JP5643363B2 - Medical image display device and medical image display program - Google Patents

Description

  The present invention relates to a medical image display device and a medical image display program, and more particularly, to improve the efficiency of diagnosis when a doctor interprets a medical image such as an X-ray image, CT image, or MR image. The present invention relates to a medical image display device and a medical image display program.

2. Description of the Related Art Conventionally, an image degeneration method is known in which an image with a large number of pixels is degenerated by simple sampling or arithmetic sampling to generate an image with a small number of pixels (see Patent Document 1).
On the other hand, a display device that independently drives three sub-pixels constituting one pixel of an LCD is known (see Patent Document 2).
JP-A-10-327402 JP 2003-228337 A

For example, there is a case where a medical image captured using an FPD (Flat Panel Detctor) having 4096 × 3328 pixels is desired to be displayed on an LCD having 1600 × 1200 pixels.
In this case, if the medical image of 4096 × 3328 obtained by FPD is reduced to an image of 1477 × 1200 using the image reduction method of Patent Document 1, it is displayed on an LCD having 1600 × 1200 pixels. It is possible to make it. Note that the number of short-axis pixels of the reduced image is 1200 because it matches the number of short-axis pixels of the LCD, and the number of long-axis pixels of the reduced image is 1477 because the aspect ratio of the original image is preserved. It is.
However, since 2.774 (= 4096/1477 = 3328/1200) pixels are reduced to one pixel of the LCD in each of the short axis direction and the long axis direction of the original image, the original resolution is large. There is a problem that is damaged.
Specifically, in breast cancer diagnosis, it is important to discriminate minute lesions such as microcalcifications of 100 μm or less, and the pixel pitch of the FPD is 70 μm, which has a necessary resolution. However, in the LCD, since the pixel pitch converted to the pixel pitch of the FPD is 194 (= 70 * 2.774) μm, there is a problem that a necessary resolution cannot be obtained.

Here, one pixel of the LCD of the monochrome liquid crystal display device is composed of three sub-pixels arranged in the major axis direction of the LCD. Therefore, if the LCD subpixels are driven independently as in the display device of Patent Document 2, the resolution in the major axis direction of the LCD can be improved.
However, the shape of the subpixel is a rectangle, and there is a problem that the aspect ratio of the original image cannot be preserved when the image reduction method disclosed in Patent Document 1 on the premise of a square or circular pixel is applied.

  Accordingly, an object of the present invention is to provide a medical image display device and a medical image display program capable of improving the efficiency of diagnosis when a doctor interprets a medical image such as an X-ray image, CT image, or MR image. It is to provide.

In the first aspect, the present invention relates to an original image of m rows and n columns in a monochrome original medical image as q rows (3 * n * q / m) where m, n, and q are natural numbers and m> q. Image conversion means for converting into a display image composed of pixels in a column, and monochrome liquid crystal for displaying the display image by independently driving three rectangular sub-pixels constituting each square pixel arranged in the row direction of the monochrome LCD A medical image display device comprising a display device, wherein the image conversion means performs a weighted average operation on the original image in a column direction so that all pixels of the original image are reflected in the display image, and m rows Medical image display apparatus characterized in that the number of columns is changed from n columns to (3 * n * q / m) by performing interpolation calculation in the row direction so that the aspect ratio of the original image is maintained while decreasing from n to q rows I will provide a.
In the medical image display apparatus according to the first aspect, the original image of m rows and n columns is degenerated and converted into a display image composed of pixels of q rows (3 * n * q / m) columns. Since the number of pixels in the row direction corresponding to the aspect ratio (n / m) of the original image is (n * q / m), the number of pixels in the row direction of the display image is equal to the aspect ratio (n / m) of the original image. The number of corresponding pixels is three times. However, in each pixel of the monochrome LCD, three subpixels are arranged in the row direction, and the subpixels are driven independently to display a display image. Therefore, the number of pixels in the row direction of the LCD that displays the display image is (3 * n * q / m) / 3 = (n * q / m). Therefore, the aspect ratio of the display image is (n * q / m) / q = (n / m). That is, the aspect ratio of the original image is stored.
Then, the resolution in the row direction can be tripled compared to the case where the subpixels are not driven independently. Since the LCD sub-pixel pitch converted to the FPD pixel pitch is n / (3 * n * q / m) = m / (3 * q) times the FPD pixel pitch, for example, the FPD pixel pitch is If 70 μm, m = 3328, and q = 1200, the subpixel pitch of the LCD converted to the pixel pitch of FPD is 65 (= 70 * 3328 / (3 * 1200)) μm. As a result, the resolution of 100 μm necessary for breast cancer diagnosis can be sufficiently handled. That is, when a doctor interprets a medical image such as an X-ray image, a CT image, or an MR image, the diagnosis can be made more efficient.

