An improved joint non-negative matrix factorization for identifying surgical treatment timing of neonatal necrotizing enterocolitis

Guoqiang Qi; Shoujiang Huang; Dengming Lai; Jing Li; Yonggen Zhao; Chen Shen; Jian Huang; Tianmei Liu; Kai Wei; Jinfa Tou; Qiang Shu; Gang Yu

doi:10.17305/bjbms.2022.7046

Authors

Guoqiang Qi Department of Data and Information, The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, China; Sino-Finland Joint AI Laboratory for Child Health of Zhejiang Province, Hangzhou, China; National Clinical Research Center for Child Health, Hangzhou, China. https://orcid.org/0000-0002-7863-6223
Shoujiang Huang National Clinical Research Center for Child Health, Hangzhou, China; Department of Neonatal Surgery, The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, China.
Dengming Lai National Clinical Research Center for Child Health, Hangzhou, China; Department of Neonatal Surgery, The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, China.
Jing Li Department of Data and Information, The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, China; Sino-Finland Joint AI Laboratory for Child Health of Zhejiang Province, Hangzhou, China; National Clinical Research Center for Child Health, Hangzhou, China. https://orcid.org/0000-0002-3626-5815
Yonggen Zhao Department of Data and Information, The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, China; Sino-Finland Joint AI Laboratory for Child Health of Zhejiang Province, Hangzhou, China; National Clinical Research Center for Child Health, Hangzhou, China.
Chen Shen Department of Data and Information, The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, China; Sino-Finland Joint AI Laboratory for Child Health of Zhejiang Province, Hangzhou, China; National Clinical Research Center for Child Health, Hangzhou, China.
Jian Huang Department of Data and Information, The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, China; Sino-Finland Joint AI Laboratory for Child Health of Zhejiang Province, Hangzhou, China; National Clinical Research Center for Child Health, Hangzhou, China. https://orcid.org/0000-0002-1955-4316
Tianmei Liu National Clinical Research Center for Child Health, Hangzhou, China; Department of Radiology, The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, China.
Kai Wei Department of Electronic Engineering, College of Information Engineering, Shanghai Maritime University, Shanghai, China
Jinfa Tou National Clinical Research Center for Child Health, Hangzhou, China; Department of Neonatal Surgery, The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, China.
Qiang Shu Department of Data and Information, The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, China; National Clinical Research Center for Child Health, Hangzhou, China.
Gang Yu Department of Data and Information, The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, China; Sino-Finland Joint AI Laboratory for Child Health of Zhejiang Province, Hangzhou, China; National Clinical Research Center for Child Health, Hangzhou, China. https://orcid.org/0000-0001-9935-9969

DOI:

https://doi.org/10.17305/bjbms.2022.7046

Keywords:

Neonatal necrotizing enterocolitis, joint non-negative matrix factorization, surgical indications, multimodal clinical data

Abstract

Neonatal necrotizing enterocolitis is a severe neonatal intestinal disease. Timely identification of surgical indications is essential for newborns in order to seek the best time for treatment and improve prognosis. This paper attempts to establish an algorithm model based on multimodal clinical data to determine the features of surgical indications and construct an auxiliary diagnosis model. The proposed algorithm adds hypergraph constraints on the two modal data based on Joint Nonnegative Matrix Factorization (JNMF), aiming to mine the higher-order correlations of the two data features. In addition, the adjacency matrix of the two kinds of data is used as a network regularization constraint to prevent overfitting. Orthogonal and L1-norm regulations were introduced to avoid feature redundancy and perform feature selection, respectively, and confirmed 14 clinical features. Finally, we used three classifiers, random forest, support vector machine, and logistic regression, to perform binary classification of patients requiring surgery. The results show that when the features selected by the proposed algorithm model are classified by random forest, the area under the ROC curve is 0.8, which has high prediction accuracy.