We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Prognosis Evaluation of Ovarian Granulosa Cell Tumor Based on Co-forest Intelligence Model.
- Authors
Xin Liao; Xin Zheng; Juan Zou; Min Feng; Liang Sun; Yan Li; Kaixuan Yang
- Abstract
Ovarian granulosa cell tumor (GCT) has different recurrence periods, which dramatically decreases after the 5-year survival period. Prognosis evaluation has important clinical values and is a research hotspot. Prognosis evaluation methods include logistic regression, Chi-square analysis, and other traditional statistical methods; however, these techniques cannot solve problems, such as limited samples and ambiguous prognosis-related pathologic features, and have poor reliability and validity of assessment results. In this study, an artificial intelligence theory was introduced, and the prognosis evaluation of ovarian GCT based on co-forest intelligence model was proposed to find a method applicable to the pathological data of ovarian GCT with limited samples and ambiguous prognosis features. First, data preprocessing of ovarian GCT samples was performed. This procedure included deleting unqualified data and standardizing and normalizing data. Second, prognosis evaluation of ovarian GCT was accomplished by using co-forest intelligence algorithm. Finally, the validity of the proposed prognosis evaluation method was verified by 75 patients with ovarian GCT in the West China Second Hospital of Sichuan University. Results indicate that: (1) the accuracy of prognosis evaluation based on the feature set selected by Log-Rank test increases by 12.1% compared with that (4.1%) based on the direct use of standardized and normalized feature set, and (2) the co-forest algorithm can be used for the model analysis of small pathological datasets of ovarian GCT. Moreover, this method can be used to explore effective characteristics from the candidate feature dataset through automatic learning with prediction accuracy of up to 95.7%. This study reveals the reliability and effectiveness of the proposed prognosis evaluation method of ovarian GCT based on co-forest intelligence model. Conclusions are beneficial for clinicians to accurately understand the development laws of ovarian GCT, take the initiative to master the diagnosis and treatment, and increase the long-term survival rate of patients.
- Subjects
GRANULOSA cell tumors; PROGNOSIS; LOGISTIC regression analysis; ARTIFICIAL intelligence; PATHOLOGICAL anatomy
- Publication
Journal of Engineering Science & Technology Review, 2018, Vol 11, Issue 2, p135
- ISSN
1791-2377
- Publication type
Article
- DOI
10.25103/jestr.112.19