Methods and apparatus are provided for enhancement of bronchial airways representations using vascular morphology.
In one embodiment, a representation of the bronchial airways of a subject and a representation of the corresponding vascular network are initially obtained. In the bronchial airways representation, spatial coordinates are identified of termini of airways where portions of the airways are not included in the representation though they are in fact continuous airways in the subject's anatomy. A vessel portion is then identified in the vascular network representation that corresponds to an unrepresented airway portion between first and second identified termini of airways in the bronchial airways representation. With this data, an enhanced bronchial airways representation is produced that reflects an airway extending between the first and second termini. The enhanced bronchial airways representation can be displayed on a monitor of a medical apparatus to facilitate bronchoscopic procedures.
In the vascular network representation, locations that correspond to the identified termini of airways in the bronchial airways representation can be identified. When two identified locations in the vascular network representation are connected by a vascular path of a blood vessel represented in the vascular network representation, the identified termini can be selected as the first and second termini and used to enhance the bronchial airways representation, reflecting an airway extending therebetween.
The method can be performed so that the bronchial airways representation and the vascular network representation are obtained from a single computed tomography (CT) scan. The bronchial airways representation can be generated by setting a Hounsfield unit threshold for region growing segmentation of the CT scan for airway detection. The enhanced bronchial airways representation can be generated by setting a lower Hounsfield unit threshold for region growing segmentation of the CT scan around detected vessels both in terms of location and orientation of detected vessels. Alternatively, the bronchial airways representation and the vascular network representation can be obtained from separate CT scans. The CT scan used to obtain the vascular network representation can be enhanced via a contrast agent.
A display of the vascular network representation can be used by a physician or other professional to manually identify a vessel portion in the vascular network representation that corresponds to an unrepresented airway portion. Alternatively, the identifying of locations in the vascular network representation that correspond to the termini of airways in the bronchial airways can be determined by a processor.
An example apparatus for providing an enhanced representation of bronchial airways of a subject includes a memory configured to store an initial data representation of the bronchial airways of the subject. The data representation includes spatial coordinates of termini of airways where the airways continue in the subject, but continuations of the airways are not included in the initial bronchial airways representation. The memory is further configured to store a data representation of a vascular network corresponding to the bronchial airways of the subject. The apparatus also includes a processor configured to identify a vessel portion in the vascular network representation that corresponds to an unrepresented airway portion between first and second identified termini of airways in the bronchial airways representation. The processor is further configured to produce an enhanced bronchial airways representation that reflects an airway extending between the first and second termini.
To do this, the processor can be configured to identify locations in the vascular network representation that correspond to the identified termini of airways in the bronchial airways representation. When two identified locations in the vascular network representation are connected by a vascular path of a blood vessel represented in the vascular network representation, the processor can be configured to select the two identified termini as the first and second termini for which the enhanced bronchial airways representation reflects an airway extending therebetween. In this way, the apparatus can identify a vessel portion in the vascular network representation that corresponds to an unrepresented airway between first and second identified termini of airways in the bronchial airways representation.
The apparatus can include a display configured to display the enhanced bronchial airways representation. The display can be a monitor of a medical apparatus configured to assist navigation of a bronchoscope in a bronchoscopic procedure. The display can also be used to manually identify a vessel portion in the vascular network representation that corresponds to an unrepresented airway portion.
The memory can be configured to store the bronchial airways representation and the vascular network representation obtained from a single computed tomography scan. The processor can be configured to generate this bronchial airways representation by setting a Hounsfield unit threshold for region growing segmentation of the CT scan for airway detection. The processor may generate the enhanced bronchial airways representation by setting a lower Hounsfield unit threshold for region growing segmentation of the CT scan around detected vessels both in terms of location and orientation of detected vessels. Alternatively, the memory can be configured to store the bronchial airways representation and the vascular network representation as obtained from separate computed tomography scans. The memory can be further configured to store the vascular network representation as obtained from a computed tomography scan enhanced via a contrast agent.
