一种用于识别肝细胞癌的预测模型及其构建方法和应用
By constructing a support vector machine algorithm model GT2.0 based on 12 specific oligosaccharide chains and clinical indicators, the shortcomings of traditional serological markers in the early diagnosis of hepatocellular carcinoma were overcome, and a highly sensitive non-invasive diagnosis of very early-stage hepatocellular carcinoma was achieved, especially with a high detection rate in AFP/DCP-negative patients.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- JIANGSU XIANSIDA BIOTECH CO LTD
- Filing Date
- 2026-05-06
- Publication Date
- 2026-07-17
AI Technical Summary
In the current technology, traditional serological markers such as AFP and DCP have insufficient sensitivity in the early diagnosis of hepatocellular carcinoma, especially in very early-stage hepatocellular carcinoma and AFP/DCP negative patients, which limits their diagnostic efficacy and cannot meet clinical needs.
By acquiring data on the abundance, gender, and age of 12 specific oligosaccharide chains from the training sample set, a prediction model (GlycanTest 2.0, GT2.0) was constructed using the Bootstrap-integrated support vector machine algorithm to achieve high-precision prediction of hepatocellular carcinoma.
The GT2.0 model achieved a diagnostic sensitivity of 74.77% in very early-stage hepatocellular carcinoma, significantly improving the detection rate of early-stage hepatocellular carcinoma. In particular, its sensitivity exceeded 83% in AFP/DCP-negative patients, providing a highly sensitive and specific non-invasive diagnostic method that fills the diagnostic gap of existing biomarkers.
Smart Images
Figure CN122135792B_ABST