In a second aspect, the present invention provides a square in which three m rows and n columns of original images in a monochrome original medical image are arranged in the row direction of the monochrome LCD when m, n, and q are natural numbers and m> q. In order to independently drive the rectangular sub-pixels constituting each of the pixels and display them on the monochrome liquid crystal display device, the computer displays the original image consisting of pixels of q rows (3 * n * q / m) columns. Image conversion means for converting into an image, and display image output means for outputting the display image to the monochrome liquid crystal display device so as to display the display image by independently driving subpixels of the monochrome LCD A medical image display program, wherein the image conversion means reduces the original image from m rows to q rows by performing a weighted average operation in the column direction so that all pixels of the original image are reflected in the display image. And a medical image display program comprising a step of interpolating in the row direction so as to maintain the aspect ratio of the original image and changing from n columns to (3 * n * q / m) columns To do.
In the medical image display program according to the second aspect, the original image of m rows and n columns is degenerated and converted into a display image composed of pixels of q rows (3 * n * q / m) columns. Since the number of pixels corresponding to the aspect ratio (n / m) of the original image is (n * q / m), the number of pixels in the row direction of the display image is the pixel corresponding to the aspect ratio (n / m) of the original image. Three times the number. However, in each pixel of a monochrome LCD, three subpixels are arranged in the row direction, and these subpixels are driven independently to display a display image. Therefore, the number of pixels in the row direction for displaying the display image is ( 3 * n * q / m) / 3 = (n * q / m). Therefore, the aspect ratio of the display image is (n * q / m) / q = (n / m). That is, the aspect ratio of the original image is stored.
Then, the resolution in the row direction can be tripled compared to the case where the subpixels are not driven independently. Since the LCD sub-pixel pitch converted to the FPD pixel pitch is n / (3 * n * q / m) = m / (3 * q) times the FPD pixel pitch, for example, the FPD pixel pitch is If 70 μm, m = 3328, and q = 1200, the subpixel pitch of the LCD converted to the pixel pitch of FPD is 65 (= 70 * 3328 / (3 * 1200)) μm. As a result, the resolution of 100 μm necessary for breast cancer diagnosis can be sufficiently handled. That is, when a doctor interprets a medical image such as an X-ray image, a CT image, or an MR image, the diagnosis can be made more efficient.

  According to the image display device and the image display program of the present invention, the resolution in the long axis direction of the LCD can be improved, and the reduced image can be displayed by storing the aspect ratio of the original image. When a medical image such as an X-ray image, a CT image, or an MR image is subjected to interpretation diagnosis, the efficiency of the diagnosis can be improved.

1 is a configuration explanatory diagram illustrating a medical image display device according to Embodiment 1. FIG. It is explanatory drawing which shows the pixel arrangement | sequence of medical image data, and the pixel arrangement | sequence of LCD. It is explanatory drawing which shows LCD and the image shrunk | reduced. It is the elements on larger scale which show the sub pixel of LCD. It is a flowchart which shows the procedure of the medical image viewer process which concerns on Example 1. FIG. FIG. 6 is a flowchart subsequent to FIG. 5. It is explanatory drawing which shows the pixel arrangement | sequence of a medical image data, and the pixel arrangement | sequence of a short-axis weighted average image. It is explanatory drawing which shows a short-axis weighted average calculation process. It is explanatory drawing which shows the linear interpolation process of a short-axis direction. It is explanatory drawing which shows the pixel arrangement | sequence of a short-axis weighted average image and the pixel arrangement | sequence of a short-axis interpolation intermediate image. It is explanatory drawing which shows the linear interpolation process of a major axis direction. It is explanatory drawing which shows the pixel arrangement | sequence of a short-axis interpolation intermediate image, and the pixel arrangement | sequence of a display image. It is explanatory drawing which shows the frequency emphasis process of a display image. It is explanatory drawing which shows the pixel arrangement | sequence of a display image, and how it looks on an actual display screen.

