JP2014158619A - Detection data processing device and image display device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate an evaluation of a measurement result and improve convenience by providing a detection data processing device capable of highly accurately evaluating a wall thickness, an image display device, a detection data processing program, an image display program, a detection data processing method and an image display method.SOLUTION: The detection data processing device acquires a plurality of line data on the whole periphery of a blood bessel B of a living body in the tube axis direction W1, where the line data have been detected on a shaft diameter direction W2 of a probe P inserted inside the blood bessel B. The detection data processing device detects an inside wall and an outside wall of the blood bessel B from the line data, detects wall thicknesses on the basis of the detected inside wall and the outside wall, and, when a wall thickness is thinner than a reference thickness, determines it as a dangerous and thin area.

Description

  The present invention relates to a detection data processing device that detects, for example, a wall of a tubular organ, and an image display device that displays a detection result as an image.

  Conventionally, various treatments for blood vessels which are tubular organs of living bodies have been performed. For this treatment, it is desired to measure the state of the blood vessels. In addition, as a treatment for eliminating a stenosis of a blood vessel such as a coronary artery, a stent treatment in which a stent is placed in the blood vessel is performed. In this stent treatment, it is required to determine an optimal stent before the treatment and to confirm whether the stent after the treatment is properly placed and the stenosis is resolved.

  Here, an optical ultrasonic tomographic image measurement method capable of obtaining a tomographic image of a blood vessel using optical ultrasonic waves has been proposed (see Patent Document). This method is said to synchronize the optical system and the ultrasonic wave, increase the measurement depth, and obtain high resolution.

However, this method can only determine the state by visually observing the tomographic image. For this reason, it is difficult to understand the boundary portion that becomes the inner wall and the outer wall, particularly the boundary portion of the outer wall, and it is difficult to appropriately evaluate the wall thickness and the like.
In addition, even if the state of the blood vessel in the circumferential direction and the axial diameter direction can be visually determined, it is difficult to imagine the state of the blood vessel in the tube axis direction. For this reason, even if, for example, a stenosis portion of a blood vessel or a thin wall portion is known, it is difficult to understand how and how these continue in the tube axis direction.

JP 2005-224399 A

  In view of the above-described problems, the present invention provides a detection data processing device, an image display device, a detection data processing program, an image display program, a detection data processing method, and an image display method capable of accurately evaluating wall thickness, and measurement results. The purpose is to facilitate the evaluation of the above and improve convenience.

  The present invention provides a data acquisition means for acquiring a plurality of line data detected in the axial direction of a probe inserted in a tubular organ of a living body in the tube axis direction for the entire circumference of the tubular organ, and an inner wall of the tubular organ from the line data. It is a detection data processing device comprising a wall part detecting means for detecting an outer wall and a wall part related data output means for outputting wall part related data based on the detected inner wall and outer wall. The present invention can also be an image display device that displays an image based on the detection result of the detection data processing device, or a detection data processing program, an image display program, a detection data processing method, or an image display method. . The range in which a plurality of line data are collected in the tube axis direction can be a predetermined range determined as appropriate, such as a distance that the probe can pull back or a distance determined by the operator.

  According to the present invention, the wall thickness can be accurately evaluated.

Explanatory drawing of the method of obtaining the detection data by optical coherence tomography. 1 is a block diagram showing a system configuration of an imaging system. Explanatory drawing of the data which deleted the reflection of the catheter. Explanatory drawing explaining the detection method of an inner wall and an outer wall with a graph. Explanatory drawing explaining the detection method of an inner wall, an outer wall, and wall thickness by a tomographic image. Explanatory drawing explaining the method to detect a stent with a graph. Explanatory drawing explaining the complement method of a stent by a tomographic image. Explanatory drawing explaining the detection and complementation of a stent by a three-dimensional image. Explanatory drawing of the three-dimensional image which displays an inner wall, an outer wall, a thin area | region, and a stent. Explanatory drawing of the method of discriminating and imaging a stent at high speed, without using the radiation of Example 2. FIG. The block diagram which shows the system configuration | structure of the stent image display system of Example 2. FIG. Explanatory drawing of the data structure of the template data of Example 2, and stent data. Explanatory drawing of the screen which displays the stent image of Example 2. FIG.

The inventor has invented a method capable of quantitatively evaluating the wall thickness of a tubular organ of a living body, and invented a method capable of easily grasping the state of the wall thickness in the tube axis direction by three-dimensional display.
Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

  This method uses detection data obtained by optical coherence tomography (OCT). FIG. 1 is an explanatory diagram for explaining the basic principle of this method.

  As shown in the enlarged cross-sectional view of FIG. 1A, optical coherence tomography is performed by inserting the catheter K into the blood vessel B and performing line data using near infrared rays irradiated from the probe P in the catheter K in the axial radial direction W2. This is a method for obtaining detection data in which a predetermined range is collected in the tube axis direction W1 with respect to the entire 360 ° circumference in the axial diameter direction W2 as the probe P rotates. In this method, detection data is acquired in the blood vessel B after placement of the stent S. The detection data may be acquired in the blood vessel B before the stent S is placed.

  FIG. 1B shows a detection data display image in which detection data at an arbitrary position in the tube axis direction is displayed, with the vertical axis representing the rotation angle of the probe P and the horizontal axis representing the distance from the probe P (the distance in the blood vessel radial direction). G1. A portion where the reflected signal from the tissue such as the blood vessel B or the stent S is strong is displayed brightly as the high luminance pixel G11. The high-intensity pixel G11 is displayed as undulating according to the distance from the probe P because the probe P is not necessarily present at the center of the radial cross section of the blood vessel B. Note that Lines 1 and 3 illustrated in FIG. 1B are data of lines where the stent S exists, and Line2 is data of lines where the stent S does not exist. In this detection data display image G1, there is also a high luminance pixel G12 in which the catheter K is reflected.

  FIG. 1C shows detection data at an arbitrary angle at an arbitrary position in the axial direction of the tube, with the vertical axis representing luminance (16-bit image luminance) and the horizontal axis representing distance (depth) in the axial diameter direction from the probe P. It is the displayed graph. In the figure, the detection data L1 to L3 on the three lines Line1 to Line3 in FIG. 1B are displayed. Here, the line indicates one direction in the axial direction from the probe P, and the detection data of the line means detection data (line data) in the one direction.

  As shown in the figure, it can be seen that peaks appear not only in the detection data L1 (Line1) and detection data L3 (Line3) of the line with the stent S but also in the detection data L2 (Line2) where the stent S does not exist.

  FIG. 2 is a block diagram showing a system configuration of the imaging system 1. The imaging system 1 includes a detection device 3, and an image processing device 5 that functions as a detection data processing device and an image display device. This imaging system 1 can obtain a tomographic image of a tubular portion (blood vessel, trachea, esophagus, duodenum, large intestine, biliary tract, etc.) of a human body, and in this embodiment, an example of obtaining a tomographic image of a blood vessel will be described. To do.

  The detection device 3 includes a light source 11, a spectrometer 12, a scanning unit 13, a reference unit 14, a detector 15, an amplifier 16, a signal processing unit 17, and a data output unit 18.

  The light source 11 oscillates an optical signal using near infrared rays. The spectroscope 12 separates the optical signal from the light source 11. The scanning unit 13 reflects an optical signal supplied via the spectroscope 12 by 90 ° with a probe, rotates 360 ° in a direction perpendicular to the supplied optical signal, and scans the optical signal. Two-dimensional information indicating the layer structure inside is obtained. This is moved in the tube axis direction to obtain a plurality of two-dimensional information forming a multilayer structure. The reference unit 14 includes a reference mirror. The optical distance from the reference mirror to the spectroscope 12 and the optical distance from the scanning mirror of the scanning unit 13 to the spectroscope 12 are equal. The detector 15 detects the reflected light and obtains a detection signal. The amplifier 16 amplifies the detection signal detected by the detector 15. The signal processing unit 17 performs signal processing on the detection signal and converts it into detection luminance data 71. The data output unit 18 outputs the detected luminance data 71 to an external device such as the image processing device 5.

  The detected luminance data 71 is luminance data of reflected light reflected from near-infrared rays by a vascular tissue or the like, and line data detected in the axial radial direction from the inside of the blood vessel in which the stent is placed is predetermined in the axial direction of the tube for the entire circumference of the blood vessel. Scan data collected in a range (a predetermined number of pages). The line data is luminance data of each depth (distance from the probe center) in the axial direction. The 360 ° line data becomes data for one page (one tomographic data), and a plurality of data (for a plurality of pages) is collected in the tube axis direction to form scanning data. Therefore, a two-dimensional tomographic image of a blood vessel including a stent can be obtained from one page of data, and a two-dimensional tomographic image of the corresponding page can be obtained by designating an arbitrary position in the tube axis direction.

  The image processing device 5 is configured by a computer. The image processing device 5 includes, as hardware, a control unit 21 that executes various controls and various calculations, a storage unit 22 that stores data and programs, a display unit 25 such as a display that displays images, a mouse that receives input operations, and An input unit 26 such as a keyboard and a medium processing unit 27 that reads and writes information from and on a recording medium 29 such as a CD-ROM are provided.

