Classification of vitamin A deficiency levels by ocular changes in Japanese black cattle

To help Japanese black cattle farmers diagnose vitamin A deficiency (VAD) levels in cattle, eye images of 40 cattle were recorded monthly during their vitamin A manipulated period using a 2-CCD camera. Ocular features extracted from images, including pupil colour, pupillary light reflex and light reflection, were investigated. Multivariate classification methods (SIMCA and PLS-DA) were used to classify cattle into mild, moderate and severe VAD groups. Five variables (r, I, CA, IPR, I_RFL) were used for classification. The mild and severe VAD groups could be classified with over 85% correct classification rate. However, the moderate VAD group could not be discriminated adequately. A VAD index was developed and proven to be effective in representing VAD status. The results showed the potential for ocular changes to be utilised as an aid to farm management.