A hybrid classifier for automated radiologic diagnosis: preliminary results and clinical applications
From Semantic Portal Wiki
{{#vardefine:category|Publication}}{{#vardefine:templatename|i.publication}}{{#vardefine:package|smwbp_instance_templates}}
| Edit |
Reference: {{#vardefine:pagename|a hybrid classifier for automated radiologic diagnosis: preliminary results and clinical applications }}
- [[]]
bibtex
{{#vardefine:pagename|A hybrid classifier for automated radiologic diagnosis: preliminary results and clinical applications }}{{#vardefine:key| }}
abstract: We describe the design, implementation, and preliminary evaluation of acomputer system to aid clinicians in the interpretation of cranial magnetic-resonance (MR) images. The system classifies normal and pathologic tissues ina test set of MR scans with high accuracy. It also provides a simple, rapidmeans whereby an unassisted expert may reliably label an image with his bestjudgment of its histologic composition, yielding a gold-standard image; thisstep facilitates objective evaluation of classifier performance. This systemconsists of a preprocessing module; a semiautomatic, reliable procedure forobtaining objective estimates of an expert's opinion of an image's tissuecomposition; a classification module based on a combination of the maximum-likelihood (ML) classifier and the ISODATA unsupervised-clustering algorithm;and an evaluation module based on confusion-matrix generation. The algorithmsfor classifier evaluation and gold-standard acquisition are advances overprevious methods. Furthermore, the combination of a clustering algorithm anda statistical classifier provides advantages not found in systems using eithermethod alone.
download:
- paper:
- slides:
| Abstract | We describe the design, implementation, an … We describe the design, implementation, and preliminary evaluation of acomputer system to aid clinicians in the interpretation of cranial magnetic-resonance (MR) images. The system classifies normal and pathologic tissues ina test set of MR scans with high accuracy. It also provides a simple, rapidmeans whereby an unassisted expert may reliably label an image with his bestjudgment of its histologic composition, yielding a gold-standard image; thisstep facilitates objective evaluation of classifier performance. This systemconsists of a preprocessing module; a semiautomatic, reliable procedure forobtaining objective estimates of an expert's opinion of an image's tissuecomposition; a classification module based on a combination of the maximum-likelihood (ML) classifier and the ISODATA unsupervised-clustering algorithm;and an evaluation module based on confusion-matrix generation. The algorithmsfor classifier evaluation and gold-standard acquisition are advances overprevious methods. Furthermore, the combination of a clustering algorithm anda statistical classifier provides advantages not found in systems using eithermethod alone. found in systems using eithermethod alone. |
| Author | Edward Herskovits + |
| Bibtype | techreport + |
| Institution | Knowledge Systems, AI Laboratory + |
| Key | KSL-90-46 + |
| Number | KSL-90-46 + |
| Tag | Computer science + |
| Title | A Hybrid Classifier for Automated Radiologic Diagnosis: Preliminary Results and Clinical Applications + |
| Tr id | KSL-90-46 + |
| Year | 1990 + |
