A non-small cell lung cancer medical data mining method
By constructing a single-cell transcriptome data and multilayer perceptron model based on peripheral blood mononuclear cells, the problems of non-invasiveness, accuracy, and robustness in predicting the efficacy of immune checkpoint inhibitors in non-small cell lung cancer were solved, providing an efficient efficacy assessment tool.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- MOBIDROP (ZHEJIANG) CO LTD
- Filing Date
- 2024-09-11
- Publication Date
- 2026-06-02
AI Technical Summary
Existing methods for predicting the efficacy of immune checkpoint inhibitors in non-small cell lung cancer based on ctDNA levels suffer from challenges in detection, insufficient accuracy, limited information, and the inability to perform completely non-invasive examinations. Furthermore, there is a lack of robust prediction models that are easy to use.
By acquiring single-cell transcriptome data of peripheral blood mononuclear cells from non-small cell lung cancer patients before medication, a T-cell expression profile matrix was constructed, and a multilayer perceptron model was used for prediction. Combined with machine learning technology, an easy-to-use one-click prediction model was established, and efficacy scoring was performed based on single-cell data.
It achieves non-invasive testing, improves the accuracy and robustness of efficacy prediction, avoids invasive tissue sampling of patients, and provides an efficient efficacy prediction tool.
Smart Images
Figure CN119560017B_ABST