A time-domain diffuse optical tomography reconstruction method and system
By employing a partitioning strategy and a quasi-Newton method to optimize the model in time-domain diffuse optical tomography, the optical parameters of shallow and deep tissues are reconstructed using logarithmic integral intensity and mean time of flight, respectively. This solves the problems of reconstruction instability and high computational complexity caused by single feature data, and achieves efficient and reliable optical parameter reconstruction.
Patent Information
- Application Number
- CN202610866432.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-16
- Publication Date
- 2026-07-17
AI Technical Summary
When using integral transform data for optical reconstruction, time-domain diffuse optical tomography often uses a single feature data type, resulting in poor reliability and stability of the reconstruction, high computational complexity, and difficulty in simultaneously improving spatial computational accuracy and deep spatial sensitivity.
A partitioning strategy based on spatial sensitivity distribution and depth information is adopted. The optical parameters of shallow and deep tissues are reconstructed by using logarithmic integral intensity and mean flight time, respectively. The quasi-Newton method is combined with the optimization model, and the adjoint method is used to reduce computational complexity and memory usage.
It significantly improves the accuracy and stability of optical parameter reconstruction, reduces computational overhead, enhances the reliability and efficiency of reconstruction results, and alleviates the ill-conditioned problem.
Smart Images
Figure CN122415799A_ABST