Detection of Collaterals from Cone-Beam CT Images in Stroke

Collateral vessels play an important role in the restoration of blood flow to the ischemic tissues of stroke patients, and the quality of collateral flow has major impact on reducing treatment delay and increasing the success rate of reperfusion. Due to high spatial resolution and rapid scan time, advance imaging using the cone-beam computed tomography (CBCT) is gaining more attention over the conventional angiography in acute stroke diagnosis. Detecting collateral vessels from CBCT images is a challenging task due to the presence of noises and artifacts, small-size and non-uniform structure of vessels. This paper presents a technique to objectively identify collateral vessels from non-collateral vessels. In our technique, several filters are used on the CBCT images of stroke patients to remove noises and artifacts, then multiscale top-hat transformation method is implemented on the pre-processed images to further enhance the vessels. Next, we applied three types of feature extraction methods which are gray level co-occurrence matrix (GLCM), moment invariant, and shape to explore which feature is best to classify the collateral vessels. These features are then used by the support vector machine (SVM), random forest, decision tree, and K-nearest neighbors (KNN) classifiers to classify vessels. Finally, the performance of these classifiers is evaluated in terms of accuracy, sensitivity, precision, recall, F-Measure, and area under the receiver operating characteristics curve. Our results show that all classifiers achieve promising classification accuracy above 90% and able to detect the collateral and non-collateral vessels from images.

Download Full-text

Automatic Grading of Stroke Symptoms for Rapid Assessment Using Optimized Machine Learning and 4-Limb Kinematics: Clinical Validation Study (Preprint)

10.2196/preprints.20641 ◽

2020 ◽

Author(s):

Eunjeong Park ◽

Kijeong Lee ◽

Taehwa Han ◽

Hyo Suk Nam

Keyword(s):

Machine Learning ◽

Medical Staff ◽

Machine Learning Algorithms ◽

Training Data ◽

Support Vector ◽

Grading System ◽

Operating Characteristics ◽

Stroke Patients ◽

Motor Weakness ◽

Automatic Grading

BACKGROUND Subtle abnormal motor signs are indications of serious neurological diseases. Although neurological deficits require fast initiation of treatment in a restricted time, it is difficult for nonspecialists to detect and objectively assess the symptoms. In the clinical environment, diagnoses and decisions are based on clinical grading methods, including the National Institutes of Health Stroke Scale (NIHSS) score or the Medical Research Council (MRC) score, which have been used to measure motor weakness. Objective grading in various environments is necessitated for consistent agreement among patients, caregivers, paramedics, and medical staff to facilitate rapid diagnoses and dispatches to appropriate medical centers. OBJECTIVE In this study, we aimed to develop an autonomous grading system for stroke patients. We investigated the feasibility of our new system to assess motor weakness and grade NIHSS and MRC scores of 4 limbs, similar to the clinical examinations performed by medical staff. METHODS We implemented an automatic grading system composed of a measuring unit with wearable sensors and a grading unit with optimized machine learning. Inertial sensors were attached to measure subtle weaknesses caused by paralysis of upper and lower limbs. We collected 60 instances of data with kinematic features of motor disorders from neurological examination and demographic information of stroke patients with NIHSS 0 or 1 and MRC 7, 8, or 9 grades in a stroke unit. Training data with 240 instances were generated using a synthetic minority oversampling technique to complement the imbalanced number of data between classes and low number of training data. We trained 2 representative machine learning algorithms, an ensemble and a support vector machine (SVM), to implement auto-NIHSS and auto-MRC grading. The optimized algorithms performed a 5-fold cross-validation and were searched by Bayes optimization in 30 trials. The trained model was tested with the 60 original hold-out instances for performance evaluation in accuracy, sensitivity, specificity, and area under the receiver operating characteristics curve (AUC). RESULTS The proposed system can grade NIHSS scores with an accuracy of 83.3% and an AUC of 0.912 using an optimized ensemble algorithm, and it can grade with an accuracy of 80.0% and an AUC of 0.860 using an optimized SVM algorithm. The auto-MRC grading achieved an accuracy of 76.7% and a mean AUC of 0.870 in SVM classification and an accuracy of 78.3% and a mean AUC of 0.877 in ensemble classification. CONCLUSIONS The automatic grading system quantifies proximal weakness in real time and assesses symptoms through automatic grading. The pilot outcomes demonstrated the feasibility of remote monitoring of motor weakness caused by stroke. The system can facilitate consistent grading with instant assessment and expedite dispatches to appropriate hospitals and treatment initiation by sharing auto-MRC and auto-NIHSS scores between prehospital and hospital responses as an objective observation.

