Endoscopic training system
A training system for an endoscopic procedure, the training system including an endoscope having an image sensor configured to generate images; an image processing device including a housing and a processing circuit in the housing, the processing circuit including a processor and memory, the memory having graphical user interface (GUI) logic and a trained single-pass neural network model, the processor being configured to process the neural network model and the GUI logic to present a GUI; and a physical model including cavities corresponding to interconnected human cavities. The GUI is configured to receive a user input corresponding to a training mode or an exam mode. In the training mode the GUI is configured to present navigation feedback, and in the exam mode the GUI is configured to not present navigation feedback.
1 . A training system for an endoscopic procedure, the training system comprising:
an endoscope having an image sensor configured to generate image sensor images;
an image processing device including a housing and a processing circuit in the housing, the processing circuit including a processor and memory, the memory comprising graphical user interface (GUI) logic, proximity suppression logic and a single-pass neural network model trained with training images corresponding to the endoscopic procedure and defining anatomic references observable in the training images, the processor being configured to process the neural network model, the proximity suppression logic and the GUI logic to present a GUI; and
a physical model, the physical model being a three-dimensional model comprising cavities corresponding to interconnected human cavities,
wherein the neural network model is configured to process in real-time the image sensor images generated by the image sensor, to detect the anatomic references in the image sensor images, and to output a set of anatomic references including identifiers and confidence values representing the likelihoods that the anatomic reference identifiers are correct,
wherein the proximity suppression logic is configured to change the confidence values of the anatomic references in the set of anatomic references based on a prior position of the endoscope to identify an anatomic reference indicative of the current position of the endoscope,
wherein the GUI is configured to receive a user input corresponding to a training mode or an exam mode,
wherein in the training mode the GUI is configured to present navigation feedback, and
wherein in the exam mode the GUI is configured to not present navigation feedback.
2 . The training system of claim 1 , wherein the navigation feedback comprises anatomic reference labels and/or a structured progression score.
3 . The training system of claim 1 , wherein the navigation feedback comprises a structured progression score, wherein the memory comprises object detection logic configured to determine whether the user missed an anatomic reference from an ordered list of anatomic references and to calculate a structured progression score by subtracting values assigned to missed anatomic references from a maximum structured progression score.
4 . The training system of claim 2 , wherein the navigation feedback comprises a diagram of the anatomic references marking viewed anatomic references and visualized anatomic references.
5 . The training of claim 1 , wherein the proximity suppression logic comprises a proximity suppression map including anatomical references and weights corresponding to the anatomical references, the weights predetermined based on a proximity of an anatomical reference to the other anatomical references, wherein the proximity suppression logic matches the prior position of the endoscope with one of the anatomical references in the map and changes the confidence values by multiplying the confidence values and the respective weights.
6 . The training system of claim 5 , wherein the image processing device further comprises a medical device interface configured to receive the image sensor images from the endoscope, the image processing device configured to determine a type of endoscopic procedure, and to select the diagram of the anatomic references, based on the medical device interface.
7 . The training system of claim 1 , wherein the training images are generated in the physical model or a copy thereof.
8 . A training system for an endoscopic procedure, the training system comprising:
an endoscope having an image sensor configured to generate images;
an image processing device including a housing and a processing circuit in the housing, the processing circuit including a processor and memory, the memory comprising graphical user interface (GUI) logic and a single-pass neural network model trained with training images corresponding to the endoscopic procedure and defining a set of anatomic references including anatomic references observable in the training images, the processor being configured to process the neural network model and the GUI logic to present a GUI; and
a physical model, the physical model being a three-dimensional model comprising cavities corresponding to interconnected human cavities,
wherein the neural network model is configured to process in real-time image sensor images generated by the image sensor and to detect the anatomic references in the image sensor images,
wherein the GUI is configured to receive a user input corresponding to a training mode or an exam mode,
wherein in the training mode the GUI is configured to present navigation feedback, and
wherein in the exam mode the GUI is configured to not present navigation feedback.
9 . The training system of claim 8 , wherein the navigation feedback comprises anatomic reference labels and/or a structured progression score.
10 . The training system of claim 8 , wherein the navigation feedback comprises a structured progression score, wherein the memory comprises object detection logic configured to determine whether the user missed an anatomic reference from an ordered list of anatomic references and to calculate a structured progression score by subtracting values assigned to missed anatomic references from a maximum structured progression score.
11 . The training system of claim 9 , wherein the navigation feedback comprises a diagram of the anatomic references marking viewed anatomic references and visualized anatomic references.
12 . The training system of claim 8 , wherein the processor comprises a proximity suppression map including weights for the anatomical references, and wherein the anatomic references of the set of anatomic references include identifiers and confidence values representing likelihoods that the anatomic reference identifiers are correct.
13 . The training system of claim 12 , wherein the processor comprises proximity suppression logic configured to change the confidence values of the anatomic references in the set of anatomic references based on a prior position of the endoscope to identify an anatomic reference indicative of the current position of the endoscope.
14 . The training system of claim 13 , wherein the weights are based on a proximity of an anatomical reference to the other anatomical references, wherein the proximity suppression logic matches the prior position of the endoscope with one of the anatomical references in the map and changes the confidence values by multiplying the confidence values and the respective weights.
15 . The training system of claim 8 , wherein the image processing device further comprises a medical device interface configured to receive the image sensor images from the endoscope, the image processing device configured to determine a type of endoscopic procedure, and to select the diagram of the anatomic references, based on the medical device interface.
16 . The training system of claim 8 , wherein the training images are generated in the physical model or a copy thereof.