Abstract
The differentiation of focal liver lesions in magnetic resonance (MR) imaging is primarily based on the intensity and homogeneity of lesions with different imaging sequences. However, these imaging findings are falsely interpreted in some patients because of the complexities involved. Our aim is to establish a computer-aided diagnosis system named LiverANN for classifying the pathologies of focal liver lesions into five categories using the artificial neural network (ANN) technique. On each MR image, a region of interest (ROI) in the focal liver lesion was delineated by a radiologist. The intensity and homogeneity within the ROI were calculated automatically, producing numerical data that were analyzed by feeding them into the LiverANN as inputs. Outputs were the following five pathologic categories of hepatic disease: hepatic cyst, hepatocellular carcinoma, dysplasia in cirrhosis, cavernous hemangioma, and metastasis. Of the 320 MR images obtained from 80 patients (four images per patient) with liver lesions, our LiverANN classified 50 cases of a training set into five types of liver lesions with a training accuracy of 100% and 30 test cases with a testing accuracy of 93%. The experiment demonstrated that our LiverANN, which functions as a computer-aided differentiation tool, can provide radiologists with a second opinion during the radiologic diagnostic procedure.
Similar content being viewed by others
References
Landis S, Murray T, Bolden S. Cancer statistics. CA Cancer J Clin. 1999;49:8–31.
Matsui O, Kadoya M, Kameyama T. Adenomatous hyperplastic nodules in the cirrhotic liver: differentiation from hepatocellular carcinoma with MR imaging. Radiology. 1989;173:123–6.
Garra B, Insana M, Shawker T, Russell M. Quantitative estimation of liver attenuation and echogenicity: normal state versus diffuse liver disease. Radiology. 1994;14:1099–108.
Hartman P, Oosterveld B, Thijssen J, Rosenbusch G. Detection and differentiation of diffuse liver disease by quantitative echography: a retrospective assessment. Invest Radiol. 1993;28:1–6.
Gosnik B, Lemon S, Scheible W, Leupold G. Accuracy of ultrasonography in diagnosis of hepatocellular disease. Am J Roentgenol. 1979;133:19–23.
Foster K, Dewbury K, Griffith A, Wright R. The accuracy of ultrasound in the detection of fatty infiltration of the liver. Br J Radiol. 1980;53:440–2.
Ogawa K, Fukushima M, Kubota K, Hisa N. Computer-aided diagnosis system for diffuse liver diseases with ultrasonography by neural network. IEEE Trans Nucl Sci. 1998;45:3069–74.
Wang Y, Itph K, Sum S. Study on tissue characterization by texture analysis with co-occurrence matrix method using ultrasonography and CT imaging. J Med Ultrason. 2002;29:211–23.
Bae K, Giger M, Chen C. Automatic segmentation of liver structure in CT images. Med Phys. 1993;20:71–8.
Selle D, Preim B, Schenk A. Analysis of vasculature for liver surgical planning. IEEE Trans Med Imaging. 2002;21:1344–57.
Tombropoulos R, Shiffman S, Davidson C. A decision aid for diagnosis of liver lesions on MRI. Proc Annu Symp Comput Appl Med Care. 1993;439–43.
Maclin PS, Dempsey J. How to improve a neural network for early detection of hepatic cancer. Cancer Lett. 1994;77:95–101.
Genevieve L, Andrea P, Williamand W, Peter F, Wolfgang S, Sanjay S. Addition of gadolinium chelates to heavily T2-weighted MR imaging: limited role in differentiating hepatic hemangiomas from metastases. Am J Roentgenol. 2000;174:477–85.
Honda H, Matsuura Y, Onitsuka H. Differential diagnosis of hepatic tumors (hepatoma, hemangioma, and metastasis) with CT: value of two-phase incremental imaging. Am J Roentgenol. 1992;159:735–40.
Whitney W, Herfkens R, Jeffrey R, McDonnell C, Li K, Dalsem W, et al. Dynamic breath-hold multiplanar spoiled gradient-recalled MR imaging with gadolinium enhancement for differentiating hepatic hemangiomas from malignancies at 1.5 T. Radiology. 1993;189:863–70.
Hamm B, Fischer E, Taupitz M. Differentiation of hepatic hemangiomas from metastases by dynamic contrast-enhanced MR imaging. J Comput Assist Tomogr. 1990;14:205–16.
