tendinopathic tendons and provides a potential tool to objectively quantify damage in tendinopathy.

Development and Clinical Application of Artificial Intelligence Assistant System for Rotator Cuff Ultrasound Scanning

OBJECTIVE

We developed an intelligent assistance system for shoulder ultrasound imaging, incorporating deep-learning algorithms to facilitate standard plane recognition and automatic tissue segmentation of the rotator cuff and its surrounding structures. We evaluated the system's performance using a dedicated data set of rotator cuff ultrasound images to assess its feasibility in clinical practice.

METHODS

To fulfill the system's primary functions, we designed a standard plane recognition module based on the ResNet50 network and an automatic tissue segmentation module using the Mask R-CNN model. The modules were trained on carefully curated data sets. The standard plane recognition module automatically identifies a specific standard plane based on the ultrasound image characteristics. The automatic tissue segmentation module effectively delineates and segments anatomical structures within the identified standard plane.

RESULTS

With the use of 59,265 shoulder joint ultrasound images, the standard plane recognition model achieved an impressive recognition accuracy of 94.9% in the test set, with an average precision rate of 96.4%, recall rate of 95.4% and F1 score of 95.9%. The automatic tissue segmentation model, tested on 1886 images, exhibited a commendable average intersection over union value of 96.2%, indicating robustness and accuracy. The model achieved mean intersection over union values exceeding 90.0% for all standard planes, indicating its effectiveness in precisely delineating the anatomical structures.

CONCLUSION

Our shoulder joint musculoskeletal intelligence system swiftly and accurately identifies standard planes and performs automatic tissue segmentation.

A comparison of point-tracking algorithms in ultrasound videos from the upper limb

Tracking points in ultrasound (US) videos can be especially useful to characterize tissues in motion. Tracking algorithms that analyze successive video frames, such as variations of Optical Flow and Lucas-Kanade (LK), exploit frame-to-frame temporal information to track regions of interest. In contrast, convolutional neural-network (CNN) models process each video frame independently of neighboring frames. In this paper, we show that frame-to-frame trackers accumulate error over time. We propose three interpolation-like methods to combat error accumulation and show that all three methods reduce tracking errors in frame-to-frame trackers. On the neural-network end, we show that a CNN-based tracker, DeepLabCut (DLC), outperforms all four frame-to-frame trackers when tracking tissues in motion. DLC is more accurate than the frame-to-frame trackers and less sensitive to variations in types of tissue movement. The only caveat found with DLC comes from its non-temporal tracking strategy, leading to jitter between consecutive frames. Overall, when tracking points in videos of moving tissue, we recommend using DLC when prioritizing accuracy and robustness across movements in videos, and using LK with the proposed error-correction methods for small movements when tracking jitter is unacceptable.

Comparison of multifidus muscle intramuscular fat by ultrasound echo intensity and fat-water based MR images in individuals with chronic low back pain

PURPOSE

The aim of this observational cross-sectional study was to examine correlations of intramuscular fat content in lumbar multifidus (LM) by comparing muscle echo intensity (EI) and percent fat signal fraction (%FSF) generated from ultrasound (US) and magnetic resonance (MR) images, respectively.

METHODS

MRI and US images from 25 participants (16 females, 9 males) selected from a cohort of patients with chronic low back pain (CLBP) were used. Images were acquired bilaterally, at the L4 and L5 levels (e.g., 4 sites). EI measurements were acquired by manually tracing the cross-sectional border of LM. Mean EI of three US images per site were analyzed (e.g., raw EI). A correction factor for subcutaneous fat thickness (SFT) was also calculated and applied (e.g., corrected EI). Corresponding fat and water MR images were used to acquire %FSF measurements. Intra-rater reliability was assessed by intraclass coefficients (ICC). Pearson correlations and simple linear regression were used to assess the relationship between %FSF, raw EI and corrected EI measurements.

RESULTS

The intra-rater ICCs for all measurements were moderate to excellent. Correlations between %FSF vs. raw EI and corrected EI were moderate to strong (0.40 < r < 0.52) and (0.40 < r < 0.51), respectively. Moderate correlations between SFT and EI were also identified.

