Interpreting SVM for medical images using Quadtree

Shukla, Prashant Kumar; Verma, Abhishek; Abhishek, Abhishek; Verma, Shekhar; Kumar, Manish

doi:10.1007/s11042-020-09431-2

Cited by 24 publications

(10 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Further, we established and validated a stable ML model based on these features to predict edema after EVT in acute stroke patients. The 7 ML models in our study have been proven to have good prediction or classification performance in the medical imaging ( Erickson et al, 2017 ; Kim et al, 2020 ; Shukla et al, 2020 ). We demonstrated that the Bayes model based on DWI + CSF FLAIR and the RF model based on DWI + CSF DWI had the highest AUC value and the most stability, respectively.…”

Section: Discussionmentioning

confidence: 85%

MRI Radiomics Features From Infarction and Cerebrospinal Fluid for Prediction of Cerebral Edema After Acute Ischemic Stroke

Jiang

Zhang

Wang

et al. 2022

Front. Aging Neurosci.

View full text Add to dashboard Cite

Neuroimaging biomarkers that predict the edema after acute stroke may help clinicians provide targeted therapies and minimize the risk of secondary injury. In this study, we applied pretherapy MRI radiomics features from infarction and cerebrospinal fluid (CSF) to predict edema after acute ischemic stroke. MRI data were obtained from a prospective, endovascular thrombectomy (EVT) cohort that included 389 patients with acute stroke from two centers (dataset 1, n = 292; dataset 2, n = 97), respectively. Patients were divided into edema group (brain swelling and midline shift) and non-edema group according to CT within 36 h after therapy. We extracted the imaging features of infarct area on diffusion weighted imaging (DWI) (abbreviated as DWI), CSF on fluid-attenuated inversion recovery (FLAIR) (CSFFLAIR) and CSF on DWI (CSFDWI), and selected the optimum features associated with edema for developing models in two forms of feature sets (DWI + CSFFLAIR and DWI + CSFDWI) respectively. We developed seven ML models based on dataset 1 and identified the most stable model. External validations (dataset 2) of the developed stable model were performed. Prediction model performance was assessed using the area under the receiver operating characteristic curve (AUC). The Bayes model based on DWI + CSFFLAIR and the RF model based on DWI + CSFDWI had the best performances (DWI + CSFFLAIR: AUC, 0.86; accuracy, 0.85; recall, 0.88; DWI + CSFDWI: AUC, 0.86; accuracy, 0.84; recall, 0.84) and the most stability (RSD% in DWI + CSFFLAIR AUC: 0.07, RSD% in DWI + CSFDWI AUC: 0.09), respectively. External validation showed that the AUC of the Bayes model based on DWI + CSFFLAIR was 0.84 with accuracy of 0.77 and area under precision-recall curve (auPRC) of 0.75, and the AUC of the RF model based on DWI + CSFDWI was 0.83 with accuracy of 0.81 and the auPRC of 0.76. The MRI radiomics features from infarction and CSF may offer an effective imaging biomarker for predicting edema.

show abstract

Section: Discussionmentioning

confidence: 85%

MRI Radiomics Features From Infarction and Cerebrospinal Fluid for Prediction of Cerebral Edema After Acute Ischemic Stroke

Jiang

Zhang

Wang

et al. 2022

Front. Aging Neurosci.

View full text Add to dashboard Cite

show abstract

“…Deep learning (DL) models, according to recent research, can learn from irrelevant data and make decisions based on that data [ 43 , 44 , 45 , 46 ], even though high-performing networks’ performances cannot be extrapolated to real-world applications. In contrast to ordinary X-rays, the segmented lungs helped the CNN model to identify the primary Region of Interest (ROI).…”

Section: Discussionmentioning

confidence: 99%

Computer-Aided Diagnosis of COVID-19 from Chest X-ray Images Using Hybrid-Features and Random Forest Classifier