  The control unit 21 can read various programs such as the detection display program 72 recorded in the recording medium 29 from the medium processing unit 27 and install them in the storage unit 22 and execute the installed various programs. The control unit 21 functions as the blood vessel wall detection unit 30, the stent detection unit 50, and the evaluation unit 60 by the detection display program 72 stored in the storage unit 22.

  The blood vessel wall detection unit 30 includes a cut unit 31, a smoothing unit 32, an inner wall detection unit 33 that functions as an inner wall position detection unit, an inner wall smoothing unit 34, an outer wall detection unit 35 that functions as an outer wall position detection unit, and an outer wall smoothing unit. 36, a wall thickness detection unit 37, a thin region detection unit 38 functioning as a dangerous region determination means, an inner wall data storage unit 41, an outer wall data storage unit 42, and a thin region detection data storage unit 43.

  The cut unit 31 is a portion in which the catheter K (see FIG. 1A) having a metal guide wire is reflected as the high luminance pixel G12 (see FIG. 1B) for each line data in the detected luminance data 71. The process which cuts is performed. Specifically, as shown in the image diagram of the detection data display image G2 after cutting in FIG. 3A, for each line data of the detected luminance data 71, the rotation center of the probe P (see FIG. 1C). To the catheter exclusion distance Dp (see FIG. 1C). The catheter exclusion distance Dp is a predetermined constant value, and is preferably longer than the radial distance of the catheter K. Thereby, the post-cut line data D1 shown in FIG. 3A is obtained, and it is possible to prevent the catheter K in the detection data display image G1 from being erroneously detected.

  In addition, the cutting of the catheter K by the cutting part 31 is good also as another method. For example, as illustrated in the image diagram of the detection data display image G3 after the cut in FIG. 3B, the cutting unit 31 includes a measurement target inner distance Di inside the inner wall Wi and a measurement target outer distance Do outside the inner wall Wi. It is possible to adopt a configuration in which data outside the measurement target distance Da is combined. The measurement target inner distance Di and the measurement target outer distance Do are set to appropriate distances such as a predetermined distance, or a predetermined ratio based on the distance from the guide wire to the inner wall Wi. can do. The position of the inner wall Wi is calculated by the method described later. In this way, it is possible to remove in advance portions that are clearly not stents or inner walls.

  The smoothing unit 32 functions as a low-pass filter, and smoothes the post-cut line data D1 (see FIG. 3A). In this embodiment, a moving average is used for smoothing. Specifically, the calculation of averaging data (for example, 21) of a predetermined number (for example, 10) before and after the target location is repeated by moving the target location by a predetermined number (for example, one). At this time, the calculation is speeded up by deleting the predetermined number of data (for example, one) at the end and reading the data of the predetermined number (for example, one) ahead with respect to the traveling direction of the movement. The number of data may be considered as one pixel of resolution to be detected. In this embodiment, 0.17 mm is one pixel (one) when converted in actual scale. By smoothing in this way, the post-cut line data D1 shown in the graph of FIG. 4 (A) is smoothed into the post-smooth line data D2 shown in the graph of FIG. 4 (B).

  4 (A) to 4 (E), in any graph, the vertical axis represents luminance (16-bit image luminance) and the horizontal axis represents the distance (depth) in the axial diameter direction from the probe P. It is the graph which displayed the detection line data of the arbitrary angles in an axial direction position.

  The inner wall detection unit 33 reads the inner wall template data 73 functioning as inner wall determination reference data from the storage unit 22 and, as shown in FIG. 4C, from the inner side (probe P side) of the smoothed post-smoothing line data D2. The first position where the brightness higher than the inner wall template Ti continues over the entire length of the inner wall template Ti is scanned. And the inner wall detection part 33 makes the inner wall position Pi the position (depth) of the axial radial direction in which the inner end of inner wall template Ti is located, when satisfy | filling conditions initially.

  The inner wall template Ti used here has a certain level of high brightness on the inner side (probe P side) in the tube axis direction, and a lower brightness is set on the outer side (outside of the blood vessel) in the tube axis direction. It is formed in a staircase shape. The outer brightness in the tube axis direction of the inner wall template Ti is set to be higher than the brightness of the flat portion behind the peak due to the stent in the detection data at the position where the stent exists.

  The width in the depth direction of the inner wall template Ti is configured to be wider than the actual thickness of the stent, and is preferably at least twice the actual thickness of the stent, more preferably at least three times.

  Thereby, the position of the inner wall of the blood vessel can be accurately detected by grasping the feature that the brightness of the outer wall of the blood vessel is gently lowered. That is, at the position where the stent exists, as shown in FIG. 4D, the smoothed line data D3 appears as a peak with a short width in the tube axis direction. On the other hand, by determining whether or not the predetermined brightness in the tube axis direction is equal to or higher than the predetermined brightness, it can be understood that the rear of the stent is equal to or lower than the brightness of the inner wall template Ti and is not the inner wall. In this way, it is possible to accurately detect the inner wall, excluding stents and the like.

  The inner wall smoothing unit 34 performs smoothing by moving average by collecting the inner wall positions Pi for one round in the blood vessel circumferential direction. The inner wall smoothing unit 34 repeats the smoothing in one round unit in the direction of the blood vessel axis, and executes it for all inner wall positions Pi. The smoothed inner wall position Pi is stored in the inner wall data 74 of the storage unit 22 by the inner wall data storage unit 41. The inner wall smoothing unit 34 obtains smoothed data based on the inner wall position Pi around the stent S that can be detected even for the portion where the inner wall position Pi is not detected due to the presence of the stent S, and the inner wall position Pi is obtained. Can be determined.

  The outer wall detection unit 35 reads the outer wall template data 75 from the storage unit 22, and, as shown in FIG. 4E, the outer wall of the tubular organ that is opposite to the scanning direction of the inner wall with respect to the smoothed line data D2. From the outside, first, a position where the brightness of the outer wall template To (outer wall determination reference data) is obtained is scanned. The outer wall detection unit 35 first sets the position where the luminance of the outer wall template To becomes the outer wall position Po. The scanning of the outer wall may be performed on all the smoothed line data including the smoothed line data D2 of the blood vessel wall part and the smoothed line data D3 of the stent part. You may perform only with respect to D2. The outer wall template data 75 may be one luminance data, but may be a stepped template like the inner wall template data 73.

  When the outer wall is scanned with the outer wall template data 75, which is one luminance data, only for the smoothed line data D2 from which the inner wall has been detected, a simple luminance determination is performed for a portion that is recognized as a blood vessel wall. Since the position of the outer wall can be detected only with this, it can be executed with high accuracy and high speed.

  The outer wall smoothing unit 36 performs smoothing by moving average by collecting the outer wall position Po for one round in the blood vessel circumferential direction. The outer wall smoothing unit 36 repeats the smoothing in one round unit in the direction of the tube axis of the blood vessel, and executes it on all outer wall positions Po. The smoothed outer wall position Po is stored in the outer wall data 76 of the storage unit 22 by the outer wall data storage unit 42. The outer wall smoothing unit 36 obtains data that is smoothed based on the outer wall position Po around the stent S that can be detected even for a portion where the outer wall position Po is not detected due to the presence of the stent S, and the outer wall position Po is obtained. Can be determined.

  By measuring the positions of the inner wall and the outer wall in this way, the inner wall detection unit 33 and the outer wall detection unit 35 can detect the detection data such as the tomographic image of the blood vessel shown in the image diagram of FIG. The inner wall position Pi and the outer wall position Po can be detected as shown in the image diagram of B), and the inner wall position having no variation as shown in the image diagram of FIG. 5C can be detected by the inner wall smoothing unit 34 and the outer wall smoothing unit 36. Pi and outer wall position Po can be acquired. In addition, the inner wall detection part 33, the inner wall smoothing part 34, the outer wall detection part 35, and the outer wall smoothing part 36 function as a wall part detection means.

  As shown in the explanatory diagram of the tomographic image of FIG. 5D, the wall thickness detection unit 37 obtains the inner wall closed curve Ri from the inner wall position Pi for one round after smoothing, and the inner wall closed curve Ri for each inner wall position Pi. The outer wall position Po in the normal line NL direction is obtained, and the distance from the inner wall position Pi to the outer wall position Po is obtained as the wall thickness Wd (the thickness of the blood vessel wall). The wall thickness detection part 37 calculates | requires this wall thickness Wd about 1 round, and also calculates | requires over the full length of a pipe-axis direction.

  Thus, by calculating the distance in the direction of the normal line NL of the inner wall closed curve Ri, the thickness of the blood vessel wall can be measured more accurately than by measuring the distance from the probe P. Further, since the inner wall can be detected more accurately than the outer wall, the wall thickness can be calculated more accurately by taking the normal line NL from the inner wall closed curve Ri rather than taking the normal line from the outer wall closed curve Ro of the outer wall.