CONCLUSION

US is a low-cost, non-invasive, accessible, and reliable method to examine muscle composition, and presents a promising solution for assessing and monitoring the effect of different treatment options for CLBP in clinical settings.

An Efficient Prediction System for Coronary Heart Disease Risk Using Selected Principal Components and Hyperparameter Optimization

Medical science-related studies have reinforced that the prevalence of coronary heart disease which is associated with the heart and blood vessels has been the most significant cause of health loss and death globally. Recently, data mining and machine learning have been used to detect diseases based on the unique characteristics of a person. However, these techniques have often posed challenges due to the complexity in understanding the objective of the datasets, the existence of too many factors to analyze as well as lack of performance accuracy. This research work is of two-fold effort: firstly, feature extraction and selection. This entails extraction of the principal components, and consequently, the Correlation-based Feature Selection (CFS) method was applied to select the finest principal components of the combined (Cleveland and Statlog) heart dataset. Secondly, by applying datasets to three single and three ensemble classifiers, the best hyperparameters that reflect the pre-eminent predictive outcomes were investigated. The experimental result reveals that hyperparameter optimization has improved the accuracy of all the models. In the comparative studies, the proposed work outperformed related works with an accuracy of 97.91%, and an AUC of 0.996 by employing six optimal principal components selected from the CFS method and optimizing parameters of the Rotation Forest ensemble classifier.

Intelligent Automatic Segmentation of Wrist Ganglion Cysts Using DBSCAN and Fuzzy C-Means

Ganglion cysts are common soft tissue masses of the hand and wrist, and small size cysts are often hypoechoic. Thus, identifying them from ultrasonography is not an easy problem. In this paper, we propose an automatic segmentation method using two artificial intelligence algorithms in sequence. A density based unsupervised learning algorithm called DBSCAN is performed as a front-end and its result determines the number of clusters used in the Fuzzy C-Means (FCM) clustering algorithm for quantification of ganglion cyst object. In an experiment using 120 images, the proposed method shows a higher extraction rate (89.2%) and lower false positive rate compared with FCM when the ground truth is set as the human expert's decision. Such human-like behavior is more apparent when the size of the ganglion cyst is small that the quality of ultrasonography is often not very high. With this fully automatic segmentation method, the operator subjectivity that is highly dependent on the experience of the ultrasound examiner can be mitigated with high reliability.

A new method of contrast enhancement of musculoskeletal ultrasound imaging based on fuzzy inference technique

Improving the clarity and visual quality of Musculoskeletal Ultrasound Images (MUI) can help clinicians to detect diseases more easily and accurately. In this work, we described how to enhance the contrast of MUI locally based on a fuzzy inference system. Local Fuzzy Inference Technique (LFIT) was introduced as a novel technique to enhance the contrast of MUI. The input data used musculoskeletal ultrasound images were collected from healthy volunteers. Local Fuzzy Inference Technique (LFIT) was compared with a recent fuzzy technique of the image enhancement and validated based on assessment metrics (second-derivative-like measure of enhancement (SDME)). The results advocated an improved quality of the musculoskeletal ultrasound images based on the LFIT technique with approximately 11% greater than recent technique and computation time of LFIT is 28.4% is less. It is possible to apply a proposal technique on the other types of image (panoramic image and video). Furthermore, observed improvements on the MUI quality could potentially invested as a pre-processing step before performing other computer vision applications, such as image segmentation, tracking, and 3D image reconstruction.

