US 12,092,569 B2
Method and system for detecting a biological substance by hyperspectral imaging
Danny S. Moshe, Ramat-HaSharon (IL)
Assigned to Green Vision Systems Ltd., Tel Aviv (IL)
Filed by Green Vision Systems Ltd., Tel-Aviv (IL)
Filed on Oct. 21, 2021, as Appl. No. 17/506,762.
Claims priority of provisional application 63/184,319, filed on May 5, 2021.
Claims priority of provisional application 63/094,391, filed on Oct. 21, 2020.
Prior Publication US 2022/0120674 A1, Apr. 21, 2022
Int. Cl. G01N 21/31 (2006.01); G01N 1/44 (2006.01); G01N 33/497 (2006.01); G06K 9/00 (2022.01); G06K 9/46 (2006.01); G06V 10/58 (2022.01); G06V 10/60 (2022.01); G06V 20/69 (2022.01); A61B 5/097 (2006.01)
CPC G01N 21/31 (2013.01) [G01N 1/44 (2013.01); G01N 33/497 (2013.01); G06V 10/58 (2022.01); G06V 10/60 (2022.01); G06V 20/693 (2022.01); A61B 5/097 (2013.01); G06V 2201/03 (2022.01)] 13 Claims
OG exemplary drawing
 
1. A method of detecting a biological substance in a sample, the method comprising:
collecting the sample as an aerosol;
adding to the sample a bio-marker that directly or indirectly binds to the biological substance in the sample;
illuminating the sample by light;
at a hyperspectral imaging system, imaging the illuminated sample by Fourier transform hyperspectral imaging a plurality of times, to generate a plurality of hyperspectral images;
accessing a computer readable medium containing a library of hyperspectral signatures of bio-markers;
at an image processor, analyzing said plurality of hyperspectral images using said library to detect the biological substance in the sample by identifying hyperspectral signatures in said plurality of hyperspectral images and comparing said identified hyperspectral signatures to hyperspectral signatures of bio-markers in said library; and
estimating a level of the identified biological substance in the sample, using a linear function of a percentage of occurrences of the identified hyperspectral signature within the hyperspectral image, wherein for each hyperspectral image of said plurality of hyperspectral images said linear function is calculated based on a difference between a normalized percentage of pixels identified as a hyperspectral signature and a normalized percentage of pixels identified as a reference hyperspectral signature in a reference hyperspectral image, wherein said percentage of pixels identified as said hyperspectral signature is relative to a number of non-background pixels in said hyperspectral image, wherein said percentage of pixels identified as a reference hyperspectral signature in said reference hyperspectral image is relative to a number of non-background pixels in said reference hyperspectral image, and wherein said non-background pixels are pixels for which an intensity is above a predetermined intensity threshold.