  The thin region detection unit 38 determines whether or not the calculated wall thickness is thinner than a predetermined reference thickness, and if it is thin, the inner wall position Pi (or outer wall position Po) is pathologically thin. It is detected as a TCF area determined to be dangerous. FIG. 5E is an image diagram of a tomographic image showing a thin region TCF which is a TCF region (dangerous region). The reference thickness can be an appropriate thickness, for example, 100 microns. The thin area detection unit 38 determines all wall thicknesses and detects the TCF area over the entire circumference of the detected luminance data 71. The detected TCF region is stored in the thin wall region data 77 of the storage unit 22 by the thin region detection data storage unit 43. The display unit 25 that displays the thin region TCF functions as a wall-related data output unit that displays wall-related data.

  The stent detection unit 50 includes a cut unit 51, a narrowing unit 52, a stent primary detection unit 53, a stent complementing unit 54, a template selection unit 56, and a stent data storage unit 57 after noise removal.

  The cut unit 51 reads the detected luminance data 71 from the storage unit 22 and cuts the data from the rotation center of the probe to the catheter exclusion distance Dp and the data below the image luminance Bn in the detected luminance data 71. That is, in the detected luminance data 71, the number of pixels included in each line is calculated such that the distance from the rotation center of the probe is equal to or greater than the catheter exclusion distance Dp and equal to or greater than the image luminance Bn.

The image brightness Bn can be a predetermined brightness determined as appropriate, can be set to a predetermined brightness of 1/2 or less with respect to the maximum value of the template, and is preferably set to a predetermined brightness of 1/4 or less. It is more preferable that the predetermined luminance is about 1/5. As a result, a line having a predetermined number of pixels (number of data) equal to or higher than the image luminance Bn can be set as the stent S candidate, and the processing speed can be improved.
Note that the data cut to prevent erroneous detection of the catheter K is performed by cutting data outside the measurement target distance Da as described in the cut unit 31 instead of cutting the data up to the catheter exclusion distance Dp. It is good also as composition to do.

  The narrowing-down unit 52 stents line data in which the number of pixels calculated by the cut unit 51 is equal to or less than a predetermined ratio (for example, 60% or less) of the average of all lines in the detected luminance data 71 of the line remaining after the cut by the cut unit 51. Extracted as detection data of existence candidates.

  The template selection unit 56 selects an appropriate template from the stent template data 78. This template is selected manually or automatically depending on the type of probe. When selecting automatically, it may be mechanically determined in accordance with the characteristics of the probe, such as the overall brightness of the detection data. Further, the number of templates to be selected is not limited to one, but may be plural. The template selection unit 56 functions as a template switching unit.

  The primary stent detection unit 53 performs matching processing on the detection data narrowed down by the narrowing-down unit 52 using the template selected by the template selection unit 56. In this matching process, a portion where the average luminance error is within a predetermined range (within a threshold value) is detected while moving the template in the depth direction with respect to the luminance profile of the detection data of each line. If it is within the predetermined range, it is determined as a stent. At this time, if the luminance is equal to or higher than the upper limit luminance Bu (Bu is a predetermined value), it is handled as the upper limit luminance Bu. When detection data having an average luminance error within a predetermined range is found, the position of the maximum luminance of the template at the template position at that time is determined as the stent position.

  The template used for the primary detection of the stent is a template that captures the characteristics of data that can appear when detecting the stent S attached to the tubular organ of the living body and matches the peak rear portion of the stent S.

  More specifically, FIG. 6A shows detection of an arbitrary angle at an arbitrary position in the axial direction of the tube, with the vertical axis representing luminance (image luminance) and the horizontal axis representing the distance (depth) in the axial diameter direction from the probe P. It is the graph which displayed data L4 and L5.

  The detection data L4 shows a graph when the stent S just placed in the blood vessel B is detected. As shown in the figure, the detection data L4 shows a peak with a sharp increase in luminance at the depth at which the stent S exists.

  The detection data L5 shows a graph when a site where a blood vessel wall is newly born after a lapse of time after the stent S is placed in the blood vessel B is detected. As shown in the figure, it can be seen that the luminance gradually increases due to the reflection on the blood vessel wall and rapidly increases slightly near the stent S to show a peak.

  In FIG. 6B, the vertical axis is the luminance (image luminance) and the horizontal axis is the distance (depth) in the axial diameter direction from the probe P, and the peak rear portion of the detection data L6 in which the stent S exists is expanded by the stent S. It is the displayed graph.

  As shown in the figure, the template T has a shape in which the brightness rapidly decreases to the slope changing portion T2 in the sharply decreasing portion T1 behind the peak, and the low brightness continues gently in the low brightness portion T3 behind the slope changing portion. is there. Accordingly, the template T has a shape that focuses on the suddenly decreasing portion T1 and the subsequent low luminance portion.

  More specifically, the template T has a shape in which the luminance decreases rapidly from the high luminance along the depth direction, and thereafter changes little and shows low luminance (especially decreases little by little or shows the same luminance). Further, the coordinate C of the detection data corresponding to the first location with the highest luminance in the template is defined as a stent pixel (stent position). Thereby, the line data in which a stent exists and the stent position in the line data can be determined appropriately. The stent position can be determined at a predetermined position in the template depending on the shape of the template, for example, a predetermined position determined from the highest luminance portion of the template (for example, not the same position but a predetermined number of pixels of resolution). A shallow position).

  In matching, if the error M between the detection data L6 and the template T is within a predetermined range, the stent S is determined. Thereby, detection accuracy can be improved.

  When a plurality of points matching the template are found in one detection data, the primary stent detection unit 53 determines the stent position using the best matching point. When a plurality of templates are used, a stent is determined to be a stent if the average luminance error with any template is within a predetermined range (within a threshold value).

  The stent primary detection unit 53 and the cut unit 51 and the narrowing unit 52 described above function as a stent position detection unit. The image processing apparatus 5 including the primary search unit 31, the template selection unit 56, and the stent primary detection unit 53 functions as a stent detection processing apparatus.

  First, the stent complementing unit 54 detects the stent candidate Ss (see FIG. 7A) by making the conditions for detecting a stent gentler than the stent detection by the narrowing-down unit 52 and the stent primary detection unit 53. I do. More specifically, the tolerance value of the average error in the narrowing-down unit 52 is increased to obtain detection data of a stent presence candidate, and the stent primary detection unit 53 detects the detected data, and the detected stent is set as a stent candidate Ss. . At this time, the same position as the already detected stent S is excluded from the stent candidates Ss. The “same position” is determined by the rotation angle of the probe P and the distance from the probe P.

  Next, as shown in the explanatory view of the perspective view of FIG. 7A, the stent complementing unit 54, for the stent candidate Ss, in the tomographic images I1 and I3 before and after the tomographic image I2 in which the stent candidate Ss exists, It is confirmed whether or not the stents S1 and S3 exist within a predetermined range (in the example shown in the circle with the radius Lo) from the same position as the stent candidate Ss, and if it exists in either one of the tomographic images I1 or I3, the stent candidate Ss. Is a stent, and a complement determination process is performed to determine that the stent candidate Ss is not a stent, and the stent S is complemented from the stent candidate Ss. The radius Lo is preferably shorter than the inter-image distance between successive tomographic images, more preferably 3/4 or less of the inter-image distance, and even more preferably about half of the inter-image distance.

  This makes it possible to complement the stent that could not be detected by the primary detection by using the characteristics depending on the shape of the stent that should appear close to the continuous tomographic image I, and improve the detection accuracy.

  For example, when viewing three tomographic images I4 to I6 that are continuous with FIGS. 7B, 7C, and 7D, the tomographic image I4 in FIG. 7B and FIG. In the tomographic image I6, the stent S can be detected correctly, but in the tomographic image I5 in FIG. 7C, the stent S may not be detected in the region E.

  Here, the stent candidate Ss is detected in the tomographic image I7 in FIG. 7E by the complementary candidate extraction process in which the allowable value of the average error is increased. This stent candidate Ss is detected not only in the region E but also outside the region E, and includes those that are erroneously detected.

  From here, by performing complementary determination processing using the previous and subsequent tomographic images I4 and I6, as shown in the tomographic image I8 shown in FIG. The candidate Ss remains, and the stent candidate Ss (see FIG. 7E) that is not continuous with another continuous tomographic image I is excluded as noise.

Thus, by performing the complementation by the stent complementation unit 54 for all the tomographic images I, the undetected stent S can be complemented, and noise associated with the complementation can be excluded.
The effect of this complementation is noticeable when the stent S is displayed in three dimensions. FIG. 8A is a three-dimensional image diagram in which the stent S detected by the stent primary detection unit 53 is three-dimensionally displayed, and FIG. 8B shows all the stent candidates Ss extracted in the complementary candidate extraction process. FIG. 8C is a three-dimensional image diagram in which the stent S after complementation by the stent complementing unit 54 is three-dimensionally displayed. In FIG. 8A, there are many undetected stents and in FIG. 8B there is a lot of noise, but in FIG. 8C, the stents are detected with high accuracy and are clearly displayed.
In addition, in the complementation determination process using the front and rear tomographic images I4 and I6, the supplementation is performed when the stent S is in one of the predetermined ranges of the front and rear tomographic images I4 and I6. It is good also as a structure complemented when the stent S exists in both the predetermined ranges. In addition, the stent may be supplemented even if the stent candidate Ss is not the stent S but exists in each of the predetermined ranges of the front and rear tomographic images I4 and I6. Even in such a configuration, the stent can be appropriately complemented.

