An Improved Computer-aided Diagnosis Scheme Using the Nearest Neighbor Criterion for Determining Histological Classification of Clustered Microcalcifications

R. Nakayama; R. Watanabe; K. Namba; K. Takeda; K. Yamamoto; S. Katsuragawa; K. Doi

doi:10.1055/s-0038-1625433

Methods of Information in Medicine, Inhaltsverzeichnis

Methods Inf Med 2007; 46(06): 716-722
DOI: 10.1055/s-0038-1625433

Original Article

Schattauer GmbH

An Improved Computer-aided Diagnosis Scheme Using the Nearest Neighbor Criterion for Determining Histological Classification of Clustered Microcalcifications

R. Nakayama

¹Department of Radiology, Mie University School of Medicine, Tsu, Japan

,

R. Watanabe

²Department of Breast Surgery, Hakuaikai Hospital, Fukuoka, Japan

,

K. Namba

³Breastopia Namba Hospital, Miyazaki, Japan

,

K. Takeda

¹Department of Radiology, Mie University School of Medicine, Tsu, Japan

,

K. Yamamoto

⁴Medical Informatics Section, Mie University School of Medicine, Tsu, Japan

,

S. Katsuragawa

⁵Department of Health Sciences, Kumamoto University School of Medicine, Kuhonji, Japan

,

K. Doi

⁶Kurt Rossmann Laboratories for Radiologic Image Research, Department of Radiology, The University of Chicago, Chicago, IL, USA

› Institutsangaben

Abstract

Summary

Objectives : Our purpose was to evaluate the potential usefulness of the nearest neighbor case which was assumed to be the similar case in a CAD scheme for determining the histological classification of clustered microcalcifications.

Methods : Our database consisted of current and previous magnification mammograms obtained from 93 patients before and after three-month follow-up examination. It included 11 invasive carcinomas, 19 noninvasive carcinomas of the comedo type, 25 noninvasive carcinomas of the noncomedo type, 23 mas- topathies, and 15 fibroadenomas. Six objective features on clustered microcalcifications were first extracted from each of the current and the previous images. The nearest neighbor case was then identified by the Euclidean distance in the previous and current feature-space. The histological classification of an unknown new case in question was assumed to be thesame as that of the nearest neighbor case which has the shortest Euclidean distance in our database.

Results : The classification accuracies were 90.9% for invasive carcinoma, 89.5% for noninvasive carcinoma of the comedo type, 96.0% for noninvasive carcinoma of the noncomedo type, 82.6% for mastopathy, and 93.3% for fibroadenoma. These results were substantially higher than those with our previous CAD scheme.

Conclusion : The nearest neighbor criterion was useful in a CAD scheme for determining the histological classification.

Keywords

Computer-aided diagnosis - clustered microcalcification - histological classification - nearest neighbor criterion - follow-up examination

Volltext

Referenzen

References
1. Sickles EA. Mammographic features of early breast cancer. AJR 1984; 143 (461) 464.
2. Sickles EA. Mammographic features of 300 consecutive nonpalpable breast cancers. AJR 1986; 146: 661-663.
3. Adler DD, Helvie MA. Mammographic biopsy recommendations. Current Opinion in Radiology 1992; 4: 123-129.
4. Kopans DB. The positive predictive value of mammography. AjR 1992; 158: 521-526.
5. Kopans DB. Breast imaging: second edition. New York, NY: Lippincott-Raven publishers 1997
6. Morimoto T, Sasa M. Atlas of Screening Mammography. Tokyo: Digital-Press; 1996 (in Japanese)
7. Sakamoto G, Haga S. Fundamental and clinic of ductal carcinoma in situ. Tokyo: Shinoharashinsha; 2001 (in Japanese)
8. Nakayama R, Uchiyama Y, Watanabe R, Katsuragawa S, Namba K, Doi K. Computer-aided Diagnosis Scheme for Histological Classification of Clustered Microcalcifications on Magnification Mammograms. Med Phys 2004; 31: 789-799.
9. Nakayama R, Watanabe R, Namba K, Takeda K, Yamamoto K, Katsuragawa S, Doi K. Computer-aided diagnosis scheme for identifying histological classification of clustered microcalcifications by use of follow-up magnification mammograms. Acad Radiol 2006; 13: 1219-1228.
10. Muramatsu C, Li Q, Schmidt R, Suzuki K, Shiraishi J, Newstead G, Doi K. Experimental determination of subjective similarity for pairs of clustered microcalcifications on mammograms: observer study results. Med Phys 2006; 33: 3460-3468.
11. Muramatsu C, Li Q, Suzuki K, Schmidt RA, Shiraishi J, Newstead GM, Doi K. Investigation of psychophysical measure for evaluation of similar images for mammographic masses: preliminary results. Med Phys 2005; 32: 2295-2304.
12. Nakayama R, Uchiyama Y. Development of new filter bank for detection of nodular patterns and line patterns in medical images. Systems and Computers in Japan 2005; 36: 81-91.
13. Nakayama R, Uchiyama Y, Yamamoto K, Watanabe R, Namba K. Computer-aided diagnosis scheme using a filter bank for detection of microcalcification clusters in mammograms. IEEE Trans Biomed Eng 2006; 53: 273-283.
14. Gonzales RC, Woods RE. Digital Image Processing. MA: Addison-Wesley; 1992
15. Sergios T, Konstantinos K. Pattern Recognition. London: Academic Press; 2003
16. Hermann G, Keller RJ, Drossman S, Caravella BA, Tartter P, Panetta RA, Bleiweiss IJ. Mammographic pattern of microcalcifications in the preoperative diagnosis of comedo ductal carcinoma in situ: histopathologic correlation. Can Assoc Radiol J 1999; 50: 235-240.
17. Fondrinier E, Lorimier G, Guerin-Boblet V, Bertrand AF, Mayras C, Dauver N. Breast microcalcifications: multivariate analysis ofradiologic and clinical factors for carcinoma. World J Surg 2002; 26: 290-296.
18. Alberdi E, Taylor P, Lee R. Elicitation and representation of expert knowledge for computer aided diagnosis in mammography. Methods Inf Med 2004; 43: 239-246.
19. Johnson RA. Wichern DW Applied Multivariate Statistical Analysis. NJ: Prentice-Hall Englewood Cliffs; 1992
20. Langlotz CP. Fundamental measures ofdiagnostic examination performance: Usefulness for clinical decision making and research. Radiology 2003; 228: 3-9.