JP2009219610A - Organ area specifying method and organ area specifying instrument - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To automatically set a threshold value suitable for original image data changing in accordance with subjects, cases, conditions for imaging or the like by making the changing of a set value by a diagnostician unnecessary. <P>SOLUTION: The organ area specifying method includes a process of setting a threshold value corresponding to a target organ, a process of extracting pixels by comparing an pixel value with the threshold value, and a process of specifying the area of the target organ on the basis of extracted pixels. In the process of setting the threshold value, a threshold value is set for every target organ from the pixel value given to the pixels of the original image data. The relation between the pixel value given to a prescribed organ other than the target organ and a threshold value to specify the target organ is obtained in advance. By using the correlation, a threshold value suitable for the target organ of the original image data is set. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a method and apparatus for identifying a region where an organ is present from medical image data, and can be applied to the identification of an organ from these image data when making a diagnosis using a PET image or CT image. It is.

  In general, attempts have been made to substitute a computer for a diagnosis process performed by a doctor using medical test data (Non-Patent Document 1), but it is difficult to relate the complexity of the human body and the test data to medical knowledge. So, it is not effective.

  In addition, it is difficult to put a diagnosis into practice using a computer using a method different from a diagnostic method by a doctor from the viewpoint of reliability. Therefore, in diagnosis support, the use of a computer is generally auxiliary such as image conversion and image processing.

  In addition, a method for diagnosing in-vivo information represented by CT, MRI, and PET using a tomographic image has been proposed. In the medical field, for example, it is known to perform a medical diagnosis using a tomographic image obtained by a tomographic imaging apparatus such as a PET image. In medical diagnosis using image data such as tomographic images and numerical data, medical experts interpret captured images and extract and infer suspicious areas by comparing them with specialized knowledge and experience.

  In a PET-CT image, the PET image represents the distribution of a drug (FDG) that easily collects in cancer, and the X-ray CT image (hereinafter referred to as CT image) represents the hardness of the tissue by the transmittance of radiation. In cancer diagnosis using whole body FDG-PET (FDG-Positron Emission Tomography) images, radioactive drugs similar to glucose called FDG (fluorodeoxyglucose) are labeled in the body with radioactive fluorine (f-18) as a label. The FDG-PET image is obtained by photographing the distribution in the body. The degree of accumulation of FDG is represented by a value called SUV, and the part where there is a cancer shows a high SUV value.

Extraction of a region suspected of cancer from a PET image utilizes the feature that FDG accumulates at a site with high sugar metabolism and the feature that cancer basically has high sugar metabolism and easily accumulates FDG.
EH Shortliffe, “Computer based Medical Consultations: MYCIN”, American: Elsevier, 1976

  In whole body PET diagnosis, tomographic imaging from the thigh to the top of the head is performed at intervals of about 3 mm in the body axis direction, and about 300 tomographic images (slice images) are taken for each patient. In this whole-body PET diagnosis, since the SUV value appears high even in the absence of cancer in an inflamed site or an organ / tissue (such as kidney, bladder, liver, etc.) that is likely to take up FDG, the diagnostician correctly diagnoses It takes a lot of burden to do.

  While group screening based on whole-body PET diagnosis is being considered, interpretation of tomographic images requires a long time and a great amount of labor as described above, and it is expected that it cannot cope with the diagnosis of a large number of people.

  As a technique for reducing the burden on the diagnostician related to the interpretation of the tomographic image, it is conceivable to cause the computer to perform a diagnosis based on the interpretation of the tomographic image performed by the diagnostician.

  When diagnosis is performed using FDG, even if FDG is physiologically normal, sugar metabolism is performed, but it accumulates at a high site, and therefore, normal / abnormal judgment criteria differ depending on the organ. Therefore, when a computer interprets a PET image, it is necessary to first identify an organ region with significantly different glucose metabolism, and then extract a region suspected of abnormality for the identified organ.

  FIG. 1 shows the degree of accumulation of FDG accumulated in each organ in the PET image and the degree of radiation transmission of each organ in the CT image. Although the PET image and the CT image have different pixel value size ranges depending on each organ, they partially overlap each other, so that it is difficult to specify an organ region only by threshold processing.

