Search

This paper presents an unsupervised clustering random-forest-based metric for affinity estimation in large and high-dimensional data. The criterion used for node splitting during forest construction can handle rank-deficiency when measuring cluster compactness. The binary forest-based metric is extended to continuous metrics by exploiting both the common traversal path and the smallest shared parent node.The proposed forest-based metric efficiently estimates affinity by passing down data pairs in the forest using a limited number of decision trees. A pseudo-leaf-splitting (PLS) algorithm is introduced to account for spatial relationships, which regularizes affinity measures and overcomes inconsistent leaf assign-ments. The random-forest-based metric with PLS facilitates the establishment of consistent and point-wise correspondences. The proposed method has been applied to automatic phrase recognition using color and depth videos and point-wise correspondence. Extensive experiments demonstrate the effectiveness of the proposed method in affinity estimation in a comparison with the state-of-the-art.

Liya Wang,1 Shenjiao Huang,2 Xiujie Sheng,3 Chenchen Ren,4 Qiming Wang,5 Linqing Yang,6 Shuping Zhao,7 Tianmin Xu,8 Xiaoxin Ma,9 Ruixia Guo,10 Pengming Sun,11 Yang Lin,8 Yuhong Li,1 Jiandong Wang,12 Yudong Wang1 1Department of Gynecologic Oncology, International Peace Maternity and Child Health Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai Municipal Key Clinical Specialty, Shanghai Key Laboratory of Embryo Original Disease, Shanghai, 200030, People’s Republic of China; 2Department of Obstetrics and Gynecology, Guangzhou Women and Children’s Medical Centre, Guangzhou 510623, People’s Republic of China; 3Department of Obstetrics and Gynecology, The Third Affiliated Hospital of Guangzhou Medical University, Key Laboratory of Major Obstetric Diseases of Guangdong Province, Guangzhou 510150, People’s Republic of China; 4Department of Obstetrics and Gynecology, The Third Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, People’s Republic of China; 5Department of Obstetrics and Gynecology, Women and Children’s Hospital of Ningbo City, Ningbo 315012, People’s Republic of China; 6Department of Gynecology, Affiliated Hospital of Jining Medical University, Jining 272029, People’s Republic of China; 7Department of Gynecology, Women and Children’s Hospital of Qingdao, Qingdao 266000, People’s Republic of China; 8Department of Obstetrics and Gynecology, The Second Hospital of Jilin University, Changchun 130000, People’s Republic of China; 9Department of Obstetrics and Gynecology, Shengjing Hospital Affiliated of China Medical University, Shenyang 110004, People’s Republic of China; 10Department of Gynecology, The First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, People’s Republic of China; 11Department of Obstetrics and Gynecology, Fujian Maternity and Child Health Hospital, Affiliated Hospital of Fujian Medical University, Fuzhou 350001, People’s Republic of China; 12Department of Gynecologic Oncology, Beijing Obstetrics and Gynecology Hospital, Capital Medical University, Beijing 100026, People’s Republic of ChinaCorrespondence: Yudong Wang Department of Gynecologic OncologyInternational Peace Maternity and Child Health Hospital, Shanghai Jiao Tong University School of Medicine, NO. 1961, Huashan Road, Shanghai 200030, People’s Republic of ChinaEmail gfywangyudong@126.comJiandong WangDepartment of Gynecologic Oncology, Beijing Obstetrics and Gynecology Hospital, Capital Medical University, NO. 251, Yao Jiayuan Road, Beijing 100026, People’s Republic of ChinaEmail wangjiandongxy@hotmail.comPurpose: The aim of this study was to investigate the clinical characteristics and management of malignant ovarian tumors during pregnancy, as well as the feto-maternal outcomes and analyze the influentialfactorson thepregnancyoutcomes.Patients and Methods: Eighty-five patients with ovarian malignancies during pregnancy treated at 12tertiary hospitals between 2009 and 2019 were analyzed in this study. The clinical features, histopathological characteristics, clinical management, and maternal and perinatal outcomes were retrospectively analyzed. The clinical features and managements were compared between abortion group and live birth group.Results: The following diagnoses were made: 41 (48.24%) patients with borderline ovarian tumors, 18 (21.18%) patients with epithelial ovarian cancers, 17 (20.00%) patients with non-epithelial ovarian malignancies and 9 (10.59%) patients with metastatic ovarian tumors. Thirty-six (42.45%) patients underwent conservative surgical treatment. Thirty-four (40.00%)patients opted for fertility-sparing surgery, and fifteen (17.56%)patients received radical surgery. Chemotherapy was administered to 32.94% of the patients. The proportion of ovarian malignancies diagnosed in the first trimester in the abortion group was higher than that in the live birth group (P< 0.05). However, tumor diameter, reproductive history, stage and surgical indications showed no significant differences between groups. A total of 67 live babies were recorded in this study, including 19 premature babies and 1 full-term newborn who died of respiratory distress. All of the BOTs were diagnosed with stage I, among whom 38 (92.68%) patients exhibited disease-free survival. Twenty-eight ovarian cancers were in stage I–II and 26 of them had disease-free survival with the longest follow-up time of 10 years. Five of the sixteen patients in advanced stage (stage III–IV) died, four of whom had metastatic tumors.Conclusion: Pregnant women with early-stage malignant ovarian tumors appear to have favorable outcomes. Conservative surgery is acceptable for early-stage borderline ovarian tumors during pregnancy. The gestational age of ovarian malignancy detection is key for pregnancy outcomes.Keywords: malignant ovarian tumor, pregnancy, management, prognosis

Additional file 1: Table S1. There are 34 coordinates of the landmarks entering into the discriminant equations. Each of them are represented as variables Xn in the equations.