Artificial intelligence guidance system for robotic surgery
This invention is a system and method for utilizing artificial intelligence to operate a surgical robot (e.g., to perform a laminectomy), including a surgical robot, an artificial intelligence guidance system, an image recognition system, an image recognition database, and a database of past procedures with sensor data, electronic medical records, and imaging data. The image recognition system may identify the tissue type present in the patient and if it is the desired tissue type, the AI guidance system may remove a layer of that tissue with the end effector on the surgical robot, and have the surgeon define the tissue type if the image recognition system identified the tissue as anything other than the desired tissue type.
1. A method implemented in a computing system for at least partially controlling a robotic surgical apparatus to perform surgical actions, the method comprising:
obtaining a first image of a region of interest associated with a subject;
identifying a type of tissue shown in the first image based, at least in part, on a neural network model trained on an image training set;
in response to determining that the identified type of tissue belongs to a set of targeted types, causing the robotic surgical apparatus to perform a first surgical action with respect to the region of interest in accordance with a surgical plan;
obtaining a second image of the region of interest after completion of the first surgical action; and
in response to a failure to identify a type of tissue shown in the second image based, at least in part, on the neural network model:
presenting the second image to a user; and
receiving an input from the user indicating a type of tissue shown in the second image.
2. The method of claim 1 , further comprising:
in response to determining that the user indicated type of tissue belongs to the set of targeted types:
updating the image training set with the second image; and
causing the surgical apparatus to perform a second surgical action with respect to the region of interest in accordance with the surgical plan.
3. The method of claim 2 , further comprising causing the neural network model to be re-trained based, at least in part, on the updated image training set.
4. The method of claim 1 , wherein the first surgical action includes removal of a portion of tissue.
5. The method of claim 1 , wherein the region of interest is in front of an end effector of the robotic surgical apparatus.
6. The method of claim 1 , wherein the image training set includes at least a plurality of pre-collected images and metadata of tissue type associated with each pre-collected image.
7. The method of claim 1 , wherein an end effector of the robotic surgical apparatus is changed after completion of the first surgical action.
8. The method of claim 1 , further comprising enabling a user to halt the execution of the method in response to a fail-safe trigger.
9. The method of claim 8 , wherein the fail-safe trigger is based on at least one of a time sequence of identified tissue types, a length of time between identification of normal tissue types, a trend of probability associated with tissue type identification, or a quality of obtained image.
10. A non-transitory computer-readable storage medium storing content that, when executed by one or more processors, causes the one or more processors to perform actions including:
obtaining a first image of a region of interest associated with a surgery subject;
identifying a type of tissue shown in the first image based, at least in part, on a neural network model;
in response to determining that the identified type of tissue belongs to a set of targeted types, causing a robotic surgical apparatus to perform a first surgical action with respect to the region of interest in accordance with a surgical plan; and
obtaining a second image of the region of interest after completion of the first surgical action;
in response to a failure to identify a type of tissue shown in the second image based, at least in part, on the neural network model:
presenting the second image to a user; and
receiving an input from the user indicating a type of tissue shown in the second image.
11. The computer-readable storage medium of claim 10 , wherein the identified type of tissue includes at least one of bone, cartilage, ligaments, disc, solid organs, or aberrant tissue.
12. The computer-readable storage medium of claim 10 , wherein the first surgical action is performed by a first end effector of the robotic surgical apparatus.
13. The computer-readable storage medium of claim 12 , wherein the actions further include determining to replace the first end effector with a second end effector in accordance with the surgical plan.
14. The computer-readable storage medium of claim 10 , wherein a first end effector of the robotic surgical apparatus includes at least one of a robotic gripper, cutting instrument, cannula, reamer, rongeur, scissors, retractor tube, incision marker, bone removal tool, or skin and/or muscle fascia incision tool.
15. A system for at least partially controlling a robotic surgical apparatus, comprising:
one or more processors;
at least one display; and
memory storing contents that, when executed by the one or more processors, cause the system to perform actions including:
obtaining one or more first images of a region of interest associated with a surgery subject;
identifying one or more types of tissue shown in the one or more first images based, at least in part, on a neural network model;
displaying the identified one or more types of tissue via a graphical user interface (GUI) on the at least one display;
receiving one or more interactions with the GUI from a user that indicate one or more surgical configurations;
causing a robotic surgical apparatus to perform one or more surgical actions with respect to the region of interest based, at least in part, on the one or more surgical configurations;
obtaining one or more second images of the region of interest after completion of at least one of the surgical actions; and
in response to a failure to identify one or more types of tissue shown in the one or more second images based, at least in part, on the neural network model:
presenting at least one of the second images to the user; and
receiving an input from the user indicating one or more types of tissue shown in the at least one of the second images.
16. The system of claim 15 , wherein the at least one display includes a touch screen.
17. The system of claim 15 , wherein the one or more surgical configurations include at least one of targeted tissue to be removed or nontargeted tissue to be protected.
18. The system of claim 15 , wherein the one or more surgical configurations include at least one of an area of tissue to be removed, a sensitive area to stay away from, or operating parameters of the robotic surgical apparatus.
19. The system of claim 18 , wherein the operating parameters of the robotic surgical apparatus further include at least one of penetration depth, operating distance, operating speed, end effector path, hold location, or transition between hold locations.
20. The system of claim 15 , wherein displaying the identified one or more types of tissue comprises displaying at least one of an estimated shape, suggested color, or meta-tag data associated with each type of identified tissue.
21. The system of claim 15 , wherein displaying the identified one or more types of tissue comprises displaying one or more boundary indicators for indicating at least one of targeted tissue to be removed, protected tissue, delivery instrument placement, or an end effector working space within the subject.