Methods Inf Med 1993; 32(05): 388-395
DOI: 10.1055/s-0038-1634951
Original Article
Schattauer GmbH

Influence of “Outliers” on the Association between Laboratory Data and Histopathological Findings in Liver Biopsy

E. Krusinska
1   Department of Medical Informatics, Linköping University, Sweden
2   Department of Mathematics and Informatics, Conservatoire National des Arts et Métiers, France
3   Institute of Electrical Metrology, Technical University of Wroclaw, Poland
,
U. L. Mathiesen
4   Department of Internal Medicine, Linköping University Hospital, Sweden
6   Department of Internal Medicine, Oskarshamn Hospital, Sweden
,
L. Franzén
5   Department of Pathology, Linköping University Hospital, Sweden
,
G. Bodemar
4   Department of Internal Medicine, Linköping University Hospital, Sweden
,
O. Wigertz
1   Department of Medical Informatics, Linköping University, Sweden
› Institutsangaben
Weitere Informationen

Publikationsverlauf

Publikationsdatum:
08. Februar 2018 (online)

Abstract

Discriminant analysis techniques were used to predict the histopathological findings in liver biopsy specimens in asymptomatic patients with slightly to moderately raised routine liver tests. Moderate to severe fibrosis and/or inflammation were treated as indication for biopsy. Two methods were used to classify patients. One was the dichotomous discrimination between “biopsy necessary” or “biopsy not necessary” groups of patients. The other involved combining two discriminant functions trained separately for recognition of fibrosis or inflammation, and then combined to predict the biopsy necessity. Detection of outliers by standard techniques, directly available in the SPSS-X package, was performed before starting discrimination procedures. Both “sharp” assignment rules and continuous scoring rules were applied to the classification problem. The correct classification rate reached over 85% for the algorithms tested. In the majority of cases the classification was found to be “non-doubtful”. Elimination of outliers (especially by standardized residuals) improved the global correct classification rate, but only slightly improved assignment to the “biopsy necessary” group. Routine and complementary laboratory findings were found to be the most discriminating; answers to questionnaire and ultrasound examination were less important. Selection of the most diagnostic features based on “clean” data without outliers enabled us to find interesting medical associations, which were previously masked by extremely asymptomatic values outlying from the main body of the “biopsy necessary” group.

 
  • REFERENCES

  • 1 Hilden J, Matzen P, Malchow-Møller A, Bryant S. et al. Precision requirements in a study of computer aided diagnosis of jaundice (the COMIK study). Scand J Clin Lab Invest 1980; 40 (Suppl 1557): 125-8.
  • 2 Lindberg G, Seensalu R, Nilsson LH, Forsell P. et al. Transferability of a computer system for medical history taking and decision support in dyspepsia. A comparison of individuals for peptic ulcer disease. Scand J Gastroenterol 1987; 22 (Suppl. 128) 190-6.
  • 3 Malchow-Møller A, Thomsen C, Hilden J, Matzen P. et al. A decision tree for early differentiation between obstructive and non-obstructive jaundice. Scand J Gastroenterol 1988; 23: 391-401.
  • 4 Malchow-Møller A, Thomsen C, Matzen P, Mindeholm L. et al. Computer diagnosis of jaundice: Bayes’ rule founded on 1002 consecutive cases. J Hepat 1986; 03: 154-63.
  • 5 Matzen P, Malchow-Møller A, Hilden J, Thomsen C. et al. Differential diagnosis of jaundice: a pocket diagnostic chart. Liver 1984; 04: 360-71.
  • 6 Segaar RW, Wilson JHP, Habbema JDF, Hilden J. A computer aid for early diagnostic classification of jaundice (The COMIP program). Comput Meth Progr Biomed 1989; 28: 131-6.
  • 7 Solberg HE, Skrede S, Blomhoff JP. Diagnosis of liver diseases by laboratory results and discriminant analysis. Scand J Clin Lab Invest 1975; 35: 713-21.
  • 8 Chowdhury S, Bodemar G, Haug P, Babic A, Wigertz O. Methods for knowledge extraction from a clinical database on liver diseases. Comput Biomed Res 1991; 06: 530-48.
  • 9 Hampel FR, Ronchetti EM, Rousseeuw PJ, Stahel WA. Robust Statistics. The Approach Based on Influence Functions. New York: Wiley; 1986
  • 10 Krusinska E. Robust methods in discriminant analysis. Rivista di Statistica Applicata 1988; 21: 239-53.
  • 11 Knodell RG, Ishak KG, Black WC, Chen TS. et al. Formulation and application of a numerical scoring system for assessing histological activity in asymptomatic chronic active hepatitis. Hepatology 1981; 01: 431-5.
  • 12 Broffit B, Clarke WR, Lachenbruch PA. The effect of Huberizing and trimming on the quadratic discriminant function. Comm Statistics A. Theory and Methods 1980; 09: 13-25.
  • 13 Seber GAF. Multivariate Observations . New York: Wiley; 1984
  • 14 Krusinska E, Liebhart J. The influence of outliers on discrimination results of chronic obturative lung diseases. Meth Inform Med 1988; 27: 167-76.
  • 15 SPSS-X. Statistical Algorithms . Chicago: Mc Graw-Hill; 1983
  • 16 Lachenbruch PA. Discriminant Analysis . New York: Hafner Press; 1975
  • 17 Copas JB. Binary regression models for contaminated data. With discussion. J R Statist Soc B 1988; 50: 225-65.
  • 18 Krusinska E, Babic A, Mathiesen U, Franzén L. et al. Empirical modelling versus commonly applied data analysis techniques as used for decision support in liver diseases. In: MEDINFO 92 . Amsterdam: North-Holland Publ Comp; 1992: 949-55.
  • 19 Hoaglin DC, Mosteller F, Tukey JW. Understanding Robust and Exploratory Data Analysis . New York: Wiley; 1984