  Therefore, first, an approximate region of the organ is extracted by threshold processing. The approximate region obtained at this stage may be extracted in a state where a plurality of adjacent organs are connected without being separated. Therefore, next, individual organs are specified from a plurality of extracted organs by applying the shape and anatomical characteristics of the obtained general organs.

  At this time, the threshold value used for the threshold value process for extracting the approximate region first is an important factor in the organ specifying process performed thereafter. For example, when the threshold value is small, extraction is performed in a state where a plurality of organs are strongly connected, so that it is difficult to specify the target organ. On the other hand, when the threshold value is large, it is avoided that a plurality of organs are extracted in a strongly connected state, but since the number of extracted pixels decreases, it is difficult to specify the region and shape of the target organ. Become.

  In addition, since the pixel value of the original image data of the medical image obtained by the diagnostic device has an intensity difference (grayscale difference) depending on various conditions such as the subject's case and measurement environment, imaging conditions such as the setting parameters of the imaging device, An appropriate threshold value for the threshold processing varies depending on each condition. For this reason, it is difficult to perform appropriate threshold processing with a uniformly determined threshold, and it takes a lot of labor and time for the diagnostician to set the threshold according to each condition.

  Therefore, when specifying the region of the organ from the original image data of the medical image obtained by the diagnostic device, the threshold value used for comparison with the pixel value to extract the pixel of the target organ from the original image data, the subject, the case, The purpose is to set according to fluctuations in the original image data due to imaging conditions and the like.

  It is another object of the present invention to automatically set a threshold value suitable for original image data that changes according to a subject, a case, an imaging condition, and the like, without requiring a threshold setting change by a diagnostician.

  The present invention is used for comparison with a pixel value in order to extract pixels of a target organ from original image data when specifying each region where a plurality of target organs exist from the original image data of a medical image obtained by a diagnostic device. A threshold is set based on each original image data. Thereby, even if the original image data changes according to the subject, the case, the imaging condition, and the like, an appropriate threshold can be set by linking the threshold.

  The present invention includes an aspect of an organ area specifying method and an aspect of an organ area specifying device.

  The aspect of the organ region specifying method of the present invention includes a step of setting a threshold corresponding to each target organ, a step of extracting a pixel by comparing a pixel value with the threshold, and a target organ based on the extracted pixel. And a step of specifying a region. Here, the original image data includes a plurality of pixels and pixel values assigned to the pixels. The step of setting the threshold value sets the threshold value for each target organ from the pixel value assigned to each pixel of the original image data for each original image data. Thereby, even when the original image data changes, the threshold value can be set in conjunction with each other.

  In the aspect of the organ region specifying method of the present invention, the threshold value can be set by a plurality of forms. In the first method form, a threshold value is set using pixel values of an organ outside the target organ, and in the second and third method forms, a threshold value is set using a histogram of the target organ.

  In the first method mode, a correlation between a pixel value assigned to a predetermined organ other than the target organ and a threshold value for specifying the target organ is obtained in advance, and the target organ of each original image data is obtained using this correlation. Set a threshold suitable for.

  In the first method mode, in the step of setting a threshold included in the organ region specifying method, a pixel value assigned to a pixel of a predetermined organ is obtained from original image data, and the threshold for the pixel value of the predetermined organ is determined based on the correlation. The obtained threshold value is set as a threshold value for specifying the target organ.

  For the acquired original image data, the correlation obtains a pixel value to be given to a predetermined organ other than the target organ, and obtains a plurality of threshold values for specifying the target organ by a conventional threshold processing, and determines a predetermined value other than the target organ. This is obtained by statistically obtaining the relationship between the pixel value assigned to the organ and the threshold value for specifying the target organ. This correlation is obtained before the target organ extraction process is performed on the original image data to be diagnosed. This correlation can determine threshold values of a plurality of target organs with respect to pixel values of a predetermined organ.

  This correlation represents the relationship between the pixel value of the predetermined organ and the threshold value of the target organ. By using this correlation, the threshold value of the target organ can be obtained by acquiring the pixel value of the predetermined organ for the original image data to be diagnosed.

