biostudies-literatureDu YThe Guangxi Promotion of Appropriate Health Technologies ProjectThe Natural Science Foundation of Guangxi Zhuang Autonomous RegionThe Guangxi Medical"139" Project for Training High-level Backbone TalentsThe Guangxi Medical”139” Project for Training High-level Backbone Talents89https://www.ebi.ac.uk/biostudies/studies/S-EPMC11020982biostudies-literatureUnknown24(1)<h4>Background</h4>Accurate preoperative identification of ovarian tumour subtypes is imperative for patients as it enables physicians to custom-tailor precise and individualized management strategies. So, we have developed an ultrasound (US)-based multiclass prediction algorithm for differentiating between benign, borderline, and malignant ovarian tumours.<h4>Methods</h4>We randomised data from 849 patients with ovarian tumours into training and testing sets in a ratio of 8:2. The regions of interest on the US images were segmented and handcrafted radiomics features were extracted and screened. We applied the one-versus-rest method in multiclass classification. We inputted the best features into machine learning (ML) models and constructed a radiomic signature (Rad_Sig). US images of the BMC medical imagingUltrasound-based deep learning radiomics model for differentiating benign, borderline, and malignant ovarian tumours: a multi-class classification exploratory study.PMC11020982G201903014S20210552023GXNSFDA026010Wu JYao JGuo WXiao YDu YChen HfalseUltrasound-based deep learning radiomics model for differentiating benign, borderline, and malignant ovarian tumours: a multi-class classification exploratory study.<h4>Background</h4>Accurate preoperative identification of ovarian tumour subtypes is imperative for patients as it enables physicians to custom-tailor precise and individualized management strategies. So, we have developed an ultrasound (US)-based multiclass prediction algorithm for differentiating between benign, borderline, and malignant ovarian tumours.<h4>Methods</h4>We randomised data from 849 patients with ovarian tumours into training and testing sets in a ratio of 8:2. The regions of interest on the US images were segmented and handcrafted radiomics features were extracted and screened. We applied the one-versus-rest method in multiclass classification. We inputted the best features into machine learning (ML) models and constructed a radiomic signature (Rad_Sig). US images of the 2024-01-01T00:00:00Z2024 Apr2025-04-04T12:58:20.227Z2025-04-04T12:58:20.227ZS-EPMC110209823862254610.1186/s12880-024-01251-2