System of using artificial intelligence in canine radiography

1 . A method for predicting one or more of vertebral heart scale (VHS) and vertebral left atrial scores (VLAS) from canine radiography carried out by one or more digital devices comprising:

receiving one or more photographed canine radiography images, wherein the one or more photographed canine radiography images were captured using a camera on a mobile device to photograph one or more existing radiography images;

performing image localization and neural network pattern recognition by algorithmically determining one or more points of interest;

locating two or more positions on the one or more photographed canine radiography images, the two or more positions comprising a position on a ninth thoracic vertebrae;

comparing two or more lengths on the one or more photographed canine radiography images; and

algorithmically generating a score, wherein the score is one or more of a vertebral heart scale score and a vertebral left atrial score, and wherein the score is based at least in part on the two or more lengths on the one or more photographed canine radiography images.

2 . The method for predicting one or more of VHS and VLAS from canine radiography of claim 1 , further comprising:

receiving tagging data from one or more veterinarians, wherein the tagging data corresponds to the one or more photographed canine radiography images; and

generating one or more training points.

3 . The method for predicting one or more of VHS and VLAS from canine radiography of claim 1 , further comprising:

identifying and categorizing specific pixel colors based at least in part on the one or more photographed canine radiography images; and

generating bounding boxes around the one or more points of interest.

4 . The method for predicting one or more of VHS and VLAS from canine radiography of claim 1 , further comprising:

performing reinforcement learning by:

establishing one or more bounding boxes corresponding to the one or more photographed canine radiography images;

performing one or more shaping functions; and

assigning one or more rewards, wherein the one or more rewards are weighted based at least in part on the one or more shaping functions.

5 . The method for predicting one or more of VHS and VLAS from canine radiography of claim 1 , wherein at least one of the one or more digital devices is at least one server and at least one of the one or more digital devices is at least one user device, further comprising:

sending first data from the at least one server to the at least one user device;

displaying a web portal on the at least one user device;

displaying viewable statistics on the at least one user device; and

sending second data from the at least one user device to the at least one server.

6 . The method for predicting one or more of VHS and VLAS from canine radiography of claim 1 , further comprising:

locating positions on the one or more photographed canine radiography images, the positions comprising:

a most ventral aspect of a carina;

an apex of a heart;

a most caudal aspect of a left atrium at which the left atrium intersects with a dorsal border of a caudal vena cava;

a cranial aspect of a vertebral body of a fourth thoracic vertebrae; and

a cranial aspect of a vertebral body of the ninth thoracic vertebrae; and

calculating a first length of at least one vertebral disk on the one or more photographed canine radiography images.

7 . The method for predicting one or more of VHS and VLAS from canine radiography of claim 1 , further comprising:

calculating a confidence level; and

displaying the confidence level.

8 . The method for predicting one or more of VHS and VLAS from canine radiography of claim 1 , further comprising determining whether the one or more photographed canine radiography images are obscured.

9 . A system of one or more digital devices for predicting one or more of vertebral heart scale (VHS) and vertebral left atrial scores (VLAS) from canine radiography comprising:

at least one server;

at least one processor;

at least one non-transitory computer-readable medium; and

program instructions stored on the at least one non-transitory computer-readable medium that are executable by the at least one processor such that the system is configured to:

receive one or more photographed canine radiography images, wherein the one or more photographed canine radiography images were captured using a camera on a mobile device to photograph one or more existing radiography images;

perform image localization and neural network pattern recognition by algorithmically determining one or more points of interest;

locate two or more positions on the one or more photographed canine radiography images, the two or more positions comprising a position on a ninth thoracic vertebrae;

compare two or more lengths on the one or more photographed canine radiography images; and

algorithmically generate a score, wherein the score is one or more of a vertebral heart scale score and a vertebral left atrial score, and wherein the score is based at least in part on the two or more lengths on the one or more photographed canine radiography images.

10 . The system of one or more digital devices for predicting one or more of vertebral heart scale (VHS) and vertebral left atrial scores (VLAS) from canine radiography of claim 9 , wherein the program instructions are executable by the at least one processor such that the system is further configured to:

receive tagging data from one or more veterinarians, wherein the tagging data corresponds to the one or more photographed canine radiography images; and

generate one or more training points.

