	US 7,599,536 B2
	Optimizing image signal interpretation for analog image signals from medical image recording devices
Florian Kraus, Haar (Germany); Christian Maier, Munich (Germany); and Manfred Weiser, Munich (Germany)
Assigned to BrainLAB AG, Feldkirchen (Germany)
Filed on Aug. 01, 2005, as Appl. No. 11/194,506.
Claims priority of provisional application 60/606008, filed on Aug. 31, 2004.
Claims priority of application No. 04018173 (EP), filed on Jul. 30, 2004.
Prior Publication US 2006/0023926 A1, Feb. 02, 2006
Int. Cl. G06K 9/00 (2006.01)

U.S. Cl. 382—128 [382/132; 128/922]

12 Claims

1. A method for optimizing the interpretation of analog image signals or sequences of image signals output by medical image recording devices, comprising the steps of:

a) testing the correlation of consecutively recorded image signals;

b) establishing that the image signals depict the same image if the correlation is not less than a particular threshold value;

c) establishing that the image signals possibly depict different images if the correlation is less than the particular threshold value; and

d) dynamically adjusting the threshold value if the correlation has changed

wherein when it is established that different images are depicted, a certain number of consecutive image signals below the threshold value are evaluated, and a standard deviation of the image signals is determined and compared with a pre-set value, wherein:

c1a) if the number of changed image signals in succession is at least equal to a predetermined sample number and the standard deviation is lower than the pre-set value, it is established that the correlation deviation has been caused by changes in the manner of recording the same image; and

c1b) the threshold value is adjusted lower;

c2a) if the number of changed image signals in succession is at least equal to the predetermined sample number and the standard deviation is higher than the pre-set value, it is established that the correlation deviation has been caused by continuously detecting images from different and changing recording situations;

c2b) the threshold value is not adjusted;

c3) if the number of changed image signals in succession is less than the predetermined sample number, an image signal from the consecutive image signals is classified as a new image.