Abstract: A method includes receiving anatomic image data comprising a plurality of graphical units; generating an initial segmented set of the plurality of graphical units associated with an anatomic tree structure; receiving an indication of a first user selected graphical unit, the first user selected graphical unit being outside the initial segmented set of the plurality of graphical units; applying a cost function to determine a first sequence of graphical units from the first user selected graphical unit to the initial segmented set of graphical units; receiving an indication of a second user selected graphical unit, the second user selected graphical unit being selected from the first sequence of the plurality of graphical units; and applying the cost function to determine a second sequence of the plurality of graphical units from the second user selected graphical unit to the initial segmented set of the plurality of graphical units.

Abstract: A method for image segmentation comprises receiving volumetric image data for an anatomical region and generating a first volumetric patch from the volumetric image data. The method also comprises generating a second volumetric patch from the first volumetric patch by weighting a plurality of volumetric units in the first volumetric patch and receiving the second volumetric patch as an input to a convolutional neural network. The method also comprises conducting a down-sampling filter process and conducting an up-sampling filter process within the convolutional neural network.

Abstract: Devices, systems, methods, and computer program products for planning medical procedures are disclosed herein. In some embodiments, a system for planning a medical procedure includes a processor and memory operably coupled to the processor. The memory stores instructions that, when executed by the processor, cause the system to perform operations including receiving a three-dimensional model of an anatomic region of a patient and evaluating whether the three-dimensional model is suitable for use in a registration procedure to be performed in the anatomic region. The evaluating includes: analyzing a span of the three-dimensional model along at least two different directions, determining whether the three-dimensional model includes at least one segmented component corresponding to at least one structure of the anatomic region, and performing a registration between the three-dimensional model and a virtual registration data set.

Abstract: A method for image segmentation comprises receiving volumetric image data for an anatomical region and generating a first volumetric patch from the volumetric image data. The method also comprises generating a second volumetric patch from the first volumetric patch by weighting a plurality of volumetric units in the first volumetric patch and receiving the second volumetric patch as an input to a convolutional neural network. The method also comprises conducting a down-sampling filter process and conducting an up-sampling filter process within the convolutional neural network.

Abstract: A system and method of planning a procedure includes a planning workstation including: a display system; and a user input device. The planning workstation is configured to: display image data via the display system; receive a first user input via the user input device; display via the display system a target of a medical procedure within the displayed image data identified based at least on the first user input; display an interactive image via the display system, the interactive image including the image data, a plurality of connected anatomical passageways, and the identified target; receive a second user input via the user input device; display via the display system a trajectory between the target and an exit point along a nearest passageway of the plurality of connected anatomical passageways identified based at least on the second user input; receive a third user input via the user input device; and adjust the interactive image based at least on the defined trajectory and the third user input.

Abstract: A system and method of planning a procedure includes a planning workstation including: a display system; and a user input device. The planning workstation is configured to: display image data via the display system; receive a first user input via the user input device; display via the display system a target of a medical procedure within the displayed image data identified based at least on the first user input; display an interactive image via the display system, the interactive image including the image data, a plurality of connected anatomical passageways, and the identified target; receive a second user input via the user input device; display via the display system a trajectory between the target and an exit point along a nearest passageway of the plurality of connected anatomical passageways identified based at least on the second user input; receive a third user input via the user input device; and adjust the interactive image based at least on the defined trajectory and the third user input.

Abstract: A system comprises a display system and a control system communicatively coupled to the display system. The control system is configured to receive anatomic image data for an anatomic tree structure and generate an initial segmentation of the anatomic image data. The initial segmentation includes a trunk structure and a branched structure unconnected to the trunk structure. The control system is also configured to determine whether to connect the branched structure to the trunk structure based on an assessment of a shape of the branched structure and an assessment of a relationship between the trunk structure and the branched structure.

Abstract: A medical system may comprise a display system, a user input device, a medical instrument, and a manipulator assembly configured to support and operate the medical instrument. The medical system may also comprise a control system. The control system may be configured to display image data corresponding to a three-dimensional anatomical region via the display system, receive a first user input to generate a first curve in the three-dimensional anatomical region via the user input device, and receive a second user input to generate a second curve in the three-dimensional anatomical region via the user input device. The control system may also be configured to determine an anatomical boundary bounded by the first curve and the second curve. The control system may also be configured to control the manipulator assembly to operate medical instrument using the determined anatomical boundary to limit movement of the medical instrument.

Abstract: A method of building an anatomical branch model comprises receiving anatomical image data comprising a plurality of graphical units associated with an anatomical structure and determining a plurality of parent segments and child segments. The method also comprises determining a set of relationships between the parent segments and the child segments by determining a first set of connection costs of connecting at least one of the parent segments to a first subset of the child segments, the child segments of the first subset are separated from the at least one of the parent segments by one or more gaps, identifying a first child segment from the first subset of the child segments based on a first connection cost, and connecting the first child segment to the at least one parent segment. The method further comprising generating an image of the anatomical branch model based on the determined set of relationships.

Abstract: A method for image segmentation comprises receiving volumetric image data for an anatomical region and generating a first volumetric patch from the volumetric image data. The method also comprises generating a second volumetric patch from the first volumetric patch by weighting a plurality of volumetric units in the first volumetric patch and receiving the second volumetric patch as an input to a convolutional neural network. The method also comprises conducting a down-sampling filter process and conducting an up-sampling filter process within the convolutional neural network.

Abstract: A method for image segmentation comprises receiving volumetric image data for an anatomical region and generating a first volumetric patch from the volumetric image data. The method also comprises generating a second volumetric patch from the first volumetric patch by weighting a plurality of volumetric units in the first volumetric patch and receiving the second volumetric patch as an input to a convolutional neural network. The method also comprises conducting a down-sampling filter process and conducting an up-sampling filter process within the convolutional neural network.

Abstract: A method comprises receiving anatomic image data comprising a plurality of graphical units and generating an initial segmented set of the plurality of graphical units associated with an anatomic tree structure. The method also comprises receiving an indication of a first user selected graphical unit from the plurality of graphical units and applying a cost function to determine a first sequence of the plurality of graphical units from the first user selected graphical unit to the initial segmented set of the plurality of graphical units.

Abstract: A method of building an anatomical branch model comprises receiving anatomical image data comprising a plurality of graphical units associated with an anatomical structure and determining a plurality of parent segments and child segments to include a plurality of linked graphical units from the plurality of graphical units. The method also comprises determining a set of relationships between the parent segments and the child segments and generating an image of the anatomical branch model based on the determined set of relationships between the parent segments and the child segments.

Abstract: A system and method of planning a procedure includes a planning workstation including: a display system; and a user input device. The planning workstation is configured to: display image data via the display system; receive a first user input via the user input device; display via the display system a target of a medical procedure within the displayed image data identified based at least on the first user input; display an interactive image via the display system, the interactive image including the image data, a plurality of connected anatomical passageways, and the identified target; receive a second user input via the user input device; display via the display system a trajectory between the target and an exit point along a nearest passageway of the plurality of connected anatomical passageways identified based at least on the second user input; receive a third user input via the user input device; and adjust the interactive image based at least on the defined trajectory and the third user input.

Abstract: A method for image segmentation comprises receiving volumetric image data for an anatomical region and generating a first volumetric patch from the volumetric image data. The method also comprises generating a second volumetric patch from the first volumetric patch by weighting a plurality of volumetric units in the first volumetric patch and receiving the second volumetric patch as an input to a convolutional neural network. The method also comprises conducting a down-sampling filter process and conducting an up-sampling filter process within the convolutional neural network.