Download Full-text

Abstract 177: Absent Collateral Flow on CT Angiogram Predicts Acute Hemorrhagic Transformation in Ischemic Stroke

Stroke ◽

10.1161/str.44.suppl_1.a177 ◽

2013 ◽

Vol 44 (suppl_1) ◽

Author(s):

Richard Burgess ◽

Esteban Cheng Ching ◽

Delora Wisco ◽

Shumei Man ◽

Ken Uchino ◽

...

Keyword(s):

Ischemic Stroke ◽

Acute Ischemic Stroke ◽

Ct Angiography ◽

Hemorrhagic Transformation ◽

Collateral Flow ◽

Stroke Patients ◽

Conventional Angiography ◽

Vessel Occlusion ◽

Post Procedure ◽

Ct Angiogram

Background: In patients with a large vessel occlusion, the degree of collateral vascular supply to an ischemic territory has been shown to be a predictor of stroke outcome. Prior studies have focused on the correlation between collateral flow measured on conventional digital subtraction angiography and outcome measures, including the presence of hemorrhagic conversion. CT/CTA is more widely available and more quickly accomplished than MR or conventional angiography. In this work we demonstrate that the absence of CT angiographic collaterals predicts hemorrhage transformation in acute ischemic stroke patients that have persistent vessel occlusion. Methods: Retrospective review of patient data from a prospectively acquired database identified acute ischemic stroke patients who underwent CT angiography followed by cerebral angiography, and post procedure non-contrast CT scans. Blinded evaluators independently assessed CT angiogram collaterals, angiographic TICI scores, and the presence and severity of post procedure hemorrhagic transformation. Fishers exact test was used to compare proportions between groups. Results: 146 patients were included. The mean age was 67. The median NIHSS was 15.5 (range 0-32). 34% of patients had any type of hemorrhagic conversion. Of patients with no collaterals on CT angiography, 63% had hemorrhagic conversion versus 23%, 33%, and 38% for patients with grades 1, 2, and 3 collaterals (p<0.05 for comparisons). Patients with TICI scores of 0 or 1 and no CTA collaterals all had hemorrhagic transformation. Conclusion: The absence of collateral flow on CT angiography in patients without recanalization strongly predicts the acute development of hemorrhagic conversion.

Download Full-text

Automatic Grading of Stroke Symptoms for Rapid Assessment Using Optimized Machine Learning and 4-Limb Kinematics: Clinical Validation Study

Journal of Medical Internet Research ◽

10.2196/20641 ◽

2020 ◽

Vol 22 (9) ◽

pp. e20641

Author(s):

Eunjeong Park ◽

Kijeong Lee ◽

Taehwa Han ◽

Hyo Suk Nam

Keyword(s):

