It has already been reported that by applying the method of quantification model No.2, it is possible to screen dairy cows with functional disorders of the liver in the field.
The present investigation was carried out to obtain a discriminant function of higher correctness than before. In it, analysis was performed with the following points of improvement.
(1) The number of cows surveyed was increased. (2) Predictor variables were reexami-ned. (3) Criterion variables were classified exactly.
In the previous investigation discrimination was carried out only by using discriminant function No.l of such eigenvector as corresponding to the largest eigenvalue that had been determined by calculation. In the present investigation discriminant function No.2 of such eigenvector as corresponding to the second largest eigenvalue was also determined. Then sample scores obtained by discriminant function No.1 were plotted on the X-coordinate and those obtained by discriminant function No.2 on the Y-coordinate. Individuals were discriminated on the two-dimensional plane formed by the two coordinates. As a result, 192 dairy cows which were suspected clinically to be suffering from hepatic dysfunction were divided into three goups, 1 to 3, in which lipid pigmentation in hepatocytes was negative (-), positive (+), and strongly positive (++), respectively, on the basis of general symptons and results of simple clinical examinations.
*Dliscrimination could be performed in these groups with the following accuracy.
group 1 (43/50 head) : 86%.
group 2 (29/36 head) : 81%.
group 3 (40/43 head) : 93%.
Therefore it was suggested that this discriminant function might be applicable to the diagnosis of hepatic dysfunction in the field.

View full abstract