Anatomical measurement in a surgical system
AssigneeAuris Health, Inc.
InventorsRami Azmi Nimer Abukhalil, Amer G. Ghanem, Kenneth Fernandez Prada, Mary Lynn Dwyer-Gaddis
Life of this patent
Estimating the term…
Abstract
For anatomy measurement in a surgical system, a machine-learned model is used to determine the location or length. For example, in RYGB, the machine-learned model indicates how much of the bowel has been pulled or run. As another RYGB example, the machine-learned model indicates a length along the bowel to a current location. In yet another RYGB example, the machine-learned model indicates a location for incision based on a desired length. Using an image or images, such as a video stream from an endoscope, the artificial intelligence provides the measurement, avoiding running with estimated steps size or use an inserted ruler.
Claims20
A method for measurement for a surgical system, the method comprising:
- capturing a video stream from a camera coupled to a first robotic arm of a surgical robotic system, the video captured intra-operatively during surgery where the surgical robotic system operates a surgical tool connected to a second robotic arm;
- measuring, by a machine-learned model in response to input of the video stream, a position associated with a grasp point of the surgical tool connected to the second robotic arm, the machine-learned model configured to output the position in response to the input of the video stream to the machine-learned model;
- and
- displaying the position measured during surgery.
A surgical system for measurement during surgery, the surgical system comprising:
- a medical imager configured to capture an image of anatomy during surgery for the anatomy;
- an image processor configured to determine, by a machine-learned model in response to input of the image, a location based on a surgical tool grasp location, the machine-learned model configured by training data for ground truth measurements;
- and
- a display configured to display the location during the surgery based on the machine-learned model.
A method for machine training a model for measuring during surgery, the method comprising:
- obtaining training data comprising surgical images and lengths along anatomy in the surgical images;
- machine training the model with the training data, the model machine trained to output the lengths in response to input of the surgical images;
- and
- storing the machine-trained model.
Dependencies are inferred from the claim text.