Machine Learning ◽

Medical Staff ◽

Machine Learning Algorithms ◽

Training Data ◽

Support Vector ◽

Grading System ◽

Operating Characteristics ◽

Stroke Patients ◽

Motor Weakness ◽

Automatic Grading

Background Subtle abnormal motor signs are indications of serious neurological diseases. Although neurological deficits require fast initiation of treatment in a restricted time, it is difficult for nonspecialists to detect and objectively assess the symptoms. In the clinical environment, diagnoses and decisions are based on clinical grading methods, including the National Institutes of Health Stroke Scale (NIHSS) score or the Medical Research Council (MRC) score, which have been used to measure motor weakness. Objective grading in various environments is necessitated for consistent agreement among patients, caregivers, paramedics, and medical staff to facilitate rapid diagnoses and dispatches to appropriate medical centers. Objective In this study, we aimed to develop an autonomous grading system for stroke patients. We investigated the feasibility of our new system to assess motor weakness and grade NIHSS and MRC scores of 4 limbs, similar to the clinical examinations performed by medical staff. Methods We implemented an automatic grading system composed of a measuring unit with wearable sensors and a grading unit with optimized machine learning. Inertial sensors were attached to measure subtle weaknesses caused by paralysis of upper and lower limbs. We collected 60 instances of data with kinematic features of motor disorders from neurological examination and demographic information of stroke patients with NIHSS 0 or 1 and MRC 7, 8, or 9 grades in a stroke unit. Training data with 240 instances were generated using a synthetic minority oversampling technique to complement the imbalanced number of data between classes and low number of training data. We trained 2 representative machine learning algorithms, an ensemble and a support vector machine (SVM), to implement auto-NIHSS and auto-MRC grading. The optimized algorithms performed a 5-fold cross-validation and were searched by Bayes optimization in 30 trials. The trained model was tested with the 60 original hold-out instances for performance evaluation in accuracy, sensitivity, specificity, and area under the receiver operating characteristics curve (AUC). Results The proposed system can grade NIHSS scores with an accuracy of 83.3% and an AUC of 0.912 using an optimized ensemble algorithm, and it can grade with an accuracy of 80.0% and an AUC of 0.860 using an optimized SVM algorithm. The auto-MRC grading achieved an accuracy of 76.7% and a mean AUC of 0.870 in SVM classification and an accuracy of 78.3% and a mean AUC of 0.877 in ensemble classification. Conclusions The automatic grading system quantifies proximal weakness in real time and assesses symptoms through automatic grading. The pilot outcomes demonstrated the feasibility of remote monitoring of motor weakness caused by stroke. The system can facilitate consistent grading with instant assessment and expedite dispatches to appropriate hospitals and treatment initiation by sharing auto-MRC and auto-NIHSS scores between prehospital and hospital responses as an objective observation.

Download Full-text

Comparison between C-arm cone beam computed tomography and interventional angiography in transarterial chemoembolization of hepatocellular carcinoma

Egyptian Journal of Radiology and Nuclear Medicine ◽

10.1186/s43055-019-0053-4 ◽

2019 ◽

Vol 50 (1) ◽

Author(s):

Ibraheim Ahmed Diab ◽

Shaimaa Abdel-hamid Hassanein ◽

Hala Hafez Mohamed

Keyword(s):

Computed Tomography ◽

Hepatocellular Carcinoma ◽

Cone Beam Computed Tomography ◽

Vascular Anatomy ◽

Cone Beam ◽

Focal Lesion ◽

Focal Lesions ◽

Conventional Angiography ◽

Interventional Angiography ◽

Cross Sectional

Abstract Background Hepatocellular carcinoma (HCC) is the most common primary hepatic malignancy of adults. One of the established treatment procedures performed worldwide for HCC is transcatheter arterial chemoembolization (TACE). By using conventional angiography in TACE, we can detect and identify the vascular anatomy of the liver through obtaining 2D images. Recently C-arm cone beam computed tomography (CBCT) is introduced for obtaining cross-sectional and three-dimensional (3D) images for better visualization of small tumors and their feeding arteries. Results The number of detected focal lesions by angiography was 51 compared to 87 focal lesion detected by CBCT; of those, 45 and 77 were active lesions by both procedures respectively. For lesions, less than 1 cm CBCT detected 23 lesions while angiography detected only one lesion. Angiography detected 87 feeding arterial branch while cone beam CT-HA detected 130 branches to the same number of target lesion. Feeder tractability and confidence were better by CBCT. Conclusion CBCT is superior to angiography in tumor detectability, detection of lesions less than 1 cm, feeder detection, and feeder traction; however, conventional angiography and DSA are irreplaceable. Thus, combination of CBCT with angiography during TACE produces better results and less complication.

