| Presentation | 2001/1/18 Segment Classification Method for CT Images of Liver Cancer based on Time-Transition Pattern of CT Values Masaki Ishiguro, Shingo Inoue, Ichiro Murase, Noriyuki Moriyama, |
|---|---|
| PDF Download Page | PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | We report on a classification method for tumor segments in CT images of livers based on decision rules derived by inductive learning techniques. We also describe a method to classify types of segments based on time-trasition of CT values of segments obtained by Dynamic CT and Simple CT images. We define features of each segment or a group of segments based on morphological features which are typically used to discriminate cancers. We generate rules for discriminating cancer tumors from other ones are generated by applying decision tree analysis and score segments based on these rules. We further apply discrimination analysis to scores and original feature values to determine weight of each value for discriminating cancer segments. Experimental results of ROC analysis show that classification methods incorporating rules derived by inductive learning outperforms the one without learned rules in terms of generalization ability. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Pattern Recognition / Feature Extraction / Image Segmentation / Decision Tree Analysis / Liver Cancers |
| Paper # | MI2000-63 |
| Date of Issue |
| Conference Information | |
| Committee | MI |
|---|---|
| Conference Date | 2001/1/18(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Medical Imaging (MI) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Segment Classification Method for CT Images of Liver Cancer based on Time-Transition Pattern of CT Values |
| Sub Title (in English) | |
| Keyword(1) | Pattern Recognition |
| Keyword(2) | Feature Extraction |
| Keyword(3) | Image Segmentation |
| Keyword(4) | Decision Tree Analysis |
| Keyword(5) | Liver Cancers |
| 1st Author's Name | Masaki Ishiguro |
| 1st Author's Affiliation | Information Technologies Research Dept., Mitsubishi Research Institute, Inc() |
| 2nd Author's Name | Shingo Inoue |
| 2nd Author's Affiliation | Information Technologies Research Dept., Mitsubishi Research Institute, Inc |
| 3rd Author's Name | Ichiro Murase |
| 3rd Author's Affiliation | Information Technologies Research Dept., Mitsubishi Research Institute, Inc |
| 4th Author's Name | Noriyuki Moriyama |
| 4th Author's Affiliation | Diagnostic Radiology Division, National Cancer Center |
| Date | 2001/1/18 |
| Paper # | MI2000-63 |
| Volume (vol) | vol.100 |
| Number (no) | 596 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |