Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – New research on Support Vector Machine s is the subject of a report. According to news originating from Zhengzhou, Peop le’s Republic of China, by NewsRx correspondents, research stated, “The aim of t he study was to establish an ultrasonographic radiomics machine learning model b ased on endobronchial ultrasound (EBUS) to assist in diagnosing benign and malig nant mediastinal and hilar lymph nodes (LNs). The clinical and ultrasonographic image data of 197 patients were retrospectively analyzed.” Our news journalists obtained a quote from the research from Zhengzhou Universit y People’s Hospital, “The radiomics features extracted by EBUS-based radiomics w ere analyzed by the least absolute shrinkage and selection operator. Then, we us ed a support vector machine (SVM) algorithm to establish an EBUSbased radiomics model. A total of 205 lesions were randomly divided into training (n = 143) and validation (n = 62) groups. The diagnostic efficiency was evaluated by receiver operating characteristic (ROC) curve analysis. A total of 13 stable radiomics f eatures with non-zero coefficients were selected. The SVM model exhibited promis ing performance in both groups. In the training group, the SVM model achieved an ROC area under the curve (AUC) of 0.892 (95% CI: 0.885-0.899), wi th an accuracy of 85.3%, sensitivity of 93.2%, and spe cificity of 79.8%. In the validation group, the SVM model had an RO C AUC of 0.906 (95% CI: 0.890-0.923), an accuracy of 74.2% , a sensitivity of 70.3%, and a specificity of 74.1%. The EBUS-based radiomics model can be used to differentiate mediastinal and hila r benign and malignant LNs.”
NETL
NSTL