Learning Rules That Classify Ocular Fundus
Images for Glaucoma Diagnosis

Fumio Mizoguchi1, Hayato Ohwada1, Makiko Daidoji1 and Shiroteru Shirato2

1 Science University of Tokyo
Noda, Chiba 278, Japan
2 Department of Ophthalmology,
University of Tokyo, Japan


Abstract. This paper provides an empirical study of an Inductive Logic
Programming (ILP) method through the application to classifying ocular
fundus images for glaucoma diagnosis. Key issues in this study are not
only dealing with low-level measurement data such as image, but also
producing diagnostic rules that are readable and comprehensive for 
interactions to medical experts. For this purpose, we develop a 
constraint-directed ILP system, GKS, that handles both symbolic and numerical
data, and produce Horn clauses with numerical constraints. Furthermore,
we provide GKS with a sequential learning facility where GKS 
repeatedly generates a single best rule which becomes background knowledge
for the next learning phase. Since the learning target for this application
is the abnormality of each segment in image, generated rules represent
the relationships between abnormal segments. Since such relationships
can be interpreted as qualitative rules and be used as diagnostic rules 
directly, the present method provides automatic construction of knowledge
base from experts accumulated diagnostic experience. Furthermore, the
experimental result shows that induced rules have high statistical 
performance. The present study indicates the advantage and possibility of
the ILP approach to medical diagnosis from measurement data.
References

[Quinlan, 1992] Quinlan.J.R., 04.5: Programs for Machine Learning, Morgan 
Kaufmann, San Mateo,California, 1992.
[Quinlan, 1993] Quinlan.J.R.,Cameron-Jones.R.M, FOIL: A midterm report, Proc. of
European Conference on Machine Learning, Springer Verlag, pp.3-20, 1993.
[Kitazawa, 1981] Kitazawa Y., Sbirato S., Mizoguchi F., A new Computer-Aided 
Glaucoma Consultation System (G4-Expert), Royal Society of Medicine International
Congress and Symposium Series, No.44, pp.161-168, 1981.
[Lloyd, 1984] Foundations of Logic Programming, Springer-Verlag, 1984.
[Mizoguchi, 1994] Mizoguchi F., Ohwada H., Constrained Relative Least General 
Generalization for Inducing Constraint Logic Programs, New Generation Computing,
Vol.13, No.3,4, pp. 335-368, 1995.
[Mizoguchi, 1995] Mizoguchi F., Ohwada H., Using Inductive Logic Programming for
Constraint Acquisition in Constraint-based Problem Solving, Proc. of the 5th 
International Workshop on Inductive Logic Programming, Vol.13, 1995.
[Muggleton, 1991] Muggleton, S., Inductive Logic Programming, New Generation
Computing, Vol. 8, No. 4, pp.295-318, 1991.
[Muggleton, 1992] Muggleton, S., King, R. and Sternberg, M., Protein secondary 
structure prediction using logic-based machine learning, Protein Engineering, Vol. 5, No.
7, pp. 647-657, 1992.
[Muggleton, 1995] Muggleton, S., Inverse Entailment and Progol, New Generation
Computing, Vol. 13, No.3,4, pp.245-286, 1995.
[Hiraki, 1991] Hiraki K., J. Genneri, Yamamoto Y., Anzai Y., Learning Spatial 
Relations from Images, Proc. of 8th International Workshop on Machine Learning,
pp.407-411, 1991.
[Weiss, 1990] Weiss, S. M., Galen, R. S. and Tadepalli, P., Maximizing the Predictive
Value of Production Rules, Artificial Intelligence, Vol. 45, pp. 47-71, 1990.