Hamm B, Thoeni RF, Gould R, Bernardino M, Luning M, Saini S, et al. Focal liver lesions: characterization with nonenhanced and dynamic contrast material enhanced MR imaging. Radiology. 1994;190:417–23.
Semelka R, Brown E, Ascher S, Patt R, Bagley A, Li W, et al. Hepatic hemangiomas: a multi-institutional study of appearance on T2-weighted and serial gadolinium-enhanced images. Radiology. 1994;192:401–6.
Semelka R, Shoenut J, Kroeker M, Greenberg H, Simm F, Minuk G, et al. Focal liver disease: comparison of dynamic contrast-enhanced CT and T2-weighted fat suppressed, FLASH, and dynamic gadolinium enhanced MR imaging at 1.5 T. Radiology. 1992;184:687–94.
Mergo P, Ros PR. Benign lesions of the liver. Radiol Clin North Am. 1998;36:319–31.
Lombardo DM, Baker ME, Spritzer CE, Blinder R, Meyers W, Herfkens RJ. Hepatic hemangiomas vs metastases: MR differentiation at 1.5 T. Am J Roentgenol. 1990;155:55–9.
Itoh K, Saini S, Hahn P, Imam N, Ferrucci J. Differentiation between small hepatic hemangiomas and metastases on MR images: importance of size-specific quantitative criteria. Am J Roentgenol. 1990;155:61–6.
Li K, Glazer G, Quint L, Francis I, Aisen A, Ensminger W, et al. Distinction of hepatic cavernous hemangioma from hepatic metastases with MR imaging. Radiology. 1988;169:409–15.
Brown J, Lee M, Lee J, Lom K, Malchow S. Focal hepatic lesions: differentiation with MR imaging at 0.5 T. Radiology. 1991;179:675–9.
Mirowitz S, Lee J, Heiken J. Cavernous hemangioma of the liver: assessment of MR tissue specificity with a simplified T2 index. J Comput Assist Tomogr. 1990;14:223–8.
Asada N, Doi K, MacMahon H, Montner S, Giger ML, Abe C, et al. Potential usefulness of an artificial neural network for differential diagnosis of interstitial lung disease: pilot study. Radiology. 1990;177:857–60.
Fujita H, Katafuchi T, Uehara T, Nishimura T. Application of artificial neural network to computer-aided diagnosis of coronary artery disease in myocardial SPECT bull’s-eye images. J Nucl Med. 1992;33:272–6.
Zhang W, Doi K, Giger M, Nishikawa R, Schmidt R. An improved shift-invariant artificial neural network for computerized detection of clustered microcalcifications in digital mammograms. Med Phys. 1996;23:595–601.
Seki K, Fujita H, Hirako K, Hara T, Ando T. Detection of microcalcifications on mammograms using neural networks. Med Imag Tech. 1997;15:639–51.
Kako N, Zhang X, Kanematsu M. Study of computer-aided diagnosis on MR imaging. Radiology. 2002;255(P):749. (abstract).
Kako N, Zhang X, Li W. Artificial neural network method for differentiation of focal liver disease in MR imaging. Radiology. 2002;255(P):646.
Rumelhart D, McClellend J. Learning representations by back-propagating errors. Nature. 1986;323:533–6.
Towell G. Knowledge-based artificial neural networks. Artif Intell. 1994;70:119–65.
Acknowledgments
This research was supported in part by the National Natural Science Foundations of China (No. 60863014 and 60762001); in part by the Program to Sponsor Teams for Innovation in the Construction of Talent Highlands in Guangxi Institutions of Higher Learning; in part by the National Creative Plan for Undergraduate Students of Guangxi University (No. 200707 and X071036); in part by a research foundation project of the Guangxi Ministry of Education (No. 200810MS048); and in part by a research grant from the Collaborative Centre for Academy/Industry/Government of Gifu University, the Ministry of Health, Labor, and Welfare under a Grant-In-Aid for Cancer Research, and the Ministry of Education, Culture, Sports, Science and Technology under a Grant-In-Aid for Scientific Research by the Japanese Government.
Author information
Authors and Affiliations
Corresponding author
About this article
Cite this article
Zhang, X., Kanematsu, M., Fujita, H. et al. Application of an artificial neural network to the computer-aided differentiation of focal liver disease in MR imaging. Radiol Phys Technol 2, 175–182 (2009). https://doi.org/10.1007/s12194-009-0062-5
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12194-009-0062-5