# US 12,183,462 B2 · Method for predicting lung cancer development based on artificial intelligence model, and analysis device therefor > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,183,462 B2 - Document ID: US12183462B2 - Status: Granted - Kind: B2 - Application number: 17/641,692 - Filed: 2020-04-03 - Published: 2022-09-22 - Granted: 2024-12-31 - Earliest claimed: 2019-09-10 - Assignees: Biolink Corporation - Applicants: BIOLINK Inc - Inventors: Hae Won Kim - Web page: https://patentgrep.com/patent/US12183462B2 ## CPC - G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems - A61B 6/032: {Transmission computed tomography [CT]} - A61B 6/037: {Emission tomography} - A61B 6/5217: {extracting a diagnostic or physiological parameter from medical diagnostic data} - G06N 3/02: Neural networks - G06T 7/0012: {Biomedical image inspection} - G06T 7/0014: {using an image reference approach} - G06V 10/22: by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition - G06V 10/764: using classification, e.g. of video objects - G06V 10/774: Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting - G06V 10/82: using neural networks - G16B 20/00: ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations - G16B 40/20: Supervised data analysis - G16H 30/40: for processing medical images, e.g. editing - G16H 40/67: for remote operation - G16H 50/30: for calculating health indices; for individual health risk assessment - G16H 50/70: for mining of medical data, e.g. analysing previous cases of other patients - G06T 2207/10081: Computed x-ray tomography [CT] - G06T 2207/10104: Positron emission tomography [PET] - G06T 2207/20076: Probabilistic image processing - G06T 2207/20081: Training; Learning - G06T 2207/20084: Artificial neural networks [ANN] - G06T 2207/20132: Image cropping - G06T 2207/30061: Lung - G06V 2201/03: Recognition of patterns in medical or anatomical images - G16B 20/20: Allele or variant detection, e.g. single nucleotide polymorphism [SNP] detection - G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA - A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING - G16B: BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY ## Abstract A method for determining a probability of developing lung cancer by using an artificial intelligence model, includes: receiving, by an analysis device, a chest fluorodeoxyglucose (FDG) PET/CT image for a sample; inputting, by the analysis device, the F-18 FDG PET/CT image to a first classification neural network, and outputting prediction information related to development of lung cancer for the sample; and predicting, by the analysis device, a probability of developing lung cancer for the sample on the basis of the prediction information. The first classification neural network is trained by using chest F-18 FDG PET/CT images for healthy people and training images excluding lung cancer regions from chest F-18 FDG PET/CT images for lung cancer patients. ## Claims ### Claim 1 A method for determining a probability of developing lung cancer by using an artificial intelligence model, the method comprising: receiving, by an analysis device, a chest F-18 fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) image for a sample; inputting, by the analysis device, the chest F-18 FDG PET/CT image to a first classification neural network and outputting prediction information related to development of the lung cancer for the sample; and predicting, by the analysis device, a probability of developing lung cancer for the sample based on the prediction information, wherein the first classification neural network is trained using a chest F-18 FDG PET/CT image for a healthy person and a training image excluding a lung cancer region from a chest F-18 FDG PET/CT image for a lung cancer patient. ### Claim 2 (depends on 1) The method of claim 1, wherein the first classification neural network is a convolution neural network (CNN) and outputs the prediction information for classifying the sample into a low-risk group or a high-risk group. ### Claim 3 (depends on 1) The method of claim 1, further comprising generating an image including information on the probability of developing lung cancer by inputting, by the analysis device, feature information output from the first classification neural network to a decoder neural network. ### Claim 4 (depends on 1) The method of claim 1, further comprising inputting, by the analysis device, the chest F-18 FDG PET/CT image to a detection neural network to detect a lung region in the F-18 FDG PET/CT image, wherein the analysis device inputs the detected lung region to the first classification neural network. ### Claim 5 (depends on 1) The method of claim 1, wherein the training image is generated by inputting the chest F-18 FDG PET/CT image for the lung cancer patient to a separate neural network and extracting a region excluding the lung cancer region from the chest F-18 FDG PET/CT image for the lung cancer patient. ### Claim 6 (depends on 1) The method of claim 1, further comprising: receiving, by the analysis device, omics data for the sample; and inputting, by the analysis device, the omics data to a second classification neural network and outputting second prediction information on the probability of developing lung cancer for the sample, wherein the analysis device predicts the probability of developing lung cancer for the sample based on the prediction information output to the first classification neural network and the second prediction information output from the second classification neural network. ### Claim 7 (depends on 1) The method of claim 1, further comprising receiving, by the analysis device, omics data for the sample, wherein the analysis device further inputs the omics data to the first classification neural network and outputs the prediction information, and the first classification neural network changes a coefficient or a weight of a filter for at least one of a plurality of layers based on the omics data for the sample. ### Claim 8 An analysis device comprising: an input device configured to receive a chest F-18 fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) image for