Hereinafter, the present invention will be described in more detail with reference to embodiments shown in the drawings. Note that the present invention is not limited thereby.
Example 1
FIG. 1 is a configuration diagram illustrating a medical image display apparatus 1 according to the first embodiment.
The medical image display device 1 is composed of pixels of m rows and n columns, where m, n, q, and P are natural numbers and m ≦ n, q <P, n / m <P / q, and m> q. A medical image viewer program that inputs monochrome medical image data A (i, j), degenerates it, converts it into display image data M (k, L) consisting of q rows (3 * n * q / m) columns, and outputs it. A monochrome liquid crystal having a personal computer 3 that executes 31 and a monochrome LCD composed of pixels of q rows and P columns and that displays display image data M (k, L) by independently driving the sub-pixels arranged in the row direction of the monochrome LCD And a display device 4.
The monochrome medical image data A (i, j) is captured by an FPD composed of pixels of m rows and n columns of the medical image capturing device F and stored in the medical image data server D.

  When m> n, the major axis direction of the medical image data A (i, j) is often the vertical direction of the subject, so the major axis direction of the LCD is often the vertical direction. However, for convenience of explanation, it is assumed that the medical image data A (i, j) and the LCD are horizontally long.

FIG. 2A is a numerical example of monochrome medical image data A (i, j), where m = 3328 and n = 4096.
(B) of FIG. 2 is a numerical example of LCD, and is q = 1200 and P = 1600.

FIG. 3 is an explanatory diagram showing the relationship between the LCD and the display image data M (k, L).
Since there is a relationship of n / m <P / q, the monochrome medical image data A (i, j) is degenerated so that the minor axis direction of the display image data M (k, L) matches the minor axis direction of the LCD. . When the aspect ratio n / m of the monochrome medical image data A (i, j) is maintained, the number of pixels p in the major axis direction of the display image data M (k, L) is shorter than the number of pixels P in the major axis direction of the LCD. .
For example, if m = 3328, n = 4096, q = 1200, and P = 1600, then p = 4096 * 1200/3328 = 1477.

  As shown in FIG. 4, one pixel U of the LCD is composed of three subpixels u1, u2, u3 arranged in the major axis direction of the LCD. Since the pixel U is a square, the shape of the subpixels u1, u2, and u3 is a rectangle.

FIG. 5 and FIG. 6 are flowcharts showing the procedure of the medical image viewer process by the medical image viewer program 31 of the personal computer 3.
In step S 0, monochrome medical image data A (i, j) composed of m rows and n columns of pixels is read from the medical image data server 2.
In step S1, the short axis degeneration rate ry = m / q is obtained from the number m of pixels in the short axis direction of the monochrome medical image data A (i, j) and the number of pixels q in the short axis direction of the LCD. For example, if m = 3328 and q = 1200, then ry = 2.774.

  In step S2, a short axis weighted average weight b = (ry−1) * B is obtained from the short axis degeneration rate ry and the adjustment value B. For example, if ry = 2.774 and B = 0.12, then b = 0.129. By adjusting the adjustment value B, the operator can adjust the appearance of the display image (image noise or the like).

In step S3, the medical image data A (i, j) and the short axis weighted average weight b are used so that all the pixels of the medical image data A (i, j) are reflected in the display image. Short-axis weighted average image data A ′ (i, j) is obtained by weighted average of (i, j) in the short-axis direction.
FIG. 7 shows the concept of the medical image data A (i, j) and the short axis weighted average image data A ′ (i, j).
As shown in FIG. 8 and the following equation, the three adjacent medical image data A (i, j-1), A (i, j), and A (i, j + 1) are weighted and short axis weighted. Average image data A ′ (i, j) is obtained because the smallest integer larger than the assumed short axis degeneration rate ry = 2.774 is “3”.
A '(i, j) = A (i, j-1) * b + A (i, j) * (1-2 * b) + A (i, j + 1) * b
However, it is assumed that A ′ (i, 0) = A (i, 0) and A ′ (i, m−1) = A (i, m−1).
i = 0,1, ..., n-1, j = 0,1, ..., m-1
In general, when an integer larger than the minor axis degeneration rate ry is “2 * g + 1”, the adjacent medical image data A (i, jg),..., A (i, j + g) are weighted averaged. The short axis weighted average image data A ′ (i, j) may be obtained. The short axis weighted average weight may be appropriately adjusted according to “g”.