  Here, the predetermined organ may be a cerebellum or a lung, for example. Since the SUV value of FDG accumulated in the cerebellum is higher than that of other organs, it can be extracted separately from other organs according to the cerebellar shape conditions without using a threshold. In this embodiment, a threshold for extracting other organs is obtained from the correlation with other organs obtained in advance using the extracted SUV value of the cerebellum as a reference.

  In addition, since it is easy to specify the formation of a lung from a CT image, the lung can be extracted separately from other organs according to the lung shape condition without using a threshold value. In this embodiment, a threshold for extracting other organs is obtained from the correlation with other organs obtained in advance using the extracted SUV value of the lung as a reference.

  The target organ is specified by extracting pixels having pixel values equal to or greater than this threshold. In addition, the pixel value of a predetermined organ can reduce an error due to variations in pixel values by using an average value of pixel values of a plurality of pixels corresponding to the cerebellum and lungs.

  In the second and third method forms, a histogram representing the relationship between the pixel value assigned to the pixel of the original image data and the number of pixels to which the pixel value is assigned is formed, and a threshold is set using this histogram. .

  Since the histogram obtained from the original image data is obtained by superimposing the histograms of each organ, the histogram peak position obtained from the original image data substantially coincides with the peak position of the histogram of each organ, and from the original image data, The obtained histogram profile corresponds to the histogram pattern superposition of each organ.

  In the second method form, a pixel value corresponding to the peak position of the number of pixels is obtained in a histogram formed from original image data, and this pixel value is set as a threshold value for specifying the target organ. The target organ is specified by extracting pixels having pixel values equal to or greater than this threshold.

  In the third method mode, the distribution of pixel values of the target organ is obtained by applying the profile pattern of each organ to the histogram formed from the original image data, the variance is obtained from the distribution of the pixel values of the profile pattern, and this variance The threshold value is determined by using the threshold value to identify the target organ. In the third method mode, upper and lower pixel values are set with a peak value interposed therebetween, and the pixel values are set as upper and lower threshold values. The upper and lower pixel values sandwiching the peak value specify the target organ by extracting pixels existing between the upper and lower threshold values determined by the dispersion of the profile pattern.

  The aspect of the organ region specifying device of the present invention is the device configuration of the aspect of the organ region specifying method of the present invention.

  An aspect of the organ region specifying device of the present invention is a device that specifies each region where a plurality of target organs exist from original image data of a medical image obtained by a diagnostic device, and the original image data of the medical image obtained by the diagnostic device And a storage unit that stores a region of the target organ obtained by specifying the organ from the original image data, and a calculation unit that obtains the region of the target organ and stores it in the storage unit. The original image data is configured to include a plurality of pixels and pixel values assigned to each pixel. The original image data can be, for example, PET image data. In the case of a PET image, an SUV value is assigned to the pixel.

  The calculation unit includes: a threshold setting unit that sets a threshold corresponding to each target organ; a comparison / extraction unit that extracts a pixel by comparing the pixel value with the threshold; and a region of the target organ based on the extracted pixel. And an area specifying unit to be specified.

  For each original image data, the threshold setting unit sets a threshold for each target organ from the pixel value assigned to each pixel of the original image data. Since the threshold set for each target organ is set based on the original image data, it can be set in conjunction with fluctuations due to the subject, case, imaging conditions, and the like.

  The comparison / extraction unit compares the pixel value of each pixel with the threshold set by the threshold setting unit, and extracts a pixel having a pixel value equal to or greater than the threshold. The extracted pixel has an image value equal to or greater than a threshold value. When a different threshold is set for the target organ, it can be used as an index to which target organ the pixel in the original image data belongs, and the threshold is a set of pixels extracted by the same threshold. The target organ can be specified as the set target organ region.

  In addition, since the pixel values of the pixels of each organ are distributed with spread, the pixel value sizes may overlap between adjacent organs. In this case, it is difficult to specify the target organ by specifying the region based on the threshold value. In such a case, the region of the target organ is specified with reference to the arrangement and shape characteristics of the organ.