Download Full-text

Abstract TP37: Cervical Vascular Findings and Clinical Correlates in Takayasu Arteritis

Stroke ◽

10.1161/str.47.suppl_1.tp37 ◽

2016 ◽

Vol 47 (suppl_1) ◽

Author(s):

Bree Chancellor ◽

Gibran Shaikh ◽

Adam Davis ◽

Pam Rosenthal ◽

Koto Ishida

Keyword(s):

Takayasu Arteritis ◽

Common Finding ◽

Wall Thickening ◽

Collateral Flow ◽

Conventional Angiography ◽

Neurologic Symptoms ◽

Clinical Correlates ◽

American College Of Rheumatology ◽

Vessel Stenosis ◽

Symptom Exacerbation

Though luminal changes in Takayasu arteritis (TA) are well seen with conventional angiography, mural changes can be best seen with CTA. Cervical vasculature is affected in over 75% of patients. Cervical vessel findings on CTA and clinical correlates have not been fully described. Methods: Thirteen patients with TA were identified by ICD-9 diagnosis code at two urban hospitals. Diagnosis was confirmed based on American College of Rheumatology criteria for TA. Results: Of 4 male, 9 female (mean age, 37; 5 Latin Am.; 3 Asian; 3 African; 2 North Am.) patients, 10 (77%) had dedicated cervical imaging (CTA/MRA). Ten had neurologic symptoms; visual (46%); weakness/numbness (31%); syncope/dizziness (23%). Nine (69%) had active disease at time of imaging. Twelve (92%) had cervical vessel lesions; 11 (85%) with wall thickening; 11 with vessel stenosis. On average 3 vessels were affected, most commonly L subclavian (69%), L common carotid (54%). Of the 6 patients (46%) with occlusions, all had collateral flow; in 4, wall enhancement, intimal vessel hyperplasia was seen. Three patients imaged before aorto-carotid bypass grafting had an average of 7 diseased cervical vessels. All had strokes soon after bypass, two ischemic with hemorrhagic conversion, a third with IPH/IVH. Two patients had cerebrovascular symptom exacerbation during menses; one was successfully treated with hysterectomy. Findings on cervical imaging directly changed surgical or medical management in 9 (77%) cases. Conclusion: Cervical vessel involvement is pervasive in TA. Wall thickening, a common finding in early and active TA, is not part of current diagnostic criteria and may warrant inclusion. Given the prevalence of cervical vessel disease and its clinical implications, cervical vascular imaging should be considered in all TA patients, particularly those with neurologic symptoms. Combining chest/cervical CTA into a single protocol may be beneficial.

Download Full-text

Abstract 2903: Nutritional Access of Dysphagic Patients 3 Months after Onset of Acute Stroke -From Database of 5,000 Patients in a Single Stroke Center-

Stroke ◽

10.1161/str.43.suppl_1.a2903 ◽

2012 ◽

Vol 43 (suppl_1) ◽

Author(s):

Makoto Nakajima ◽

Yuichiro Inatomi ◽

Toshiro Yonehara ◽

Yoichiro Hashimoto ◽

Teruyuki Hirano

Keyword(s):

Acute Stroke ◽

Chronic Phase ◽

Oral Intake ◽

Continuous Variables ◽

Operating Characteristics ◽

Stroke Patients ◽

Nihss Score ◽

Swallowing Dysfunction ◽

Stroke Center ◽

Dysphagic Patients

Background and purpose: Prediction of swallowing function in dysphagic patients with acute stroke is indispensable for discussing percutaneous endoscopic gastrostomy (PEG) placement. We performed a retrospective study using database of a large number of acute ischemic stroke patients to clarify predictors for acquisition of oral intake in chronic phase. Methods: A total 4,972 consecutive acute stroke patients were admitted to our stroke center during 8.5 years; a questionnaire was sent to all the survivors after 3 months of onset. We investigated nutritional access after 3 months of onset in 588 patients who could not eat orally 10 days after admission, and analyzed predictive factors for their acquisition of oral intake. Continuous variables were dichotomized to identify the most sensitive predictors; the cutoff values were investigated by receiver operating characteristics curve analysis. Results: Out of 588 dysphagic patients, 75 died during the 3 months, and 143 (28%) of the residual 513 achieved oral intake after 3 months. In logistic-regression models, age ≤80 years, absence of hyperlipidemia, absence of atrial fibrillation, modified Rankin Scale score 0 before onset, and low National Institutes of Health Stroke Scale (NIHSS) score independently predicted oral intake 3 months after onset. From two different model analyses, NIHSS score ≤17 on day 10 (OR 3.63, 95% CI 2.37-5.56) was found to be a stronger predictor for oral intake than NIHSS score ≤17 on admission (OR 2.34, 95% CI 1.52-3.59). At 3 months, 17/143 (12%) patients with oral intake were living at home, while only 1/370 (0.3%) patients without oral intake were. Conclusion: A quarter of dysphagic patients with acute stroke obtained oral intake 3 months after onset. Clinicians should be cautious about PEG placement for stroke patients with severe dysphagia who were independent prior to the stroke, aged ≤80 years, and show NIHSS score ≤17 on day 10, because their swallowing dysfunction may improve in a few months.