11 . The system of one or more digital devices for predicting one or more of vertebral heart scale (VHS) and vertebral left atrial scores (VLAS) from canine radiography of claim 9 , wherein the program instructions are executable by the at least one processor such that the system is further configured to:

identify and categorize specific pixel colors based at least in part on the one or more photographed canine radiography images; and

generate bounding boxes around the one or more points of interest.

12 . The system of one or more digital devices for predicting one or more of vertebral heart scale (VHS) and vertebral left atrial scores (VLAS) from canine radiography of claim 9 , wherein the program instructions are executable by the at least one processor such that the system is further configured to:

perform reinforcement learning by:

establishing one or more bounding boxes corresponding to the one or more photographed canine radiography images;

performing one or more shaping functions; and

assigning one or more rewards, wherein the one or more rewards are weighted based at least in part on the one or more shaping functions.

13 . The system of one or more digital devices for predicting one or more of vertebral heart scale (VHS) and vertebral left atrial scores (VLAS) from canine radiography of claim 9 , wherein at least one of the one or more digital devices is at least one server and at least one of the one or more digital devices is at least one user device, wherein the program instructions are executable by the at least one processor such that the system is further configured to:

send first data from the at least one server to the at least one user device;

display a web portal on the at least one user device;

display viewable statistics on the at least one user device; and

send second data from the at least one user device to the at least one server.

14 . The system of one or more digital devices for predicting one or more of vertebral heart scale (VHS) and vertebral left atrial scores (VLAS) from canine radiography of claim 9 , wherein the program instructions are executable by the at least one processor such that the system is further configured to:

locate positions on the one or more photographed canine radiography images, the positions comprising:

a most ventral aspect of a carina;

an apex of a heart;

a most caudal aspect of a left atrium at which the left atrium intersects with a dorsal border of a caudal vena cava;

a cranial aspect of a vertebral body of a fourth thoracic vertebrae; and

a cranial aspect of a vertebral body of a cranial aspect of a vertebral body of the ninth thoracic vertebrae; and

calculate a first length of at least one vertebral disk on the one or more photographed canine radiography images.

15 . The system of one or more digital devices for predicting one or more of vertebral heart scale (VHS) and vertebral left atrial scores (VLAS) from canine radiography of claim 9 , wherein the program instructions are executable by the at least one processor such that the system is further configured to:

calculate a confidence level; and

display the confidence level.

16 . The system of one or more digital devices for predicting one or more of vertebral heart scale (VHS) and vertebral left atrial scores (VLAS) from canine radiography claim 9 , wherein the program instructions are executable by the at least one processor such that the system is further configured to determine whether the one or more photographed canine radiography images are obscured.

17 . A non-transitory computer-readable medium, wherein the non-transitory computer-readable medium is provisioned with program instructions that, when executed by at least one processor, cause one or more devices to:

receive one or more photographed canine radiography images, wherein the one or more photographed canine radiography images were captured using a camera on a mobile device to photograph one or more existing radiography images;

perform image localization and neural network pattern recognition by algorithmically determining one or more points of interest;

locate two or more positions on the one or more photographed canine radiography images, the two or more positions comprising a position on a ninth thoracic vertebrae;

compare two or more lengths on the one or more photographed canine radiography images; and

algorithmically generate a score, wherein the score is one or more of a vertebral heart scale score and a vertebral left atrial score, and wherein the score is based at least in part on the two or more lengths on the one or more photographed canine radiography images.

18 . The non-transitory computer-readable medium of claim 17 , wherein the program instructions are executable by the at least one processor such that the one or more devices are further configured to:

receive tagging data from one or more veterinarians, wherein the tagging data corresponds to the one or more photographed canine radiography images; and

generate one or more training points.

19 . The non-transitory computer-readable medium of claim 17 , wherein the program instructions are executable by the at least one processor such that the one or more devices are further configured to:

identify and categorize specific pixel colors based at least in part on the one or more photographed canine radiography images; and

generate bounding boxes around the one or more points of interest.

20 . The non-transitory computer-readable medium of claim 17 , wherein the program instructions are executable by the at least one processor such that the one or more devices are further configured to:

perform reinforcement learning by:

establishing one or more bounding boxes corresponding to the one or more photographed canine radiography images;

performing one or more shaping functions; and

assigning one or more rewards, wherein the one or more rewards are weighted based at least in part on the one or more shaping functions.