Abstract
The authors have developed an automated computeraided diagnostic (CAD) scheme by using artificial neural networks (ANNs) on quantitative analysis of image data. Three separate ANNs were applied for detection of interstitial disease on digitized chest images. The first ANN was trained with horizontal profiles in regions of interest (ROIs) selected from normal and abnormal chest radiographs for distinguishing between normal and abnormal patterns. For training and testing of the second ANN, the vertical output patterns obtained from the 1st ANN were used for each ROI. The output value of the second ANN was used to distinguish between normal and abnormal ROIs with interstitial infiltrates. If the ratio of the number of abnormal ROIs to the total number of all ROIs in a chest image was greater than a specified threshold level, the image was classified as abnormal. In addition, the third ANN was applied to distinguish between normal and abnormal chest images. The combination of the rule-based method and the third ANN also was applied to the classification between normal and abnormal chest images. The performance of the ANNs was evaluated by means of receiver operating characteristic (ROC) analysis. The average Az value (area under the ROC curve) for distinguishing between normal and abnormal cases was 0.976±0.012 for 100 chest radiographs that were not used in training of ANNs. The results indicate that the ANN trained with image data can learn some statistical properties associated with interstitial infiltrates in chest radiographs.
Key words: interstitial infiltrate, computer-aided diagnosis, artificial neural network, chest radiograph
Full Text
The Full Text of this article is available as a PDF (985.0 KB).
Footnotes
This study was supported by USPHS Grants CA24806 and CA62625.
K. Doi and H. MacMahon are shareholders of R2 Technology, Inc, Los Altos, CA. It is the policy of the University of Chicago that investigators disclose publicly actual or potential significant financial interests that may appear to be affected by the research activities.
References
- 1.MacMahon H, Liu KJM, Montner SM, et al. The nature and subtlety of abnormal findings in chest radiographs. Med Phys. 1991;18:206–210. doi: 10.1118/1.596708. [DOI] [PubMed] [Google Scholar]
- 2.Herman PG, Gerson DE, Hessel SJ, et al. Disagreement in chest roentgen interpretation. Chest. 1975;68:278–282. doi: 10.1378/chest.68.3.278. [DOI] [PubMed] [Google Scholar]
- 3.Katsuragawa S, Doi K, MacMahon H. Image feature analysis and computer-aided diagnosis in digital radiography: Detection and characterization of interstitial lung diseases in digital chest radiographys. Med Phys. 1988;15:311–319. doi: 10.1118/1.596224. [DOI] [PubMed] [Google Scholar]
- 4.Katsuragawa S, Doi K, MacMahon H. Image feature analysis and computer-aided diagnosis in digital radiography: Classification of normal and abnormal lungs with interstitial diseases in chest images. Med Phys. 1989;16:38–44. doi: 10.1118/1.596412. [DOI] [PubMed] [Google Scholar]
- 5.Chen X, Doi K, Katsuragawa S, et al. Automated selection of regions of interest for quantitative analysis of lung textures in digital chest radiographs. Med Phys. 1993;20:975–982. doi: 10.1118/1.596979. [DOI] [PubMed] [Google Scholar]
- 6.Morishita J, Doi K, Katsuragawa S, et al. Computer-aided diagnosis for interstitial infiltrates in chest radiographs: Opticaldensity dependence of texture measures. Med Phys. 1995;22:1515–1522. doi: 10.1118/1.597419. [DOI] [PubMed] [Google Scholar]
- 7.Monnier-Cholley L, MacMahon L, Katsuragawa S, et al. Computerized analysis of interstitial infiltrates on chest radiographs: A new scheme based on geometric-pattern features and Fourier analysis. Acad Radiol. 1995;2:455–462. doi: 10.1016/S1076-6332(05)80399-X. [DOI] [PubMed] [Google Scholar]
- 8.Katsuragawa S, Doi K, MacMahon H, et al. Quantitative analysis of geometric-pattern features of interstitial infiltrates in digital chest radiographs: Preliminary results. J Digit Imaging. 1996;9:137–144. doi: 10.1007/BF03168609. [DOI] [PubMed] [Google Scholar]
- 9.Abe K, Doi K, MacMahon H, et al. Computer-aided diagnosis in chest radiography: Preliminary experience. Invest Radiol. 1993;28:987–993. doi: 10.1097/00004424-199311000-00002. [DOI] [PubMed] [Google Scholar]
- 10.Asada N, Doi K, MacMahon H, et al. Potential usefulness of an artificial neural network for differential diagnosis of interstitial diseases. Pilot study. Radiology. 1990;177:857–860. doi: 10.1148/radiology.177.3.2244001. [DOI] [PubMed] [Google Scholar]
- 11.Xu XW. Computerized analysis for automated detection of lung nodules in digitized chest radiographs. Chicago, IL: University of Chicago; 1996. [Google Scholar]
- 12.Metz CE. ROC methodology in radiographic imaging. Invest Radiol. 1986;21:720–733. doi: 10.1097/00004424-198609000-00009. [DOI] [PubMed] [Google Scholar]
- 13.Metz CE. Some practical issues of experimental design and data analysis in radiographical ROC studies. Invest Radiol. 1989;24:234–245. doi: 10.1097/00004424-198903000-00012. [DOI] [PubMed] [Google Scholar]
- 14.Metz CE, Shen JH, Herman BA: New methods for estimating a binormal ROC curve from continuously distributed test results. Invited presentation at the 1990 Joint Statistical Meeting of the American Statistical Society and the Biometric Society, Anaheim, CA, 1990.
- 15.Monnier-Cholley L, Katsuragawa S, Morishita J, et al. Objective evaluation of computerized scheme for detection of interstitial infiltrates on chest radiographs. Radiology. 1994;193P:146–146. [Google Scholar]