  The stent data storage unit 57 stores the detected stent data in the storage unit 22 as stent data 79.

  Note that the cut unit 31 of the blood vessel wall detection unit 30 and the cut unit 51 of the stent detection unit 50 may read the detected luminance data 71 from the recording medium 29 by the medium processing unit 27. The cut unit 31 and the cut unit 51 that read the detected luminance data 71 function as data acquisition means.

The evaluation unit 60 includes a stent state detection unit 61 that functions as a stent attachment state determination unit, and an evaluation data storage unit 63.
The stent state detection unit 61 evaluates the mounting state of the stent S on the inner wall based on the inner wall data 74 and the stent data 79. Specifically, for each tomographic image I, an inner wall closed curve Ri obtained by connecting the continuous inner wall data 74 and a stent closed curve obtained by connecting the continuous stent data 79 are obtained, and the area where the inner wall closed curve Ri and the stent closed curve do not coincide with each other. An average mismatch amount obtained by dividing the area by the circumference of the blood vessel wall (that is, the length of the inner wall closed curve Ri) is calculated, and this average mismatch amount is used as an index indicating an average stent mounting deviation amount. It can be evaluated that the larger the average mismatch amount, the larger the shift amount, and the smaller the average mismatch amount, the smaller the shift amount. The area where the inner wall closed curve Ri and the stent closed curve do not coincide can be calculated, for example, by dividing the area of the stent closed curve from the area of the inner wall closed curve Ri. In this manner, the amount of stent displacement relative to the inner wall can be quantitatively evaluated.

  Further, if the average mismatch amount is larger than a predetermined deviation allowable value, it is preferable to output an alarm indicating that the stent mounting state is inappropriate. Thereby, the user can recognize that the mounting state is inappropriate without checking all the average mismatch amounts in the tube axis direction of the stent.

  The image processing device 5 displays the average mismatch amount numerically, or displays the average mismatch amount on the display unit 25 in an appropriate manner, such as displaying a graph with the horizontal axis as the depth in the tube axis direction and the vertical axis as the average mismatch amount. indicate. The display unit 25 that displays the average mismatch amount functions as a wall-related data output unit that outputs the average mismatch amount as wall-related data.

  The tomographic image I9 shown in the image diagram of FIG. 7G shows that the stent S is close to the inner wall closed curve Ri and is correctly attached. In the tomographic image I10 shown in the image diagram of FIG. 7 (H), it can be seen that the stent S is separated from the inner wall closed curve Ri and is not mounted correctly. In this way, by properly displaying the inner wall closed curve Ri and the stent S on the tomographic images I9 and I10, the suitability of the stent mounting state can be visually evaluated, and quantitative evaluation can be performed by calculating the above-described average mismatch amount.

  As another evaluation method, the distance between the stent S and the inner wall (for example, the distance in the normal direction NL of the inner wall) is calculated, and this distance is obtained for all the stents S in one tomographic image I, and the total distance is added. The value divided by the number of stents S may be used as an evaluation index as an average mismatch amount. Even in this case, the amount of stent displacement relative to the inner wall can be quantitatively evaluated.

  The evaluation data storage unit 63 stores the average mismatch amount of each tomographic image I calculated by the stent state detection unit 61 as the evaluation data 81 in the storage unit 22.

  The storage unit 22 includes detection luminance data 71, detection display program 72, inner wall template data 73, inner wall data 74, outer wall template data 75, outer wall data 76, wall thin region data 77, stent template data 78, stent data 79, and the like. Various data and programs are stored. The detected luminance data 71 is received from the detection device 3 or read from the recording medium 29 by the medium processing unit 27.

The three-dimensional display unit 91 and the two-dimensional display unit 92 functioning as a three-dimensional image constructing unit receive detection luminance data 71, inner wall data 74, outer wall data 76, wall thin region data 77, and stent data 79 from the storage unit 22. Reading is performed, and a three-dimensional image and a two-dimensional image are respectively displayed on the display unit 25.
The three-dimensional display unit 91 arranges the detected luminance data 71, the inner wall data 74, the outer wall data 76, the wall thin region data 77, and the stent data 79 in the axial diameter direction and the tube axis direction to reconstruct a three-dimensional image. Display the original image. Thereby, an inner wall, an outer wall, a thin area | region, and a stent can be freely displayed as a three-dimensional image individually or in combination.

  FIG. 9A is a three-dimensional image diagram that displays the inner wall Wi and the stent S in a longitudinal direction in the tube axis direction, and FIG. 9B displays the inner wall Wi and the stent S across the axial diameter direction. FIG. 9C is a three-dimensional image diagram, and FIG. 9C is a three-dimensional image displaying the outer wall Wo (or inner wall Wi) and the thin region TCF. Here, the inner wall Wi and the outer wall Wo are displayed in pink, red, or the like that makes it easy to image a blood vessel, the stent S is displayed in silver, gray, or the like that makes it easy to image a metal, and the thin region TCF is an inner wall such as yellow or white. Displayed in a color distinguishable from Wi and the outer wall Wo. Thus, the state of the blood vessel and the state of the stent can be displayed in three dimensions, and display at an arbitrary position and an arbitrary angle can be performed according to the operation of the input unit 26. The three-dimensional display unit 91 that displays the thin area TCF functions as a dangerous area specifying means.

  The two-dimensional display unit 92 displays tomographic images I (I1 to I10) by rotating 360 ° of luminance data (line data) of the entire circumference at each position (page number) in the tube axis direction in the detected luminance data 71. To do. The two-dimensional display unit 92 also displays a stent image (for example, a red dot) at a position where the stent is present at the same position (page number) in the tube axis direction among the data stored in the stent data 79. Or images with different colors or brightness such as white dots). Thereby, the stent indwelled in the blood vessel is accurately displayed on the tomographic image (two-dimensional display unit 92) of the blood vessel.

  In addition, the two-dimensional display unit 92 includes a manual adjustment unit 93 and receives an operation input of an operator such as manually adding a stent via the input unit 26. The stent data additionally input by the manual adjustment unit 93 is stored in the stent data 79.

In the detection display program 72, the part that causes the image processing device 5 to perform the functions from the primary search unit 31 to the stent data storage unit 57 is the stent detection processing program.
The image processing device 5 configured as described above functions as a stent image display device.

With the imaging system 1 described above, it is possible to detect the inner wall and outer wall of a biological tubular organ such as a blood vessel.
Moreover, since the imaging system 1 detects an inner wall and an outer wall with a different template, it can each detect with sufficient precision.
In addition, the imaging system 1 uses a stepped template when detecting the inner wall so that the brightness behind the template is higher than the brightness of the flat portion behind the stent, and the width of the entire template in the depth direction is increased. Since it is wider (longer) than the actual thickness, it is possible to prevent erroneous detection of the stent as the inner wall, and the inner wall can be detected with high accuracy.

  Moreover, since the imaging system 1 can detect the wall thickness of a living tubular organ, it can grasp the TCF region. Since the imaging system 1 detects the wall thickness based on the thickness of the inner wall in the normal direction, the wall thickness can be accurately measured regardless of the position of the probe P with respect to a blood vessel that is not a perfect circle, Reliability can be increased.

  Moreover, since the imaging system 1 performs smoothing in the detection of the inner wall and the outer wall, the shapes of the inner wall and the outer wall can be acquired with high accuracy. In particular, the imaging system 1 can effectively detect the wall thickness with high accuracy by performing smoothing on the outer wall where the detection data is likely to vary.

  In addition, the imaging system 1 can display the TCF region as a tomographic image or a three-dimensional image, and allows the user to clearly recognize the position and shape of the TCF region. In particular, by displaying the image as a three-dimensional image, the imaging system 1 allows the user to see the two-dimensional tomographic images in order, and to make the wall thin more clearly than imagined the width of the thin wall region in the tube axis direction. The size and shape of the part can be confirmed.

  Further, the imaging system 1 can determine whether or not the stent is properly mounted, and can quantitatively verify the mounting state of the stent.

  Further, the imaging system 1 can detect the stent with high accuracy by complementing the stent. This complementation can solve the conflicting problem of improving the stent detection rate while preventing noise by looking at the continuity of the stent in successive tomographic images.

  Conventionally, stent treatment in which a stent is placed in a blood vessel has been performed as a treatment for eliminating a stenosis of a blood vessel such as a coronary artery. In this stent treatment, it is required to determine an optimal stent before the treatment and to confirm whether the stent after the treatment is properly placed and the stenosis is resolved.

