Automatic steering of an introducer with a video laryngoscope
Methods and systems for automatic steering of a blind introducer based on images from an independent camera (e.g., a camera of a video laryngoscope) are disclosed. A real time image of a video laryngoscope may be analyzed by a trained ML model to output a classification for the image. The classification may be associated with steering instructions for an introducer. The steering instructions may be provided to the introducer to cause steering of a distal end of the introducer in real time. The trained ML model may be generated by AI or ML training algorithms, based on multiple sets of multiple still-shot training images associated with a finite quantity of classifications.
1 . A method for steering an introducer with a video laryngoscope, the method comprising:
receiving an image from a camera of a video laryngoscope, the image including a distal end of an introducer and an anatomical structure;
providing the image as an input to a trained machine-learning (ML) model, wherein the ML model is trained to classify images as steering instructions for the introducer, and the trained ML model is trained using a training dataset including a first set of training images pre-classified with a first steering instruction and a second set of training images pre-classified with a second steering instruction;
receiving as output from the trained ML model, a classification of the image;
based on the classification, generating a steering instruction for the introducer; and
causing the distal end of the introducer to be steered in accordance with the steering instruction.
2 . The method of claim 1 , further comprising:
receiving an indication to automatically steer the introducer, wherein generating the steering instruction for the introducer is in response to the indication to automatically steer the introducer.
3 . The method of claim 1 , wherein the anatomical structure is a trachea or vocal cords.
4 . The method of claim 1 , the method further comprising:
displaying, at the video laryngoscope, a visual indicator associated with the steering instruction.
5 . The method of claim 1 , wherein the classification includes the steering instruction and wherein the steering instruction includes a direction and a magnitude for steering the distal end.
6 . The method of claim 5 , wherein the classification of the image is selected from a finite set of at least four classifications.
7 . The method of claim 1 , wherein the image is from a video feed of the camera of the video laryngoscope.
8 . The method of claim 1 , the method further comprising:
determining that the introducer is advancing at a rate of speed, wherein generating the steering instruction for the introducer is based on the rate of speed.
9 . A video laryngoscope comprising:
a handle portion;
a display screen coupled to the handle portion;
a blade portion, coupled to the handle portion, configured to be inserted into a mouth of a patient;
a camera, positioned at a distal end of the blade portion, that acquires a video feed while the video laryngoscope is powered on;
a memory storing a trained machine-learning (ML) model; and
a processor that operates to:
receive an indication to automatically steer an introducer;
receive an image of the video feed from the camera in real time, the image comprising a single frame of the video feed depicting an introducer portion associated with a distal end of the introducer;
classify, by the trained ML model, the image, wherein the classification includes the steering instruction and the classification is selected from a finite set of at least four classifications;
based on the classification of the image, generate a steering instruction to steer the distal end of the introducer; and
send the steering instruction to a steering system of the introducer to bend the distal end of the blind introducer in the steering direction.
10 . The system of claim 9 , wherein the image is the only input into the trained ML model.
11 . The system of claim 9 , wherein manual steering of the introducer at the video laryngoscope is prevented while the automatic steering instructions are sent to the introducer.
12 . A method for steering a blind introducer via a third-person perspective camera, the method comprising:
receiving a video feed from a camera having a third-person view of a blind introducer, the video feed including a distal end of the blind introducer;
receiving a first image from the video feed of the camera;
classifying, by a trained machine-learning (ML) model, the first image with a first classification;
determining, based on the first classification, a first bending angle for the distal end of the blind introducer;
instructing a steering system of the blind introducer, in real time, to bend the distal end according to the determined first bending angle;
receiving a second image from the video feed of the camera; and
classifying, by the trained ML model, the second image with a second classification, in real time.
13 . The method of claim 12 , wherein the second classification is associated with no action of the distal end of the blind introducer.
14 . The method of claim 12 , further comprising:
determining, based on the second classification, a second bending angle for the distal end of the blind introducer; and
instructing a steering system of the blind introducer, in real time, to bend the distal end according to the determined second bending angle.
15 . The method of claim 12 , wherein the ML model is trained to classify images as steering instructions for the blind introducer, and the trained ML model is trained using a training dataset including a first set of training images pre-classified with a first steering instruction and a second set of training images pre-classified with a second steering instruction.
16 . The method of claim 12 , wherein the classification includes a steering instruction including a direction and a magnitude for steering the distal end.
17 . The method of claim 12 , wherein the first classification and the second classification are selected from a finite set of at least four classifications.
18 . The method of claim 12 , wherein the camera is part of a video laryngoscope.
19 . The method of claim 18 , wherein the method is performed by the video laryngoscope.
20 . The method of claim 19 , wherein the blind introducer is physically coupled to the video laryngoscope.