  In step S4, the image sequence coordinate i is initialized to 0.

  In step S5, the LCD minor axis coordinate L = 0 is initialized.

In step S6, an image reference short axis coordinate jq = int (L * ry) corresponding to the LCD short axis coordinate L is obtained.
In step S7, the short axis interpolation weight Wy (L) = (L * ry) −jq is obtained.

In step S8, as shown in FIG. 9 and the following equation, the short-axis interpolated intermediate image data C () is obtained by linear interpolation using the short-axis weighted average image data A ′ (i, j) and the short-axis interpolation weight Wy (L). i, L).
C (i, L) = A '(i, jq) * (1-Wy (L)) + A' (i, jq + 1) * Wy (L)

  In steps S9 and S10, steps S6 to S8 are repeated until L = q-1.

In steps S11 and S12, steps S5 to S10 are repeated until i = n-1.
As described above, as shown in FIG. 10, the short-axis weighted average image data A ′ (i, j) is converted into the short-axis interpolated intermediate image data C (i, L).

Proceeding to FIG. 6, in step S13, the LCD major axis pixel number p = n / ry, which is the number of pixels actually used for display in the major axis direction of the LCD, is obtained. For example, if n = 4096 and ry = 2.774, then p = 1477.
In step S14, in order to independently drive the subpixels in which three pixels are arranged in the major axis direction of the LCD, the number of subpixels (p * 3) obtained by multiplying the number of pixels p by three (p * 3) is used. Find n / (p * 3). For example, if n = 4096 and p = 1477, rx = 0.9244.

  In step S15, the LCD major axis coordinate k = 0 is initialized.

In step S16, an image reference long axis coordinate ip = int (k * rx) corresponding to the LCD long axis coordinate k is obtained.
In step S17, the long axis interpolation weight Wx (k) = (k * rx) −ip is obtained.
In step S18, the LCD minor axis coordinate L = 0 is initialized.

In step S19, as shown in FIG. 11 and the following equation, the display image data M ′ (k, L) is obtained by linear interpolation using the short-axis interpolation intermediate image data C (i, L) and the long-axis interpolation weight Wx (k). )
M ′ (k, L) = C (ip, L) * (1−Wx (k)) + C (ip + 1, L) * Wx (k)

  In steps S20 and S21, step S19 is repeated until L = q−1, and the process proceeds to step S22.

  In steps S22 and S23, steps S16 to S21 are repeated until k = p * 3-1, and the process proceeds to step S24. As a result, as shown in FIG. 12, display image data M ′ (k, L) is obtained.

In step S24, in order to improve the resolution in the minor axis direction, frequency enhancement processing by weighting calculation of three points is performed on the display image data M ′ (k, L), and the final display image data M (k, L) is obtained. obtain. As a result, final display image data M (k, L) is obtained as shown in FIG.
M (k, L) = M ′ (k, L−1) * e + M ′ (k, L) * (1-2 * e) + M ′ (k, L + 1) * e
However, it is assumed that M (k, 0) = M ′ (k, 0) and M (k, q−1) = M ′ (k, q−1).
k = 0,1, ..., p * 3-1, L = 0,1, ..., q-1
Here, e is a weighting coefficient in the frequency enhancement process, and is adjusted between, for example, −0.1 to −0.2 in consideration of the balance with the major axis direction and the impression of the displayed image.

  In step S25, the display image data M (k, L) is output from the personal computer 3 to the monochrome liquid crystal display device 4.

  In the monochrome liquid crystal display device 4, the subpixels of the LCD are driven independently to display the display image data M (k, L) as shown in FIG.