  The aspect of the organ region specifying device of the present invention can be the first to third device forms corresponding to the first to third method forms.

  The 1st form of the organ area | region identification apparatus of this invention is an apparatus structure corresponding to a 1st method form. In the first embodiment, the storage unit stores a correlation between a pixel value assigned to a predetermined organ other than the target organ and a threshold value for specifying the target organ. In the first mode, the pixel value assigned to the pixel corresponding to the predetermined organ is read from the original image data stored in the storage unit. The read pixel value is used for reading a threshold value using the correlation.

  The threshold setting unit obtains a threshold for specifying the target organ by applying the pixel value given to the predetermined organ to the correlation, and sets the obtained threshold as a threshold for specifying the target organ.

  Here, in the correlation stored in the storage unit, the predetermined organ is, for example, the cerebellum or the lung, and the pixel value given to the predetermined organ can be an average value of the pixel values corresponding to the cerebellum or the lung. The threshold value of the target organ set corresponding to the average value of the pixel values can be read out. This correlation represents, for example, a relationship between a pixel value of an SUV value indicating the degree of FDG accumulation and a threshold for identifying an organ. The relationship between the pixel value and the threshold value can be determined for each organ.

  The 2nd form of the organ area | region identification apparatus of this invention is an apparatus structure corresponding to a 2nd method form. In the second embodiment, the pixel value assigned to the pixel corresponding to the predetermined organ is read from the original image data stored in the storage unit, the pixel value assigned to the pixel, and the number of pixels to which the pixel value is assigned A histogram forming unit that forms a histogram representing the relationship between the two and a peak detection unit that obtains a pixel value corresponding to the peak of the number of pixels in the histogram. The threshold setting unit sets the pixel value obtained by the peak detection unit as a threshold for specifying the target organ.

  The third embodiment of the organ region specifying device of the present invention is a device configuration corresponding to the third method embodiment. In the third embodiment, the storage unit stores, for the target organ, a histogram profile pattern representing a relationship between a pixel value assigned to a pixel and the number of pixels to which the pixel value is assigned.

  The third form reads out the pixel value assigned to the pixel corresponding to the predetermined organ from the original image data stored in the storage unit, the pixel value assigned to the pixel, and the number of pixels to which the pixel value is assigned, And a distribution calculation unit for obtaining a distribution of pixel values of the target organ by applying a profile pattern stored in the storage unit to the histogram formed by the histogram formation unit. The threshold value setting unit calculates a threshold value from the distribution of pixel value distributions calculated by the distribution calculation unit, and determines a threshold value for specifying the target organ from the calculated threshold value.

  As described above, according to the present invention, when specifying an area of an organ from original image data of a medical image obtained by a diagnostic device, it is suitable for original image data that changes according to a subject, a case, and measurement conditions. The threshold value can be automatically set without being operated by the operator.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 2 is a flowchart illustrating a diagnostic procedure using a PET-CT image. In a diagnosis using a PET image and a CT image (X-ray CT image), when a computer is used for diagnosis support, an organ region to be diagnosed is first identified from the PET image (S1). A region suspected of being abnormal is extracted based on the diagnostic criteria determined for each organ (S2).

  The PET image is an image in which the degree of FDG accumulation is represented by the SUV value given to each pixel, and the organ cannot be identified and the target organ cannot be specified with the original image data obtained by the diagnostic device. Therefore, in the step S1, it is necessary to specify an organ region from the original image data.

  A more detailed procedure for the step of identifying the organ region from the original image data of S1 will be described with reference to the flowchart of FIG.

  For a target organ to be identified, a threshold value of an SUV value for identifying the organ is set (S11), compared with a threshold value set for the SUV value assigned to each pixel of the original image data, and an SUV equal to or greater than the threshold value. A pixel having a value is extracted.

  Usually, the original image data is composed of a plurality of pixels and pixel values assigned to the pixels. In the case of a PET image, an SUV value is assigned to each pixel as a pixel value. Since the SUV value indicating the degree of FDG accumulation varies depending on the organ, the organ can be identified from the original image data by comparing the SUV value of each pixel with a threshold corresponding to each organ.