Other objects and advantages of the invention will be apparent to those of skill in the art from the following detailed description and accompanying drawings.
The present embodiments are illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings.
Documents incorporated by reference in the present patent application may include terms that are defined in a manner that conflicts with the definitions made explicitly or implicitly in the present specification. In the event of any conflicts, the definitions in the present specification should be considered to be controlling.
Computed tomography (CT) scans are well known in the medical arts. CT scans are used not only for diagnostic purposes, but also to aid in the navigation of catheters and/or other probes for relatively non-invasive medical procedures. The CARTOâ„¢ mapping system produced by Biosense Webster Inc., of Diamond Bar, Calif., for which various details are described in U.S. Patent Application Publications 2009/0093806 and 2009/0138007 and U.S. Pat. Nos. 5,391,199, 5,443,489, 5,558,091, 5,983,126, 5,944,022, 6,172,499, and 6,177,792, 6,788,967, and 6,690,963, employs CT scan data to assist in the navigation of catheters and other probes in a subject's heart. A CT scan series of essentially two dimensional segments are analyzed to provide a three-dimensional segmentation representation of the heart for this purpose.
For non-invasive bronchial diagnostic and other medical procedures, it is desirable to provide an accurate bronchial airways representation of a subject's airways, i.e. the subject's bronchial tree. While the chambers of the heart are relatively large, the airways of the lungs form a tree, as illustrated in
As a result, the conventional use of CT scans to provide a three-dimensional representation of the airways of a subject's lungs commonly reflect inaccurate breaks or discontinuities of airways within a model derived from a CT scan. When such a representation is to be used for the navigation of a catheter or probe into distal regions of the airway tree, conventional airways representations show dead ends where in fact an airway continues in the subject.
Due to the complexity of the lungs, when a lung CT scan is segmented, sometimes a portion of the anatomy is missed. Further, when lung segmentations are inaccurate, the various airways may not appear to be connected or a physician may not be able to recognize where the airways are coming from, where they go, or how they are connected. The resulting anatomical model can be misleading to the physician. Thus, physicians can benefit from a refined approach to detect the entirety of most airways in a segmentation representation of a patient's bronchial airways.
Applicants have recognized that the use of the morphology of a corresponding representation of a subject's lungs' blood vessels can provide an enhanced airways representation. Blood vessels are often more accurately detected and represented than airways when a CT scan is conducted. Additionally, techniques such as introducing a contrast agent intravenously to a patient may be used to further enhance radiologic representation of vasculature. As discussed in more detail below, information regarding blood vessels can be used to extrapolate better information regarding unrepresented airways in the vicinity of the vessels to produce an enhanced representation of the subject's bronchial airways.
The inventors have recognized that vessels in the lungs are always accompanied by airways so that when a vessel is identified, there should always be a corresponding airway in its vicinity. Accordingly, by using the identification of corresponding vessels, missing parts of a segmentation representation of bronchial airways can be completed to provide an enhanced segmentation representation.
The sequence of segments derived from segmentation and its associated image and position coordinate data can be interpreted in several ways to provide an enhanced view of the subject's internal organs on a display 28 and/or saved to a memory 29 for use in medical diagnostics or other medical procedures. The three-dimensional aspect of a CT scan representation is generally derived from X, Y image coordinates relative to each cross section or segment with a Z spatial coordinate derived from the position of the given segment in the series of segments that make up the scan.
To provide an enhanced segmentation representation for airways segmentation 30, the morphology of a three dimensional representation of the vascular network that tracks the bronchial airway tree of the subject is utilized.
From the segment 51 illustrated in
By correlating the spatial data of the airways segmentation 30 and vascular network representation 40, the airways segmentation 30 can be improved to provide an enhanced airways segmentation representation. As illustrated in
An example method for producing the enhanced representation of bronchial airways of a subject, such as enhanced segmentation 60, is provided with reference to
A representation of a vascular network corresponding to the bronchial airways of the subject is also obtained, step 74. For example, vascular network representation 40 can be obtained of the subject 25 using CT system 20. However, it is not essential that the vascular representation be made at the same time or using the same method since the spatial coordinates of vessel branching can be coordinated with spatial coordinates of airway branching to provide sufficient correspondence data to perform the method. For example, the vascular representation can be obtained from a separate computed tomography scan that has been enhanced via a detectable substance such as, for example, a contrast agent.