Imbalanced Loss-Integrated Deep-Learning-Based Ultrasound Image Analysis for Diagnosis of Rotator-Cuff Tear

A rotator cuff tear (RCT) is an injury in adults that causes difficulty in moving, weakness, and pain. Only limited diagnostic tools such as magnetic resonance imaging (MRI) and ultrasound Imaging (UI) systems can be utilized for an RCT diagnosis. Although UI offers comparable performance at a lower cost to other diagnostic instruments such as MRI, speckle noise can occur the degradation of the image resolution. Conventional vision-based algorithms exhibit inferior performance for the segmentation of diseased regions in UI. In order to achieve a better segmentation for diseased regions in UI, deep-learning-based diagnostic algorithms have been developed. However, it has not yet reached an acceptable level of performance for application in orthopedic surgeries. In this study, we developed a novel end-to-end fully convolutional neural network, denoted as Segmentation Model Adopting a pRe-trained Classification Architecture (SMART-CA), with a novel integrated on positive loss function (IPLF) to accurately diagnose the locations of RCT during an orthopedic examination using UI. Using the pre-trained network, SMART-CA can extract remarkably distinct features that cannot be extracted with a normal encoder. Therefore, it can improve the accuracy of segmentation. In addition, unlike other conventional loss functions, which are not suited for the optimization of deep learning models with an imbalanced dataset such as the RCT dataset, IPLF can efficiently optimize the SMART-CA. Experimental results have shown that SMART-CA offers an improved precision, recall, and dice coefficient of 0.604% (+38.4%), 0.942% (+14.0%) and 0.736% (+38.6%) respectively. The RCT segmentation from a normal ultrasound image offers the improved precision, recall, and dice coefficient of 0.337% (+22.5%), 0.860% (+15.8%) and 0.484% (+28.5%), respectively, in the RCT segmentation from an ultrasound image with severe speckle noise. The experimental results demonstrated the IPLF outperforms other conventional loss functions, and the proposed SMART-CA optimized with the IPLF showed better performance than other state-of-the-art networks for the RCT segmentation with high robustness to speckle noise.

Automatic Characterizations of Lumbar Multifidus Muscle and Intramuscular Fat with Fuzzy C-means based Quantization from Ultrasound Images

BACKGROUND

Low Back Pain (LBP) is a common disorder involving the muscles and bones and about half of the people experience LBP at some point of their lives. Since the social economic cost and the recurrence rate over the lifetime is very high, the treatment/rehabilitation of chronic LBP is important to physiotherapists, both for clinical and research purposes. Trunk muscles such as the lumbar multifidi is important in spinal functions and intramuscular fat is also important in understanding pain control and rehabilitations. However, the analysis of such muscles and related fat require many human interventions and thus suffers from the operator subjectivity especially when the ultrasonography is used due to its cost-effectiveness and no radioactive risk.

AIMS

In this paper, we propose a fully automatic computer vision based software to compute the thickness of the lumbar multifidi muscles and to analyze intramuscular fat distribution in that area.

METHODS

The proposed system applies various image processing algorithms to enhance the intensity contrast of the image and measure the thickness of the target muscle. Intermuscular fat analysis is done by Fuzzy C-Means (FCM) clustering based quantization.

RESULTS

In experiment using 50 DICOM format ultrasound images from 50 subjects, the proposed system shows very promising result in computing the thickness of lumbar multifidi.

CONCLUSION

The proposed system have minimal discrepancy(less than 0.2 cm) from human expert for 72% (36 out of 50 cases) of the given data. Also, FCM based intramuscular fat analysis looks better than conventional histogram analysis.

Segmentation of finger tendon and synovial sheath in ultrasound image using deep convolutional neural network

BACKGROUND

Trigger finger is a common hand disease, which is caused by a mismatch in diameter between the tendon and the pulley. Ultrasound images are typically used to diagnose this disease, which are also used to guide surgical treatment. However, background noise and unclear tissue boundaries in the images increase the difficulty of the process. To overcome these problems, a computer-aided tool for the identification of finger tissue is needed.

RESULTS

Two datasets were used for evaluation: one comprised different cases of individual images and another consisting of eight groups of continuous images. Regarding result similarity and contour smoothness, our proposed deeply supervised dilated fully convolutional DenseNet (D2FC-DN) is better than ATASM (the state-of-art segmentation method) and representative CNN methods. As a practical application, our proposed method can be used to build a tendon and synovial sheath model that can be used in a training system for ultrasound-guided trigger finger surgery.

CONCLUSION

We proposed a D2FC-DN for finger tendon and synovial sheath segmentation in ultrasound images. The segmentation results were remarkably accurate for two datasets. It can be applied to assist the diagnosis of trigger finger by highlighting the tissues and generate models for surgical training systems in the future.