a sample; a storage device configured to store a first classification neural network that generates prediction information on a probability of developing lung cancer using the chest F-18 FDG PET/CT image; and a computing device configured to input the received chest F-18 FDG PET/CT image to the first classification neural network to predict a probability of a subject developing lung cancer, wherein the first classification neural network is trained using a chest F-18 FDG PET/CT image for a healthy person and a training image excluding a lung cancer region from a chest F-18 FDG PET/CT image for a lung cancer patient. ### Claim 9 (depends on 8) The analysis device of claim 8, wherein the first classification neural network is trained using a PET/CT image labeled with F-18 FDG. ### Claim 10 (depends on 8) The analysis device of claim 8, wherein the first classification neural network is a convolution neural network (CNN) and outputs the prediction information for classifying the sample into a low-risk group or a high-risk group. ### Claim 11 (depends on 8) The analysis device of claim 8, wherein the storage device further stores a detection neural network for detecting a lung region in the chest F-18 FDG PET/CT image, and the computing device inputs the chest F-18 FDG PET/CT image to the detection neural network to detect a lung region in the chest F-18 FDG PET/CT image and inputs the detected lung region to the first classification neural network. ### Claim 12 (depends on 8) The analysis device of claim 8, wherein the storage device further stores a decoder neural network that receives feature information output from the first classification neural network and generates an image including information on the probability of developing lung cancer, and the computing device inputs classification information and the feature information output from the first classification neural network to the decoder neural network to generate an image including the classification information. ### Claim 13 (depends on 8) The analysis device of claim 8, wherein the training image is generated by inputting the chest F-18 FDG PET/CT image for the lung cancer patient to a separate neural network and extracting a region excluding the lung cancer region from the chest F-18 FDG PET/CT image for the lung cancer patient. ### Claim 14 (depends on 8) The analysis device of claim 8, wherein the input device further receives omics data for the sample, the storage device further stores a second classification neural network that receives the omics data and outputs second prediction information on the probability of developing lung cancer for the sample, and the computing device predicts the probability of developing lung cancer for the sample based on the prediction information output to the first classification neural network and the second prediction information output from the second classification neural network. ### Claim 15 (depends on 8) The analysis device of claim 8, wherein the input device further receives omics data for the sample, the computing device further inputs the omics data to the first classification neural network and outputs the prediction information, and the first classification neural network changes a coefficient or a weight of a filter for at least one of a plurality of layers based on the omics data for the sample. ### Claim 16 A method for determining a probability of developing lung cancer by using an artificial intelligence model, the method comprising: inputting, by an analysis device, a chest F-18 fluorodeoxyglucose (FDG) positron emission tomography (PET) image for a sample to a first classification neural network and outputting first prediction information related to the development of lung cancer for the sample; inputting, by the analysis device, a chest computed tomography (CT) image for the sample to a second classification neural network and outputting second prediction information related to the development of lung cancer for the sample; and predicting, by the analysis device, a probability of developing lung cancer for the sample based on the first prediction information and the second prediction information, wherein the first classification neural network is trained using a chest F-18 FDG PET image for a healthy person and a first training image excluding a lung cancer region from a chest F-18 FDG PET image for a lung cancer patient, and the second classification neural network is trained using a chest CT image for a healthy person and a second training image excluding the lung cancer region from a chest CT image for the lung cancer patient. ### Claim 17 (depends on 16) The method of claim 16, further comprising inputting, by the analysis device, the chest F-18 FDG PET image to a first detection neural network to detect a lung region in the chest F-18 FDG PET image, the analysis device inputting the detected lung region to the first classification neural network, and inputting, by the analysis device, the chest CT image to a second detection neural network to detect the lung region in the chest CT image, the analysis device inputting the detected lung region in the chest CT image to the second classification neural network. ### Claim 18 (depends on 16) The method of claim 16, wherein the first training image is generated by inputting the chest F-18 FDG PET image for the lung cancer patient to a separate neural network and extracting a region excluding the lung cancer region from the chest F-18 FDG PET image for the lung cancer patient, and the second training image is generated by inputting the chest CT image for the lung cancer patient to a separate neural network and extracting a region excluding the lung cancer region from the chest CT image for the lung cancer patient. ### Claim 19 (depends on 16) The method of claim 16, further comprising receiving, by the analysis device, omics data for the sample, wherein the analysis device further inputs the omics data to the first classification neural network to output the first prediction information and further inputs the omics data to the second classification neural network to output the second prediction information.