In step 76, a vessel portion in the vascular network representation is identified that corresponds to an unrepresented airway portion between first and second identified termini of airways in the bronchial airways representation. To do this, locations in the vascular network representation can be identified that correspond to the identified termini of airways in the bronchial airways representation. For example, if segment 51 is the last segment where airways 35a and 35b can be ascertained, that segment can define the spatial location of the respective termini 34a, 34b. A search of the segment 51 can then be made for the closest blood vessel, in order, for example, to identify the vessel locations 45a, 47a, illustrated in
Similarly, if segment 53 is the first segment where airways 37a and 37b can be ascertained, that segment can define the spatial location of the respective termini 36a, 36b. A search of the segment 53 can then be made for the closest blood vessel, in order, for example, to identify the vessel locations 45b, 47b, illustrated in
Upon a condition that two identified locations in the vascular network representation are connected by a vascular path of a blood vessel represented in the vascular network representation, the bronchial airways representation can be enhanced by indicating an airway extending between the two termini that correspond to the vascular path of the two identified locations. For example in
In step 78, an enhanced bronchial airways representation is produced that reflects an airway extending between the first and second termini that have accordingly been determined as the ends of the unrepresented airway. For example, in the enhanced airways representation 60, airway portion 61 is provided between airways 35a, 37b. Similarly, an airway portion 62 is provided between airways 35b, 37a. This can be accomplished, for example, by setting a lower Hounsfield unit threshold for region growing the segmentation from the CT scan around detected vessels both in terms of location and orientation of detected vessels.
In one example, a processor and associated memory, for example the processor 26 and the associated memory 29 of
The memory 29 can be configured to store a data representation of the bronchial airways of the subject 25 that identifies spatial coordinates of termini of airways where the airways continue in the subject, but continuations of the airways are not included in the bronchial airways representation. The memory can also be configured to store a data representation of a vascular network corresponding to the bronchial airways of the subject.
The processor 26 can be configured to identify a vessel portion in the vascular network representation that corresponds to an unrepresented airway portion between first and second identified termini of airways in the bronchial airways representation. The processor 26 can then be configured to produce an enhanced bronchial airways representation that reflects an airway extending between the first and second termini that can be displayed on the display 28.
To do this the processor 26 can be configured to identify locations in the vascular network representation that correspond to the identified termini of airways in the bronchial airways representation, such as corresponding vascular locations 45a, 45b, 47a and 47b in
The bronchial airways representation enhancement system is not dependent on being incorporated into a CT-Scan system such as the system 20 depicted in
The example medical apparatus 80 includes a controller 82 that controls the positioning of a bronchoscope 84 via a robotic handle 86 for conducting a diagnostic or other bronchoscopic procedure. Navigation of a distal end of the bronchoscope 84 is facilitated by displaying a representation of the airways of a patient on a system monitor 88.
The controller 82 includes processor and memory components for facilitating the operation of the medical apparatus 80. In such a system 80, data for previously generated bronchial airways and vascular network representations can be stored in memory of the controller 82 which can be configured to display those representations on the system monitor 88. The controller 82 can be configured to process the previously generated bronchial airways representation referencing the morphology of the previously generated vascular network representation to produce and display an enhanced bronchial airways representation as discussed above. This is highly advantageous when it is desired to navigate the bronchoscope 86 to a position in the patient's lungs that must traverse portions of airways that are not represented in the previously generated bronchial airways representation.
The examples discussed above are non-limiting. Although representations based on CT-scan are used for the above examples, the invention is applicable to representations derived from use of other technologies as well. Other variations and embodiments will be apparent to those skilled in the art and are within the scope of this disclosure.