  Here, an image analysis apparatus and an image analysis method that can easily confirm the validity of the determined stent before treatment have been proposed (see Patent Document 1: Japanese Patent Application Laid-Open No. 2009-240359). In this method, a simulated stent image for prior confirmation is displayed by being superimposed on an image obtained by an intravascular ultrasound (IVUS) apparatus. Thereby, it is supposed that the validity of the stent can be easily confirmed before treatment.

  Also, a method for reconstructing a three-dimensional image of a target has been proposed (see Patent Document 2: Japanese Patent Laid-Open No. 2001-155144). The method includes the steps of acquiring a digital radiographic image with a camera that rotates around the organ, marking the position of the target from image to image, and collecting successively identified targets in successive images. And reconstructing a three-dimensional image of the target from the captured image and position. Thereby, it is supposed that the three-dimensional image of the stent can be reconstructed.

  Further, a storage unit that stores data of a plurality of slice images related to a medical instrument placed in the subject, and an image generation unit that generates data of a plurality of projection images having different projection ranges from the plurality of stored slice images. A three-dimensional image processing apparatus has been proposed (see Patent Document 3: Japanese Patent Application Laid-Open No. 2010-88803). According to this apparatus, it is said that a slice image of a stent or the like can be obtained using X-rays.

However, since the method of Patent Document 1 is to be confirmed in a simulated manner, the state of the stent after placement cannot be confirmed.
In addition, the method of Patent Document 2 does not describe how to identify a stent, and the specific method and the final output are unclear. In addition, exposure is performed because X-ray radiation is used. This can cause problems.
The apparatus of Patent Document 3 uses X-ray radiation. Therefore, the problem of radiation exposure can occur, and it is not preferable when the stenosis has not recurred in the stent placement part periodically.

  The present inventor has studied a method for rapidly imaging and imaging a stent without using radiation. This method uses detection data obtained by optical coherence tomography (OCT). FIG. 10 is an explanatory diagram for explaining this method.

  As shown in the enlarged cross-sectional view of FIG. 10A, optical coherence tomography is performed by inserting the catheter K into the blood vessel B and irradiating the line data with near infrared rays irradiated from the probe P in the catheter K in the axial radial direction W2. , And detection data is obtained by collecting a predetermined range of this line data in the tube axis direction W1 for the entire 360 ° circumference in the axis diameter direction W2. In this method, detection data is acquired in the blood vessel B after placement of the stent S.

  FIG. 10B shows a detection data display image in which detection data at an arbitrary position in the tube axis direction is displayed with the vertical axis representing the rotation angle of the probe P and the horizontal axis representing the distance from the probe P (the distance in the blood vessel radial direction). G1. A portion where the reflected signal from the tissue such as the blood vessel B or the stent S is strong is displayed brightly as the high luminance pixel G11. The high-luminance pixel G11 is displayed as wavy depending on the distance from the probe P because the probe P is not necessarily present at the center of the radial cross section of the blood vessel B. Note that Lines 1 and 3 illustrated in FIG. 10B are data of lines where the stent S exists, and Line2 is data of lines where the stent S does not exist.

  FIG. 10C shows detection data at an arbitrary angle at an arbitrary position in the axial direction of the tube, with the vertical axis representing luminance (16-bit image luminance) and the horizontal axis representing the distance (depth) in the axial direction from the probe P. It is the displayed graph. In the figure, the detection data L1 to L3 on the three lines Line1 to Line3 in FIG. 10B are displayed. Here, the line indicates one direction in the axial direction from the probe P, and the detection data of the line means detection data (line data) in the one direction.

  As shown in the figure, it can be seen that peaks appear not only in the detection data L1 (Line1) and detection data L3 (Line3) of the line with the stent S but also in the detection data L2 (Line2) where the stent S does not exist.

  In FIG. 10D, the vertical axis represents luminance (image luminance), and the horizontal axis represents distance (depth) in the axial diameter direction from the probe P. Detection data L4 and L5 at arbitrary angles at arbitrary tube axis positions. It is the graph which displayed.

  The detection data L4 shows a graph when the stent S just placed in the blood vessel B is detected. As shown in the figure, the detection data L4 shows a peak with a sharp increase in luminance at the depth at which the stent S exists.

  The detection data L5 shows a graph when a site where a blood vessel wall is newly born after a lapse of time after the stent S is placed in the blood vessel B is detected. As shown in the figure, it can be seen that the luminance gradually increases due to the reflection on the blood vessel wall and rapidly increases slightly near the stent S to show a peak.

  In order to reconstruct a two-dimensional tomographic image or a three-dimensional image of the stent S using such detection data, the pixel where the stent S is detected (hereinafter referred to as a stent pixel) must be determined from the detection data. I must.

Here, the stent pixel is characterized in that it has higher luminance than other pixels.
However, there is a problem that peaks appear not only in the detection data L1 and L3 with the stent S but also in the detection data L2 without the stent S. Further, the detection data L2 without the stent S has a wider depth range (range in the horizontal tube axis direction in FIG. 10D) than the detection data L1 and L3 with the stent S, but the blood vessel wall is newly formed. Even in the detection data L5 of the stent S, there is a problem that the depth range where the brightness is high is wide. Furthermore, the peak in the vicinity of the stent S is lower in the detection data L5 of the stent S in which the blood vessel wall is newly formed than in the detection data L4 of the stent S that has just been inserted, and cannot be determined only by the high brightness. There is also. Thus, since the actual luminance of the stent pixel and the shape of the peak near the stent pixel vary depending on the state around the stent S, it is difficult to extract the stent pixel appropriately and at high speed. If the shape of the template for extracting the stent pixel is complicated, there is a problem that the calculation time for extracting the stent pixel becomes long and high-speed operation cannot be expected.

  As a result of intensive research on such data, the present inventors have invented a method for appropriately and quickly discriminating stent pixels.

  In FIG. 10E, the vertical axis is the luminance (image luminance) and the horizontal axis is the distance (depth) in the axial diameter direction from the probe P, and the peak rear portion of the detection data L6 where the stent S is present due to the stent S is enlarged. It is the displayed graph.

  As shown in the figure, the brightness of the sharply decreasing portion T1 behind the peak rapidly decreases to the slope changing portion T2, and the low brightness portion T3 behind the slope changing portion continues to have a low brightness. The stent S was able to be extracted appropriately by creating the template T that focuses on the suddenly decreasing portion T1 and the subsequent low-luminance portion.

  That is, although the brightness profile of the detection data shows various aspects depending on the stent S and the tissue, the present inventor suddenly decreases the brightness at a position deeper than the stent S due to the high reflectance of the stent S with respect to near infrared rays. A feature has been found, and a method of appropriately discriminating the stent S from this feature has been invented.

  The template has a shape in which the luminance decreases sharply from high luminance along the depth direction, and thereafter changes little and exhibits low luminance (particularly, gradually decreases or shows the same luminance). Further, the coordinate C of the detection data corresponding to the first location with the highest luminance in the template is defined as a stent pixel (stent position). Thereby, the line data in which a stent exists and the stent position in the line data can be determined appropriately. The stent position can be determined at a predetermined position in the template depending on the shape of the template, for example, a predetermined position determined from the highest luminance portion of the template (for example, not the same position but a predetermined number of pixels of resolution). A shallow position).

  Further, at the time of matching, if the error M between the detection data L6 and the template T is within a predetermined range, the stent S is determined to be determined. Thereby, detection accuracy can be improved.

  Further, since the luminance of the stent S portion becomes very high, the luminance of the detection data that is equal to or higher than the upper limit luminance Bu is preferably treated as the highest luminance portion of the template. Thereby, the stent S can be discriminated more appropriately.

  Further, in order to improve the calculation speed, a line having a predetermined number of pixels (data number) of the image brightness Bn or higher in the primary search is set as a stent S candidate, and only the candidate data is compared with the template, and within a predetermined error. By distinguishing the data from the stent S, the processing speed was further improved. The image brightness Bn can be a predetermined brightness determined as appropriate, can be set to a predetermined brightness of 1/2 or less with respect to the maximum value of the template, and is preferably set to a predetermined brightness of 1/4 or less. It is more preferable that the predetermined luminance is about 1/5.

  In the primary search, it is preferable to exclude data that is equal to or smaller than the catheter exclusion distance Dp on the shallower side of the detection data. Thereby, the erroneous detection by the probe P and the catheter appearing can be prevented. Here, the catheter exclusion distance Dp can be a predetermined distance that is the same as or slightly longer or slightly shorter than the range in which the catheter including the probe P is shown. Accordingly, it is possible to prevent the data of the stent S from being excluded when the probe P is close to the stent S while eliminating the erroneous detection by the probe P.

Further, in the primary search, by removing data having an image luminance Bn or less that is the magnitude of the base noise, erroneous detection due to noise can be prevented.
When a plurality of points matching the template are found in the comparison with the template in the secondary search (template matching), the point that most matches the template is set as the stent S. Thereby, a stent position can be distinguished appropriately.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 11 is a block diagram illustrating a system configuration of the imaging system 101. The imaging system 101 includes a detection device 3 and an image processing device 105. The imaging system 101 can obtain a tomographic image of a tubular portion of a human body (blood vessel, trachea, esophagus, duodenum, large intestine, biliary tract, etc.). In this embodiment, an example of obtaining a tomographic image of a blood vessel will be described. To do.

