| 摘要 |
PROBLEM TO BE SOLVED: To provide an analytical method or the like for analyzing a time-serial data as to a brain image, capable of estimating residual radioactivity without generating an error, and capable of obtaining an excellent image, when using a divided dose method. SOLUTION: This analytical method is input with the time-serial data 2 as to the brain image. The input time-serial data 2 as to the brain image is learnt in a neural network 4. The time-serial data 2 as to the brain image in a specified period is input into the neural network 4 submitted to the learning to predict the time-serial data in a period after the specified period. An input data vector comprises, as elements, brain radioactive data within prescribed period as to a plurality of positions in brains about a plurality of subjects collected in a load test using a brain blood flow SPECT. The minimum measure is a Euclidean distance. The specified period is a period before the second tracer dosage after the first tracer dosage in the load test by the divided dose method using the blood flow SPECT. COPYRIGHT: (C)2008,JPO&INPIT
|