et al. 2023

View full text Add to dashboard Cite

In recent years, a lot of attention has been paid to using radiology imaging to automatically find COVID-19. (1) Background: There are now a number of computer-aided diagnostic schemes that help radiologists and doctors perform diagnostic COVID-19 tests quickly, accurately, and consistently. (2) Methods: Using chest X-ray images, this study proposed a cutting-edge scheme for the automatic recognition of COVID-19 and pneumonia. First, a pre-processing method based on a Gaussian filter and logarithmic operator is applied to input chest X-ray (CXR) images to improve the poor-quality images by enhancing the contrast, reducing the noise, and smoothing the image. Second, robust features are extracted from each enhanced chest X-ray image using a Convolutional Neural Network (CNNs) transformer and an optimal collection of grey-level co-occurrence matrices (GLCM) that contain features such as contrast, correlation, entropy, and energy. Finally, based on extracted features from input images, a random forest machine learning classifier is used to classify images into three classes, such as COVID-19, pneumonia, or normal. The predicted output from the model is combined with Gradient-weighted Class Activation Mapping (Grad-CAM) visualisation for diagnosis. (3) Results: Our work is evaluated using public datasets with three different train–test splits (70–30%, 80–20%, and 90–10%) and achieved an average accuracy, F1 score, recall, and precision of 97%, 96%, 96%, and 96%, respectively. A comparative study shows that our proposed method outperforms existing and similar work. The proposed approach can be utilised to screen COVID-19-infected patients effectively. (4) Conclusions: A comparative study with the existing methods is also performed. For performance evaluation, metrics such as accuracy, sensitivity, and F1-measure are calculated. The performance of the proposed method is better than that of the existing methodologies, and it can thus be used for the effective diagnosis of the disease.

show abstract

“…As a result, the Radial Sigma SVM model performed better than other machine learning models. Through linear or nonlinear kernel functions (Radial SVM), SVM is a supervised learning classification algorithm that constructs a hyperplane in a higher dimensional space [ 69 ]. Across all tasks, we found that overall prediction accuracy was high.…”

Section: Discussionmentioning

confidence: 99%

An Exploration of Pepino (Solanum muricatum) Flavor Compounds Using Machine Learning Combined with Metabolomics and Sensory Evaluation

Sun

Zhao

et al. 2022

Foods

View full text Add to dashboard Cite

Flavor is one of the most important characteristics that directly determines the popularity of a food. Moreover, the flavor of fruits is determined by the interaction of multiple metabolic components. Pepino, an emerging horticultural crop, is popular for its unique melon-like flavor. We analyzed metabolomics data from three different pepino growing regions in Haidong, Wuwei, and Jiuquan and counted the status of sweetness, acidity, flavor, and overall liking ratings of pepino fruit in these three regions by sensory panels. The metabolomics and flavor ratings were also integrated and analyzed using statistical and machine learning models, which in turn predicted the sensory panel ratings of consumers based on the chemical composition of the fruit. The results showed that pepino fruit produced in the Jiuquan region received the highest ratings in sweetness, flavor intensity, and liking, and the results with the highest contribution based on sensory evaluation showed that nucleotides and derivatives, phenolic acids, amino acids and derivatives, saccharides, and alcohols were rated in sweetness (74.40%), acidity (51.57%), flavor (56.41%), and likability (33.73%) dominated. We employed 14 machine learning strategies trained on the discovery samples to accurately predict the outcome of sweetness, sourness, flavor, and liking in the replication samples. The Radial Sigma SVM model predicted with better accuracy than the other machine learning models. Then we used the machine learning models to determine which metabolites influenced both pepino flavor and consumer preference. A total of 27 metabolites most important for pepino flavor attributes to distinguish pepino originating from three regions were screened. Substances such as N-acetylhistamine, arginine, and caffeic acid can enhance pepino‘s flavor intensity, and metabolites such as glycerol 3-phosphate, aconitic acid, and sucrose all acted as important variables in explaining the liking preference. While glycolic acid and orthophosphate inhibit sweetness and enhance sourness, sucrose has the opposite effect. Machine learning can identify the types of metabolites that influence fruit flavor by linking metabolomics of fruit with sensory evaluation among consumers, which conduces breeders to incorporate fruit flavor as a trait earlier in the breeding process, making it possible to select and release fruit with more flavor.

show abstract

Interpreting SVM for medical images using Quadtree

Cited by 24 publications

References 32 publications

MRI Radiomics Features From Infarction and Cerebrospinal Fluid for Prediction of Cerebral Edema After Acute Ischemic Stroke

MRI Radiomics Features From Infarction and Cerebrospinal Fluid for Prediction of Cerebral Edema After Acute Ischemic Stroke

Computer-Aided Diagnosis of COVID-19 from Chest X-ray Images Using Hybrid-Features and Random Forest Classifier

An Exploration of Pepino (Solanum muricatum) Flavor Compounds Using Machine Learning Combined with Metabolomics and Sensory Evaluation

Contact Info

Product

Resources

About