  The detection device 3 includes a light source 11, a spectrometer 12, a scanning unit 13, a reference unit 14, a detector 15, an amplifier 16, a signal processing unit 17, and a data output unit 18.

  The light source 11 oscillates an optical signal using near infrared rays. The spectroscope 12 separates the optical signal from the light source 11. The scanning unit 13 reflects an optical signal supplied via the spectroscope 12 by 90 ° with a probe, rotates 360 ° in a direction perpendicular to the supplied optical signal, and scans the optical signal. Two-dimensional information indicating the layer structure inside is obtained. This is moved in the tube axis direction to obtain a plurality of two-dimensional information forming a multilayer structure. The reference unit 14 includes a reference mirror. The optical distance from the reference mirror to the spectroscope 12 and the optical distance from the scanning mirror of the scanning unit 13 to the spectroscope 12 are equal. The detector 15 detects the reflected light and obtains a detection signal. The amplifier 16 amplifies the detection signal detected by the detector 15. The signal processing unit 17 performs signal processing on the detection signal and converts it into detection luminance data 151. The data output unit 18 outputs the detected luminance data 151 to an external device such as the image processing device 105.

  The detected luminance data 151 is luminance data of reflected light reflected from near-infrared rays by a vascular tissue or the like, and line data detected in the axial direction from the inside of the blood vessel in which the stent is placed is predetermined in the tube axial direction for the entire circumference of the blood vessel. Scan data collected in a range (a predetermined number of pages). The line data is luminance data of each depth (distance from the probe center) in the axial direction. The 360 ° line data becomes data for one page (one tomographic data), and a plurality of data (for a plurality of pages) is collected in the tube axis direction to form scanning data. Therefore, a two-dimensional tomographic image of a blood vessel including a stent can be obtained from one page of data, and a two-dimensional tomographic image of the corresponding page can be obtained by designating an arbitrary position in the tube axis direction.

  The image processing apparatus 105 is configured by a computer. The image processing apparatus 105 includes, as hardware, a control unit 130 that executes various controls and various calculations, a storage unit 150 that stores data and programs, a display unit 155 such as a display that displays images, a mouse that receives input operations, and An input unit 156 such as a keyboard and a medium processing unit 157 that reads and writes information from and on a recording medium 158 such as a CD-ROM are provided. The control unit 130 can read various programs such as the stent detection display program 152 recorded in the recording medium 158 from the medium processing unit 157 and install them in the storage unit 150, and execute the installed various programs.

  The control unit 130 uses the stent detection display program 152 stored in the storage unit 150 to perform a data adjustment unit 131, a template selection unit 135, a matching unit 137, a stent data storage unit 138, a three-dimensional display unit 141, and a two-dimensional display unit 142. And the manual adjustment unit 143.

  The storage unit 150 stores various data and programs such as detected luminance data 151, a stent detection display program 152, template data 153, and stent data 154. The detected luminance data 151 is received from the detection device 3 or read from the recording medium 158 by the medium processing unit 157.

  The data adjustment unit 131 includes a cut unit 132 and a narrowing unit 133. The data adjustment unit 131 reads the detected luminance data 151 from the storage unit 150 and executes processing. The detected luminance data 151 may be read from the recording medium 158 by the medium processing unit 157. The data adjustment unit 131 that reads the detected luminance data 151 functions as a data acquisition unit.

  The cutting unit 132 cuts the data from the rotation center of the probe to the catheter exclusion distance Dp and the data below the image luminance Bn in the detected luminance data 151. That is, in the detected luminance data 151, the number of pixels included in each line is calculated such that the distance from the rotation center of the probe is not less than the catheter exclusion distance Dp and not less than the image luminance Bn.

  The narrowing-down unit 133 sets the line data in which the number of pixels calculated by the cut unit 132 is equal to or less than a predetermined ratio (for example, 60% or less) of the average of all lines of the detected luminance data 151 of the line remaining after the cut by the cut unit 132 to the stent. Extracted as detection data of existence candidates.

  The template selection unit 135 selects an appropriate template from the template data 153. This template is selected manually or automatically depending on the type of probe. When selecting automatically, it may be mechanically determined in accordance with the characteristics of the probe, such as the overall brightness of the detection data. Further, the number of templates to be selected is not limited to one, but may be plural. The template selection unit 135 functions as a template switching unit.

  The matching unit 137 executes a matching process using the template selected by the template selection unit 135 on the detection data narrowed down by the data adjustment unit 131. In this matching process, a portion where the average luminance error is within a predetermined range (within a threshold value) is detected while moving the template in the depth direction with respect to the luminance profile of the detection data of each line. If it is within the predetermined range, it is determined as a stent. At this time, if the luminance is equal to or higher than the upper limit luminance Bu (Bu is a predetermined value), it is handled as the upper limit luminance Bu. When detection data having an average luminance error within a predetermined range is found, the position of the maximum luminance of the template at the template position at that time is determined as the stent position.

  When a plurality of matching points are found in one detection data, the matching unit 137 determines the stent position using the best matching point. When a plurality of templates are used, a stent is determined to be a stent if the average luminance error with any template is within a predetermined range (within a threshold value).

  The matching unit 137 and the data adjustment unit 131 described above function as a stent position detection unit. The image processing apparatus 105 including the data adjustment unit 131, the template selection unit 135, and the matching unit 137 functions as a stent detection processing device.

  The stent data storage unit 138 stores the detected stent data in the storage unit 150 as stent data 154.

The three-dimensional display unit 141 and the two-dimensional display unit 142 read out the stent data 154 and the detected luminance data 151 from the storage unit 150 and display a three-dimensional image and a two-dimensional image on the display unit 155, respectively.
The three-dimensional display unit 141 displays a three-dimensional image in which the three-dimensional image is reconstructed by arranging the stent data 154 in the axial diameter direction and the tube axis direction.

  The two-dimensional display unit 142 displays the tomographic image by rotating 360 ° of luminance data (line data) of the entire circumference at each position (page number) in the tube axis direction in the detected luminance data 151. In addition, the two-dimensional display unit 142 displays a stent image (for example, a red dot) at a position where the stent is present at the same position (page number) in the tube axis direction among the data stored in the stent data 154. Or images with different colors or brightness such as white dots). Thereby, the stent indwelled in the blood vessel is accurately displayed on the tomographic image (two-dimensional display unit 142) of the blood vessel.

  In addition, the two-dimensional display unit 142 includes a manual adjustment unit 143 and receives an operation input of an operator who manually adds a stent via the input unit 156. The stent data additionally input by the manual adjustment unit 143 is stored in the stent data 154.

Of the stent detection display program 152, the portion that causes the image processing apparatus 105 to perform the functions from the data adjustment unit 131 to the stent data storage unit 138 is the stent detection processing program.
The image processing apparatus 105 configured as described above functions as a stent image display apparatus.

FIG. 12 is an explanatory diagram for explaining the data structure of the template data 153 and the data structure of the stent data 154.
As shown in FIG. 12A1 to FIG. 12C2, in this embodiment, the template data 153 has a probe A template 153a and a probe B first so that it can correspond to different types of probes A and B. A template 153b and a second template 153c for probe B are included. The probe A template 153a is a probe A template, and the probe B first template 153b and the probe B second template 153c are probes B templates. With respect to the probe B, if it is approximate to one of the first template for probe B 153b and the second template for probe B 153c, it is determined as a stent. These templates 153a to 153c list values obtained by rounding the luminance profile at the rear of the stent by an average value or the like for each predetermined number of pixels.

  As shown in FIG. 12A1, the probe A template 153a is composed of 10 columns of numbers 1 to 10. In each column, values A1, A1, A2, A3, A4, A5, A6, A7, A7, A7 are stored in order from number 1 to number 10. The values A1 to A7 stored in the numbers 1 to 10 are values as shown in the graph of FIG. 12 (A2). Here, the same value is stored in the number having the same sign of the value. Therefore, for the sharply decreasing portion Ta from No. 1 to No. 4 (inclination changing portion Tb), No. 1 and No. 2 continue to have the same value A1, and as shown in FIG. doing. For example, if the value of A1 is “1”, a template with a steep slope that decreases to 1/5 of the maximum value in two steps, such as “7/15” for A2 and “1/5” for A3. It has become. The low-luminance portion Tc after the number 5 has a gentle slope in which the value gradually decreases from A4 to A7, and the final value A7 of the template can be, for example, intensity “0”.

  The primary search data for the probe A template 153a has an upper limit luminance Bu that is about the same as the maximum value of the template, an image luminance Bn that is about 1/5 times the maximum value of the template, and the catheter exclusion distance Dp is It is about 1/30 of the maximum value of the template.