According to the medical image display device 1 and the medical image viewer program 31 of the first embodiment, the following effects can be obtained.
(1) The number of sub-pixels in the major axis direction actually used for display on the LCD is (p * 3), but when converted into the number of pixels, it is (n * q / m). Therefore, the aspect ratio of the display image is (n * q / m) / q = (n / m). That is, the aspect ratio of the original medical image data A (i, j) is stored.
(2) The resolution in the major axis direction can be tripled compared to the case where the subpixels are not driven independently. Since the LCD sub-pixel pitch converted to the FPD pixel pitch is n / (3 * n * q / m) = m / (3 * q) times the FPD pixel pitch, for example, the FPD pixel pitch is If 70 μm, m = 3328, and q = 1200, the subpixel pitch of the LCD converted to the pixel pitch of FPD is 65 (= 70 * 3328 / (3 * 1200)) μm. As a result, the resolution of 100 μm necessary for breast cancer diagnosis can be sufficiently handled.
(3) From the above, when a doctor makes an interpretation diagnosis of a medical image such as an X-ray image, CT image, or MR image, the diagnosis can be made more efficient.

  The medical image display apparatus and medical image display program of the present invention can be used to improve the efficiency of diagnosis when a doctor interprets a medical image such as an X-ray image, CT image, or MR image.

1 medical image display device 3 personal computer
4 Monochrome LCD 31 Medical image viewer program

Claims (6)

When m, n, and q are natural numbers and m> q, the m-by-n converted image in the monochrome original medical image is converted into a display image composed of q rows (3 * n * q / m) columns of pixels. Medical image display comprising: an image converting means for displaying and a monochrome liquid crystal display device for independently driving the rectangular sub-pixels constituting each of the square pixels arranged in the row direction of the monochrome LCD to display the display image A device,
The image converting means, the aspect ratio of the object to be converted image and converts the q-line from m rows said to be converted image in the column direction so that all the pixels is reflected on the display image of the converted image is maintained As described above, when converting from n columns to (3 * n * q / m) columns in the row direction, the column direction conversion is performed by using a plurality of pixels adjacent in the column direction as reference pixels. The number of pixels in the column direction is reduced to q pixels by weighted average calculation processing for generating substantially the same number of pixels from a plurality of pixels and column direction degenerate interpolation calculation, and conversion in the row direction is obtained by calculation in the column direction. A medical image display device that generates a display image by a row direction interpolation calculation using a plurality of pixels adjacent in the row direction as reference pixels . The medical image display apparatus according to claim 1, wherein the column direction degenerate interpolation calculation performs linear interpolation using two pixels adjacent in the column direction as reference pixels . The image conversion means uses a weighted average calculation process in the column direction with an integer value larger than m / q as a reference pixel number, where m / q is a degeneration rate when reducing the number of rows from m rows to q rows. The medical image display device according to claim 1, wherein: When m, n, and q are natural numbers and m> q, three rectangular subpixels are formed by arranging three m rows and n columns of converted images in the monochrome original medical image in the row direction of the monochrome LCD. In order to drive the pixels independently and display them on a monochrome liquid crystal display device,
Computer
The displays the display image by driving the image conversion means for converting the target converted image on the display image composed of pixels q rows (3 * n * q / m ) column, and the monochrome LCD subpixels independently Display image output means for outputting the display image to the monochrome liquid crystal display device,
A medical image display program for functioning as
The image converting means
Row direction so that all the pixels aspect ratio of the object to be converted image and converts the q rows from m rows said to be converted image in the column direction as reflected in the display image is maintained in the object to be converted image comprising the step of converting the n column (3 * n * q / m ) column,
The conversion in the column direction is performed by weighted average calculation processing for generating substantially the same number of pixels from a plurality of pixels in the column direction in the converted image using a plurality of pixels adjacent in the column direction as reference pixels, and column direction degenerate interpolation calculation. The number of pixels in the column direction is reduced to q pixels, and the conversion in the row direction generates a display image by a row direction interpolation calculation using a plurality of pixels adjacent in the row direction obtained by the column direction calculation as reference pixels. <br/> A medical image display program characterized by the above. The medical image display program according to claim 4, wherein the column-direction degenerate interpolation calculation performs linear interpolation using two pixels adjacent in the column direction as reference pixels . The image conversion means uses a weighted average calculation process in the column direction with an integer value larger than m / q as a reference pixel number, where m / q is a degeneration rate when reducing the number of rows from m rows to q rows. The medical image display program according to claim 4, wherein:

Images