Skip to main content
Fig. 1 | BioMedical Engineering OnLine

Fig. 1

From: Detecting central fixation by means of artificial neural networks in a pediatric vision screener using retinal birefringence scanning

Fig. 1

Neural network architecture. A two-layer architecture has been employed, consisting of one hidden layer and one output layer. The network has an input p containing four inputs (p1–p4), representing the four normalized RBS spectral powers, respectively P2.5/P4.5, P3.5/P4.5, P5.5/P4.5, and P6.5/P4.5. The hidden layer contains four neurons. The output of the net signals presence or absence of central fixation

Back to article page