A bone metastasis diagnosis and prognosis evaluation system based on multi-modal deep learning
By integrating radionuclide bone scintigraphy and CT images through a multimodal deep learning system, accurate detection and classification of bone metastases are achieved. Combined with clinical data, prognostic assessment is performed, which solves the problems of insufficient diagnostic accuracy and lack of prognostic assessment in existing technologies, and provides personalized prognostic support and quantitative survival prediction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ZHEJIANG CANCER HOSPITAL
- Filing Date
- 2026-06-15
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for diagnosing bone metastases suffer from limitations in single-modality diagnostic accuracy, lack of prognostic assessment methods, reliance on human experience for lesion localization, imprecise lesion segmentation and classification, and a disconnect between diagnosis and prognosis.
A bone metastasis diagnosis and prognostic assessment system based on multimodal deep learning is adopted. Through data acquisition and preprocessing modules, bone anatomy partitioning modules, dual-stream feature extraction and fusion modules, and diagnostic modules, the system integrates the functional information of radionuclide bone scintigraphy and the anatomical information of CT images to achieve accurate detection, segmentation and classification of lesions, and combines clinical data for prognostic prediction.
It improves the accuracy of bone metastasis diagnosis, reduces the false positive rate, provides individualized prognostic assessment, supports clinical decision-making, realizes data integration from imaging to prognosis, and outputs quantitative survival prediction results.
Smart Images
Figure CN122415600A_ABST