Download Full-text

Hybrid Integration Approach of Entropy with Logistic Regression and Support Vector Machine for Landslide Susceptibility Modeling

Entropy ◽

10.3390/e20110884 ◽

2018 ◽

Vol 20 (11) ◽

pp. 884 ◽

Cited By ~ 28

Author(s):

Tingyu Zhang ◽

Ling Han ◽

Wei Chen ◽

Himan Shahabi

Keyword(s):

Support Vector Machine ◽

Logistic Regression ◽

Landslide Susceptibility ◽

Shaanxi Province ◽

Slope Aspect ◽

Support Vector ◽

Operating Characteristics ◽

Study Region ◽

Susceptibility Maps ◽

Landslide Susceptibility Maps

The main purpose of the present study is to apply three classification models, namely, the index of entropy (IOE) model, the logistic regression (LR) model, and the support vector machine (SVM) model by radial basis function (RBF), to produce landslide susceptibility maps for the Fugu County of Shaanxi Province, China. Firstly, landslide locations were extracted from field investigation and aerial photographs, and a total of 194 landslide polygons were transformed into points to produce a landslide inventory map. Secondly, the landslide points were randomly split into two groups (70/30) for training and validation purposes, respectively. Then, 10 landslide explanatory variables, such as slope aspect, slope angle, altitude, lithology, mean annual precipitation, distance to roads, distance to rivers, distance to faults, land use, and normalized difference vegetation index (NDVI), were selected and the potential multicollinearity problems between these factors were detected by the Pearson Correlation Coefficient (PCC), the variance inflation factor (VIF), and tolerance (TOL). Subsequently, the landslide susceptibility maps for the study region were obtained using the IOE model, the LR–IOE, and the SVM–IOE model. Finally, the performance of these three models was verified and compared using the receiver operating characteristics (ROC) curve. The success rate results showed that the LR–IOE model has the highest accuracy (90.11%), followed by the IOE model (87.43%) and the SVM–IOE model (86.53%). Similarly, the AUC values also showed that the prediction accuracy expresses a similar result, with the LR–IOE model having the highest accuracy (81.84%), followed by the IOE model (76.86%) and the SVM–IOE model (76.61%). Thus, the landslide susceptibility map (LSM) for the study region can provide an effective reference for the Fugu County government to properly address land planning and mitigate landslide risk.

Download Full-text

A Machine Learning Based Approach of Performance Estimation for High-Pressure Compressor Airfoils

Volume 2D: Turbomachinery ◽

10.1115/gt2018-75372 ◽

2018 ◽

Author(s):

Jonas Marx ◽

Stefan Gantner ◽

Jörn Städing ◽

Jens Friedrichs

Keyword(s):

