| Sonstiges: |
- Nachgewiesen in: USPTO Patent Grants
- Sprachen: English
- Patent Number: 11682,484
- Publication Date: June 20, 2023
- Appl. No: 17/850441
- Application Filed: June 27, 2022
- Assignees: MERIT CRO, INC. (Madison, WI, US)
- Claim: 1. An optical coherence tomography (OCT) image analysis process comprising: visualizing an OCT dataset from a scan of a patients retina on a display device, wherein the image displays a plurality of cross-sectional retinal layers of the retina; indicating a portion of an edge of at least one of the retinal layers with a user input device to provide a designated retinal layer; at a user work station, calculating a patient reflectivity profile for the designated retinal layer and using the reflectivity profile to identify potential retinal locations of the designated retinal layer across the entire image; transmitting the patient reflectivity profile to a server remote from the user work station, wherein the remote server comprises a plurality of OCT datasets from normal and diseased retinas; via a processor associated with the remote server, applying one or more machine learning algorithms to analyze the patient reflectivity profile in relation to the plurality of OCT datasets from normal and diseased retinas to generate one or more algorithms that automatically segment retinal layers in an OCT image, automatically identify lesions in one or more retinal layers, and/or associate a disease with an automated segmentation or lesion identification result, wherein the algorithm facilitates displaying a refined trace of the designated retinal layer.
- Claim: 2. The process of claim 1 , further comprising the step of displaying an image with a refined trace of the designated retinal layer generated by the algorithm.
- Claim: 3. The process of claim 2 , further comprising the step of transmitting the image with a refined trace of the designated retinal layer to a user.
- Claim: 4. The process of claim 1 , wherein the algorithm associates the reflectivity profile of the designated retinal layer with a disease state of the retina wherein the disease state is selected from the group consisting of Age-related macular degeneration, diabetic retinopathy, retinitis pigmentosa, uveitis, central vein occlusion, and other retinal degenerations.
- Claim: 5. The process of claim 4 , further comprising the step of using the algorithm to associate a disease state or normal state with the patient retina.
- Claim: 6. The process of claim 1 , wherein the algorithm identifies lesions in one or more retinal layers.
- Claim: 7. The process of claim 6 , wherein the type and/or location of the lesion is used to diagnose a disease or the retina and/or designate a stage of severity of a disease of the retina.
- Claim: 8. The process of claim 5 , further comprising transmitting information about the disease state or normal state of the patient retina to a user.
- Claim: 9. The process of claim 7 , further comprising transmitting information about the lesion, disease associated with the lesion, or stage of severity of retinal disease to a user.
- Claim: 10. The process of claim 1 , further comprising the step of tagging the patient reflectivity profile with an information identification tag, wherein the information identification tag comprises information selected from the group consisting of name of the surface, location at the retina, disease indication and lesion indication, retinal location relating to known retinal landmark (e.g., fovea), age, gender, race, and animal species (other than human).
- Claim: 11. The process of claim 1 , wherein the display device is networked with an SD-OCT device.
- Claim: 12. The process of claim 11 , wherein the process further comprising the step of utilizing the SD-OCT device to obtain the OCT dataset of the scan of the patient's retina.
- Claim: 13. The process of claim 1 , wherein the server remote from the user work station is a cloud based server.
- Claim: 14. The process of claim 1 , wherein the one or more machine learning algorithms are selected from the group exemplified by but not limited to a neural network, a decision tree, a regression model, a k-nearest neighbor model, a partial least squares model, a support vector machine and an ensemble of the models that are integrated to define an algorithm.
- Claim: 15. The process of claim 1 , further comprising obtaining multiple patient retinal images over a defined time period for a given patient, via a processor automatically analyzing the multiple retinal images to identify changes in the retinal images, and displaying an image showing the identified changes in the patient retinal images.
- Patent References Cited: 20110141259 June 2011 Nakano ; 20120184845 July 2012 Ishikawa ; 20120274898 November 2012 Sadda ; 20160174830 June 2016 Rubin ; 20160183783 June 2016 Sadda ; 20170119247 May 2017 Walsh
- Primary Examiner: Wilburn, Molly
- Attorney, Agent or Firm: Casimir Jones S.C. ; Jones, J. Mitchell
|