METHODS

We propose a novel finger tendon segmentation method for use with ultrasound images that can also be used for synovial sheath segmentation that yields a more complete description for analysis. In this study, a hybrid of effective convolutional neural network techniques are applied, resulting in a deeply supervised dilated fully convolutional DenseNet (D2FC-DN), which displayed excellent segmentation performance on the tendon and synovial sheath.

Semi-dynamic Control of FCM Initialization for Automatic Extraction of Inflamed Appendix from Ultrasonography

BACKGROUND

Current naked-eye examination of the ultrasound images for inflamed appendix has limitations due to its intrinsic operator subjectivity problem.

OBJECTIVE

In this paper, we propose a fully automatic intelligent method for extracting inflamed appendix from ultrasound images. Accurate and automatic extraction of inflamed appendix from ultrasonography is a major decision making resource of the diagnosis and management of suspected appendicitis.

METHODS

The proposed method uses Fuzzy C-means learning algorithm in pixel clustering with semi-dynamic control of initializing the number of clusters based on the intensity contrast dispersion of the input image. Thirty percent of the prepared ultrasonography samples are classified into four different groups based on their intensity contrast distribution and then different number of clusters are assigned to the images in accordance with such groups in Fuzzy C-means learning process.

RESULTS

In the experiment, the proposed system successfully extracts the target without human intervention in 82 of 85 cases (96.47% accuracy). The proposed method also shows that it can cover the false negative cases occurred previously that used self-organizing map as the learning engine.

CONCLUSION

Such high level reliable correct extraction of inflamed appendix encourages to use the automatic extraction software in the diagnosis procedure of suspected acute appendicitis.

Automated localization and segmentation techniques for B-mode ultrasound images: A review

B-mode ultrasound imaging is used extensively in medicine. Hence, there is a need to have efficient segmentation tools to aid in computer-aided diagnosis, image-guided interventions, and therapy. This paper presents a comprehensive review on automated localization and segmentation techniques for B-mode ultrasound images. The paper first describes the general characteristics of B-mode ultrasound images. Then insight on the localization and segmentation of tissues is provided, both in the case in which the organ/tissue localization provides the final segmentation and in the case in which a two-step segmentation process is needed, due to the desired boundaries being too fine to locate from within the entire ultrasound frame. Subsequenly, examples of some main techniques found in literature are shown, including but not limited to shape priors, superpixel and classification, local pixel statistics, active contours, edge-tracking, dynamic programming, and data mining. Ten selected applications (abdomen/kidney, breast, cardiology, thyroid, liver, vascular, musculoskeletal, obstetrics, gynecology, prostate) are then investigated in depth, and the performances of a few specific applications are compared. In conclusion, future perspectives for B-mode based segmentation, such as the integration of RF information, the employment of higher frequency probes when possible, the focus on completely automatic algorithms, and the increase in available data are discussed.

Tendon-motion tracking in an ultrasound image sequence using optical-flow-based block matching

BACKGROUND

Tendon motion, which is commonly observed using ultrasound imaging, is one of the most important features used in tendinopathy diagnosis. However, speckle noise and out-of-plane issues make the tracking process difficult. Manual tracking is usually time consuming and often yields inconsistent results between users.

METHODS

To automatically track tendon motion in ultrasound images, we developed a new method that combines the advantages of optical flow and multi-kernel block matching. For every pair of adjacent image frames, the optical flow is computed and used to estimate the accumulated displacement. The proposed method selects the frame interval adaptively based on this displacement. Multi-kernel block matching is then computed on the two selected frames, and, to reduce tracking errors, the detailed displacements of the frames in between are interpolated based on the optical flow results.

RESULTS

In the experiments, cadaver data were used to evaluate the tracking results. The mean absolute error was less than 0.05 mm. The proposed method also tracked the motion of tendons in vivo, which provides useful information for clinical diagnosis.

CONCLUSION

The proposed method provides a new index for adaptively determining the frame interval. Compared with other methods, the proposed method yields tracking results that are significantly more accurate.