US 12,267,577 B2
Apparatus and method for imaging fundus of eye
Ilkka Jolma, Oulu (FI); Kenneth Oksanen, Oulu (FI); and Juho Pylvänäinen, Oulu (FI)
Assigned to Optomed Plc, Oulu (FI)
Filed by Optomed Plc, Oulu (FI)
Filed on Nov. 14, 2022, as Appl. No. 17/986,525.
Claims priority of application No. 21208395 (EP), filed on Nov. 16, 2021.
Prior Publication US 2023/0156320 A1, May 18, 2023
Int. Cl. H04N 23/60 (2023.01); A61B 3/00 (2006.01); A61B 3/12 (2006.01); A61B 3/14 (2006.01); H04N 23/611 (2023.01); H04N 23/63 (2023.01)
CPC H04N 23/64 (2023.01) [A61B 3/0008 (2013.01); A61B 3/12 (2013.01); A61B 3/145 (2013.01); H04N 23/611 (2023.01); H04N 23/635 (2023.01)] 20 Claims
OG exemplary drawing
 
1. An apparatus configured to image a fundus of an eye, comprising:
an optical system;
an image sensor configured to capture a still image or a video through the optical system;
a user interface including a display configured to display data to a user of the apparatus;
one or more processors configured to cause performance of at least the following:
setting the image sensor to capture an aiming video;
detecting a retina of the eye in the aiming video, wherein detecting the retina comprises using an Adaptive Boosting, AdaBoost, statistical classification meta-algorithm or one of its variants, which constructs a strong classifier by combining results of a sequence of weak classifiers;
setting the display to display the aiming video with a highlight of the detected retina;
setting the image sensor to capture one or more final still images of the fundus or a final video of the fundus;
using, by the weak classifiers, also one or more results of comparisons from previous weak classifiers in the sequence of the classifiers to improve the accuracy of a next weak classifier in the sequence of the classifiers; and
comparing, by the weak classifiers, also average probabilities of pixels in areas in the single frame belonging to the retina as determined by an already executed weak classifier as follows: comparing either an average probability of an area in the single frame relative to the pixel of interest to a third constant, or a difference of averages of probabilities of two areas in the single frame relative to the pixel of interest to a fourth constant.