  The first template 153b for the probe B is composed of 20 columns of data 1 to 20 as shown in FIG. 12 (B1). Each column stores values B1, B2, B3, B3, B3, B4, B5, B5, B5, and B6 in order from number 1 to number 10, and numbers 11 to 20 all store B6. Yes. The values B1 to B6 stored in the numbers 1 to 20 are values as shown in the graph of FIG. 12 (B2). Here, the same value is stored in the number having the same sign of the value. Therefore, as shown in FIG. 12 (B2), the value of the sharply lowered portion Ta from number 1 to number 3 (inclination changing portion Tb) is rapidly lowered, and the low luminance portion Tc after number 4 has a gentle value. It is falling. For example, if the value of B1 is “1”, B2 is “65/100”, B3 is “3/10”, and so on, with a steep slope that decreases to about 1/3 of the maximum value in two steps. It is a template. B4 to B6 are gentle slopes that gradually decrease in value, and the last value B6 of the template can be set to, for example, “1/10” of B1.

  The primary search data for the probe B first template 153b has an upper limit luminance Bu that is about the same as the maximum value of the template, an image luminance Bn that is about 1/5 times the maximum value of the template, and a catheter exclusion distance. Dp is about 3/25 of the maximum value of the template.

  As shown in FIG. 12C1, the second template 153c for the probe B is composed of 20 columns of data of number 1 to number 20. Each column stores values C1, C1, C2, C3, C4, C5, C6, C6, C6, and C6 in order from number 1 to number 10, and numbers 11 to 20 all store C6. . Values C1 to C6 stored in numbers 1 to 20 are values as shown in the graph of FIG. 12 (C2). Here, the same value is stored in the number having the same sign of the value. Accordingly, as shown in FIG. 12 (C2), the value rapidly decreases from number 2 in the sharply decreasing portion Ta from number 1 to number 4 (inclination changing portion Tb), and is compared in the low luminance portion Tc from number 5 onwards. It is a gentle decline. For example, if the value of C1 is “1”, C2 is “6/10”, C3 is “3/10”, and so on. It is a template. C4 to C6 are relatively gentle slopes that gradually decrease in value to 1/5, 1/10, and 0, and the final value C6 of the template can be set to “0”, for example.

  The primary search data for the second template for probe B 153c has an upper limit luminance Bu that is about the same as the maximum value of the template, an image luminance Bn that is about 1/5 times the maximum value of the template, and a catheter exclusion distance. Dp is about 3/25 of the maximum value of the template.

  Thus, the last of the template is preferably 1/8 or less, more preferably 1/10 or less of the initial value of the template. Further, it is preferable that a plurality of minimum values continue at the end of the template. Moreover, it is preferable that it becomes smaller than half within a predetermined number of times from the maximum value at the beginning of the template, and more preferably smaller than one third within a predetermined number of times. The predetermined number of times is preferably 10 times (10 steps) or less, more preferably 5 times (5 steps) or less, and more preferably about 3 to 4 times (3 to 4 steps). In addition, the rough shape of the template is a portion that rapidly decreases from the maximum value to half or less (preferably 1/3 or less), and a low-luminance portion (slowly decreases) in a depth range longer than the sharply decreasing depth range. Or a part having the same luminance).

  As shown in FIG. 12D, the stent data 154 includes a number (StrutNo), a page number (intCurrPage), an angle (Angle), a RectX position (Rect_x), a RectY position (Rect_y), a PolarX position (Polar_x), and PolarY. A position (Polar_y) and a specific method (Method) are included.

  Thereby, it is possible to store in which position of which page number the stent is present, whether the stent is automatically detected, manually set, or deleted. More specifically, the detection data is formed by collecting the tomographic image data of the entire circumference, which is obtained by collecting the luminance data (line data) in the depth direction, which is the distance from the probe, for each angle, for a plurality of pages in the tube axis direction. By indicating the stent position in the tomographic image for each page number, it is possible to clearly grasp the stent position in the detection data.

  In the specific method (Method) of the stent data 154, a flag indicating any one of automatic (Auto), manual (Manual), and deletion (Deleted) is stored. This makes it possible to determine whether the image processing apparatus 105 automatically detects a stent (automatic) by calculation, a stent manually registered as a stent (manual), or a deleted stent (deleted).

  13A and 13B are explanatory diagrams of a screen for displaying a stent image. FIG. 13A is a screen image diagram of the two-dimensional image display screen 160, and FIG. 13B is a screen image diagram of the three-dimensional image display screen 180.

  A two-dimensional image display screen 160 shown in FIG. 13A includes a two-dimensional image display unit 173 that displays a two-dimensional image that is a tomographic image, and accepts operations using various buttons described below.

The file open button 161 displays a selection screen for selecting a detection data file, and accepts selection of detection data for which image confirmation and various operations are to be performed.
The file display unit 162 displays the file name of the currently displayed detection data.

  The luminance display unit 163 displays the entire luminance of the two-dimensional image 174 displayed on the two-dimensional image display unit 173. When the brightness slider 164 is moved and the redraw button 168 is clicked, the overall brightness of the two-dimensional image 174 is changed, and the changed brightness is displayed on the brightness display unit 163.

  The contrast display unit 165 displays the overall contrast of the two-dimensional image 174 displayed on the two-dimensional image display unit 173. When the contrast slider 166 is moved and the redraw button 168 is clicked, the overall contrast of the two-dimensional image 174 is changed, and the changed contrast is displayed on the contrast display unit 165.

  The stent display check button 167 displays the stent image 174a on the two-dimensional image 174 displayed on the two-dimensional image display unit 173 when checked, and hides the stent image 174a when unchecked.

  When the redraw button 168 is clicked, the display of the two-dimensional image display unit 173 is redrawn. Accordingly, the two-dimensional image 174 having the luminance changed by the luminance slider 164 and the contrast changed by the contrast slider 166 is displayed on the two-dimensional image display unit 173.

  When the automatic detection button 169 is clicked, the stent is automatically detected by the data adjustment unit 131 (see FIG. 11) and the matching unit 137. At this time, the mode display unit 171 displays “Auto Mode” indicating the automatic mode.

When the manual drawing check button 170 is checked, it is accepted that the stent image is manually drawn on the two-dimensional image 174 displayed on the two-dimensional image display unit 173. The drawn stent data is stored in the stent data 154 (see FIG. 11) as manual drawing data.
The work progress status display unit 172 displays the progress status of the processing work by a slider display.

  The two-dimensional image display unit 173 displays a two-dimensional image 174. This two-dimensional image 174 is a tomographic image of a blood vessel obtained by rotating and arranging 360-degree data of the entire circumference at an arbitrary position in the tube axis direction from the detected luminance data 151. The blood vessel wall 174b or the like that reflects near infrared rays is brightly displayed because of its high luminance. In the portion where the stent image 174a (stent S) exists, most of near-infrared rays are reflected by the stent S, so there is almost no reflection by the body tissue thereafter, and there is a radial shadow toward the outside of the stent image 174a. Is displayed. The probe 174c is shown in the center.

The page slider 175 changes the number of pages of the tomographic image to be displayed (that is, the position in the tube axis direction) by a drag operation.
When the three-dimensional display button 176 is clicked, the three-dimensional image display screen 180 shown in FIG. 13B is displayed.

  When the data save button 177 is clicked, the operation input data (for example, manually added stent data) is saved in appropriate data in the storage unit 150.

  The image save button 178 saves the displayed two-dimensional image 174 in the storage unit 150 (see FIG. 11). At this time, if the stent display check button 167 is checked, an image displaying the stent is stored, and if the stent display check button 167 is unchecked, an image not displaying the stent is stored.

  FIG. 13B is a screen image diagram of the three-dimensional image display screen 180. The three-dimensional image display screen 180 displays the three-dimensional image 186 on the three-dimensional image display unit 185, and can confirm the three-dimensional image 186 from an arbitrary direction by various operations.

When operated, the vertical rotation slider 181 can rotate the three-dimensional image 186 in the vertical direction and change the viewing direction in the vertical direction.
The animation button 182 starts animation when checked, and stops animation when unchecked.

When operated, the size slider 183 enlarges / reduces the display size of the three-dimensional image 186.
When operated, the left / right rotation slider 184 can rotate the three-dimensional image 186 in the left / right direction and change the viewing direction to the left / right direction.

  The three-dimensional image display unit 185 displays the stent 187 as a three-dimensional image 186. In the three-dimensional image 186, the automatically detected stent 187a and the manually input stent 187b have different colors or different luminances. Although FIG. 13B is displayed in a monochrome image, it is difficult to understand, but the stent 187b is displayed darker (darker) than the stent 187a.

  With this three-dimensional image display screen 180, the three-dimensional image 186 can be confirmed in three dimensions from an arbitrary direction. Therefore, it is possible to confirm in detail the arrangement and extent of the stent 187. Since the automatically detected stent 187a and the manually input stent 187b are displayed in different colors, it is possible to clearly determine whether the stent is automatically detected or manually input.

  With the above configuration and operation, the position of the indwelling stent can be detected with high accuracy at high speed without using radiation. In particular, since the template data 153 (153a, 153b, 153c) (see FIG. 12) is data focusing on the portion behind the peak due to the stent, the stent can be appropriately detected regardless of the state of the tissue around the stent.