Machine Learning ◽

High Pressure ◽

Machine Learning Algorithms ◽

Performance Estimation ◽

Predictive Maintenance ◽

Support Vector ◽

K Nearest Neighbor ◽

Operating Characteristics ◽

High Pressure Compressor ◽

Pressure Compressor

In recent years, the demands of Maintenance, Repair and Overhaul (MRO) customers to provide resource-efficient after market services have grown increasingly. One way to meet these requirements is by making use of predictive maintenance methods. These are ideas that involve the derivation of workscoping guidance by assessing and processing previously unused or undocumented service data. In this context a novel approach on predictive maintenance is presented in form of a performance-based classification method for high pressure compressor (HPC) airfoils. The procedure features machine learning algorithms that establish a relation between the airfoil geometry and the associated aerodynamic behavior and is hereby able to divide individual operating characteristics into a finite number of distinct aero-classes. By this means the introduced method not only provides a fast and simple way to assess piece part performance through geometrical data, but also facilitates the consideration of stage matching (axial as well as circumferential) in a simplified manner. It thus serves as prerequisite for an improved customary HPC performance workscope as well as for an automated optimization process for compressor buildup with used or repaired material that would be applicable in an MRO environment. The methods of machine learning that are used in the present work enable the formation of distinct groups of similar aero-performance by unsupervised (step 1) and supervised learning (step 2). The application of the overall classification procedure is shown exemplary on an artificially generated dataset based on real characteristics of a front and a rear rotor of a 10-stage axial compressor that contains both geometry as well as aerodynamic information. In step 1 of the investigation only the aerodynamic quantities in terms of multivariate functional data are used in order to benchmark different clustering algorithms and generate a foundation for a geometry-based aero-classification. Corresponding classifiers are created in step 2 by means of both, the k Nearest Neighbor and the linear Support Vector Machine algorithms. The methods’ fidelities are brought to the test with the attempt to recover the aero-based similarity classes solely by using normalized and reduced geometry data. This results in high classification probabilities of up to 96 % which is proven by using stratified k-fold cross-validation.

Download Full-text

Optical Flow Approach Followed by SVM Classification Model to Recognize Abnormal Behavior of a Crowd

International Journal of Engineering & Technology ◽

10.14419/ijet.v7i3.34.18927 ◽

2018 ◽

Vol 7 (3.34) ◽

pp. 156

Author(s):

Basavaraj G.M ◽

Dr Ashok Kusagur

Keyword(s):

Optical Flow ◽

Promising Result ◽

Classification Model ◽

Abnormal Behavior ◽

Support Vector ◽

Force Model ◽

Feature Points ◽

Crowd Behavior ◽

Operating Characteristics ◽

Harris Corner

A many of researches have been carried out in the field of the crowd behavior recognition system. Recognizing crowd behavior in videos is most challenging and occlusions because of irregular human movement. This paper gives an overview of optical flow model along with the SVM (Support Vector Machine) classification model. This proposed approach evaluates sudden changes in motion of an event and classifies that event to a category: Normal and Abnormal. Geometric means of location, direction, and displacement of the feature points of each frame are estimated. Harris corner Detector is used in each frame for tracking a set of feature points. Proposed approach is very effective in real time scenario like public places where security is most important. After analyzing result ROC curve (receiver operating characteristics) is plotted which gives classification accuracy. We also presented frame level comparison with Ground truth and social force model (SFM) techniques. Our proposed approach is giving a promising result compare to all state of art methods.

Download Full-text

Analyzing Behavior of Cancer Patients using Machine Learning Techniques

International Journal of Innovative Technology and Exploring Engineering - Special Issue ◽

10.35940/ijitee.i8414.078919 ◽

2019 ◽

Vol 8 (9) ◽

pp. 1547-1556

Keyword(s):

Machine Learning ◽

Natural Language ◽

Cancer Patients ◽

Language Processing ◽

Machine Learning Techniques ◽

Support Vector ◽

Svm Classifier ◽

Operating Characteristics ◽

Decision Tree Classifier ◽

Tree Classifier

The online discussion forums and blogs are very vibrant platforms for cancer patients to express their views in the form of stories. These stories sometimes become a source of inspiration for some patients who are anxious in searching the similar cases. This paper proposes a method using natural language processing and machine learning to analyze unstructured texts accumulated from patient’s reviews and stories. The proposed methodology aims to identify behavior, emotions, side-effects, decisions and demographics associated with the cancer victims. The pre-processing phase of our work involves extraction of web text followed by text-cleaning where some special characters and symbols are omitted, and finally tagging the texts using NLTK’s (Natural Language Toolkit) POS (Parts of Speech) Tagger. The post-processing phase performs training of seven machine learning classifiers (refer Table 6). The Decision Tree classifier shows the higher precision (0.83) among the other classifiers while, the Area under the operating Characteristics (AUC) for Support Vector Machine (SVM) classifier is highest (0.98).

Download Full-text