  In addition, since the template data 153 (153a, 153b, 153c) (see FIG. 12) has a configuration that includes the sudden decrease portion Ta and the low luminance portion Tc, it is possible to detect the stent at higher speed and accuracy.

  Further, since stent candidate data is extracted by primary search and matching is performed using template data 153 (see FIG. 12), stents can be detected at higher speed.

  Further, since template data 153 is prepared in a plurality of types such as a probe A template 153a, a probe B first template 153b, and a probe B second template 153c, a template suitable for the type of probe is prepared. Can accurately detect the stent. The template selection unit 135 can select a template suitable for the probe.

  Moreover, since the brightness | luminance more than upper limit brightness | luminance Bu among detection brightness | luminance data 151 is handled as upper limit brightness | luminance Bu, it can also discriminate | determine appropriately also about a high brightness | luminance stent.

Further, in the primary search, since the data less than the catheter exclusion distance Dp on the shallower depth side is excluded from the detection data, it is possible to prevent erroneous detection due to the appearance of the probe.
In addition, since the primary search excludes data having an image luminance of Bn or less, erroneous detection due to noise can be prevented.

  When multiple points that match the template are found in the comparison with the template in the secondary search (template matching), the point that most matches the template is used as a stent, and one stent is erroneously detected as multiple stents. Can be prevented.

  Further, since the three-dimensional image 186 displays the stent and does not display the blood vessel, the three-dimensional shape of the stent can be clearly confirmed.

  The present invention is not limited only to the configuration of the above-described embodiment, and many embodiments can be obtained.

  The present invention can confirm the inside width of the blood vessel before the stent treatment for expanding the stenotic blood vessel, the wall thickness of the blood vessel, and the like, and can confirm the position, size and spread of the stent after the stent treatment, and the attached state of the stent Available for imaging systems. In addition, it can be used for a device that detects and displays a tubular portion of a living body (blood vessel, trachea, esophagus, duodenum, large intestine, biliary tract, etc.) and a stent placed in the tubular portion.

5, 105 ... Image processing device 22 ... Storage unit 25, 155 ... Display unit 33 ... Inner wall detector 34 ... Inner wall smoother 35 ... Outer wall detector 36 ... Outer wall smoother 38 ... Thin region detector 51 ... Cut unit 52 ... narrowing-down part 53 ... primary stent detection part 56, 135 ... template selection part 61 ... stent state detection part 72 ... detection display program 91 ... three-dimensional display part 131 ... primary search part 137 ... matching part 152 ... stent detection display program 153 T ... Template data 170 ... Manual drawing check button 174a ... Stent image B ... Blood vessel Bn ... Image brightness Dp ... Catheter exclusion distance P ... Probe Ri ... Inner wall closed curve S ... Stent T1, Ta ... Steeply decreasing portion T2, Tb ... Inclination change Portions T3, Tc ... Low luminance portion TCF ... Thin region Ti ... Inner wall template To ... Outer wall template W1 ... Axial W2 ... axial radial Wi ... inner wall Wo ... outer wall Wd ... wall thickness

Claims (18)

Data acquisition means for acquiring a plurality of line data detected in the axial direction of the probe inserted into the tubular organ of the living body in the axial direction of the tubular organ all around,
Wall detection means for detecting an inner wall and an outer wall of the tubular organ from the line data;
A detected data processing apparatus comprising: wall portion related data output means for outputting wall portion related data based on the detected inner wall and outer wall. The wall thickness is calculated from the distance between the inner wall and the outer wall detected from the line data, and if the wall thickness is thinner than a reference thickness, the position of the line data in the tubular organ is considered to be dangerously thin. The detection data processing apparatus according to claim 1, further comprising a dangerous area determination unit that determines a dangerous area. 3D image construction for reconstructing the position of the inner wall or outer wall detected from the line data into a three-dimensional position based on the depth and detection angle of the probe in the tube axis direction to construct a three-dimensional image of the inner wall or outer wall Means,
The detection data processing apparatus according to claim 2, further comprising dangerous area specifying means for distinguishing the dangerous area in the three-dimensional image from other areas. A stent position detecting means for detecting the position of the stent from the line data;
The detection data processing apparatus according to claim 1, 2, or 3, further comprising: a stent attachment state determination unit that determines whether the stent is appropriately attached based on the position of the inner wall and the position of the stent. The stent mounting state determination means includes
A closed curve of the position of the stent over the entire circumference of the tubular organ;
A closed curve of the position of the inner wall over the entire circumference of the tubular organ;
Determine the area that is a mismatch between the closed curve of the stent position and the closed curve of the inner wall position,
5. The detection data processing apparatus according to claim 4, wherein a value obtained by dividing the area by the circumference of the inner wall is determined to be inappropriate if it exceeds the allowable deviation of the stent, and determined to be appropriate if not exceeded. A storage unit that stores inner wall determination reference data that is a reference for determining the inner wall and outer wall determination reference data that is a reference for determining the outer wall;
The wall detection means is
A smoothing unit for smoothing the line data;
The smoothed line data after the smoothing satisfies the standard of the inner wall judgment reference data or scans from the inside of the line data, and detects the position of the inner wall when the standard is satisfied,
An outer wall position detection unit that detects whether or not the smoothed line data after the smoothing satisfies the criterion of the outer wall determination reference data or scans from the outside of the line data and detects the position of the outer wall when the criterion is satisfied. Item 6. The detection data processing device according to any one of Items 1 to 5. The inner wall determination reference data is a template having a width based on a detection range of a predetermined intensity or more continuing for a predetermined range, and the template is configured such that an outer intensity condition is lower than an inner intensity condition. Item 7. A detection data processing apparatus according to Item 6. From the line data, comprising a stent position detecting means for detecting the position where the stent exists using a stent detection template,
The detection data processing apparatus according to any one of claims 1 to 7, wherein the stent detection template is a template that matches a site behind the peak of the stent in the line data. The template for detecting a stent has a maximum value in a front part of the template, a sharply decreasing part that rapidly decreases from the maximum value to a slope changing part in a range of half or less of the maximum value, and a stage after the slope changing part. The detection data processing apparatus according to claim 8, further comprising a low-luminance portion that gradually decreases from the sudden decrease portion. A primary search for extracting, as candidate data, line data in which the number of data that is equal to or greater than a predetermined distance from the center distance and whose luminance is equal to or greater than the predetermined luminance is equal to or less than a predetermined ratio of the average value of all line data. Run,
The detection data processing apparatus according to claim 8 or 9, wherein detection using the stent detection template is performed among the data remaining in the primary search. Prepare multiple types of stent detection templates,
The detection data processing device according to claim 8, further comprising a template switching unit that switches the stent detection template used according to the line data. A stent candidate is detected under a gentler condition than the detection condition of the stent by the stent position detecting means,
If a stent candidate or a stent exists within a predetermined range from a position corresponding to the position of the stent candidate in at least one of the tomographic images before and after the tomographic image in which the stent candidate exists, the stent candidate is not a noise but a stent. The detection data processing device according to any one of claims 8 to 11, further comprising a stent complementing unit that recognizes the stent and complements the stent. The detection data processing device according to any one of claims 1 to 12,
An image display device comprising: an image display unit configured to display an image of an inner wall, an outer wall, or a stent of a tubular organ based on a detected position of an inner wall, an outer wall, or a stent of the tubular organ. The image display device according to claim 13, further comprising manual addition means capable of manually adding a stent position. Computer
Data acquisition means for acquiring a plurality of line data detected in the axial direction of the probe inserted into the tubular organ of the living body in the axial direction of the tubular organ all around,
Wall detection means for detecting an inner wall and an outer wall of the tubular organ from the line data;
A detection data processing program that functions as a wall-related data output unit that outputs wall-related data based on detected inner and outer walls. Computer
Data acquisition means for acquiring a plurality of line data detected in the axial direction of the probe inserted into the tubular organ of the living body in the axial direction of the tubular organ all around,
Wall detection means for detecting an inner wall and an outer wall of the tubular organ from the line data;
The wall thickness is calculated from the distance between the inner wall and the outer wall, and if the wall thickness is thinner than the reference thickness, the risk that the position of the line data in the tubular organ is determined to be a dangerous area is determined to be dangerous thin Area determination means;
An image display program that functions as image display means for displaying an image of the dangerous area. A plurality of line data detected in the axial direction of the probe inserted into the tubular organ of the living body in the axial direction of the tubular organ are obtained in the tube axial direction.
Detecting the inner and outer walls of the tubular organ from the line data;
A detection data processing method for outputting wall-related data based on detected inner and outer walls. A plurality of line data detected in the axial direction of the probe inserted into the tubular organ of the living body in the axial direction of the tubular organ are obtained in the tube axial direction.
Detecting the inner and outer walls of the tubular organ from the line data;
The wall thickness is calculated from the distance between the inner wall and the outer wall, and if the wall thickness is thinner than a reference thickness, the position of the line data in the tubular organ is determined as a dangerous thin area as a dangerous thin area,
An image display method for displaying an image of the dangerous area.