  Therefore, the SUV value assigned to each pixel of the original image data is compared with a threshold value, and a pixel (pixel) having an SUV value equal to or greater than the threshold value is extracted to form an extraction region (S12, S13).

  However, when the SUV value of each organ has a spread, and there are a plurality of organs in the original image data and there is an overlap in the spread of the SUV value of each organ, the organ is classified only by comparison with a threshold value. Are difficult to identify. Therefore, the formed extraction area may include other organs in addition to the target organ. Therefore, using the position and volume of the organ, a candidate organ candidate region is specified for the extracted region (S14).

  A procedure for identifying a candidate area from the extracted area will be described with reference to the flowchart of FIG. 4 and the explanatory diagram of FIG. Here, the case where the PET image includes the heart, the liver, and the kidney (the right kidney and the left kidney) is shown.

  Through S11 to S13 described above, an extraction region is formed from a PET image using a threshold value, and candidate regions are specified for the formed extraction region using organ features such as the position and volume of the organ.

  First, a candidate region including the heart, liver, and right kidney is identified from the position and volume of the organ (FIG. 5A) (S21), and then the left kidney is identified from the position and volume of the organ. (FIG. 5B) (S22).

  Among the candidate regions identified in the step S21, the heart, liver, and right kidney are identified in a state where they are not identified. On the other hand, the left kidney is specified in step S22. Since the positional relationship between the right kidney and the left kidney is known, the right kidney is identified using the positional relationship of the left kidney from the candidate region including the heart, liver, and right kidney identified in step S21 (Fig. 5 (c)) (S23).

  Since the right kidney and the left kidney are identified by the steps of S22 and S23, a candidate region including the heart and liver is obtained by removing the kidney from the candidate region including the heart, liver, and right kidney. S24).

  Finally, the heart region is separated from the candidate region excluding the kidney using the shape of the heart, and the liver region is specified (FIG. 5 (d)) (S25).

  The present invention relates to an organ region for setting a threshold value for specifying a target organ in S11, and uses a histogram of the target organ, a first form in which a threshold value is set using pixel values of an organ outside the target organ. The second and third modes for setting the threshold can be performed in a plurality of modes.

  Hereinafter, the first embodiment will be described with reference to FIGS. 6 to 11, the second embodiment will be described with reference to FIGS. 12 to 19, and the third embodiment will be described with reference to FIGS. 20 to 23.

  First, the first embodiment will be described. FIG. 6 is a diagram showing a relationship between a pixel value (SUV value) and a threshold value for specifying an organ, and FIG. 7 is a diagram for explaining a procedure for obtaining a threshold value for specifying an organ from the pixel value (SUV value). FIG. 8 is a diagram for explaining a configuration example of the organ region specifying apparatus according to the first embodiment of the present invention, and FIG. 9 explains an operation of organ region specifying according to the first embodiment of the present invention. FIG. 10 and FIG. 11 are diagrams for explaining the organ region specifying operation according to the first embodiment of the present invention.

  FIG. 6 is a diagram showing the relationship between the SUV value representing the amount of FDG accumulated in the cerebellum and the threshold value for identifying the liver and kidney from the original image data.

  The figure shown in FIG. 6 obtains a plurality of original image data under different measurement conditions such as a subject, a case, and an imaging condition, obtains an SUV value of the cerebellum for the plurality of original image data, The relationship obtained by performing an operation of specifying each organ while manually changing the threshold value for the kidney is shown. The horizontal axis represents the average SUV value of the SUV values assigned to a plurality of pixels corresponding to the cerebellum, and the vertical axis represents the threshold value obtained manually when the liver or kidney is specified. Here, by using an average SUV value obtained by averaging a plurality of SUV values applied to a plurality of pixels corresponding to the cerebellum, errors due to fluctuations in the SUV values are reduced. The horizontal axis is not limited to the average SUV value, and the SUV value having the largest number among the plurality of SUV values assigned to the cerebellar pixels may be used.

  From the relationship shown in FIG. 6, the threshold value for specifying the liver and the threshold value for specifying the kidney have a predetermined correlation with the SUV value of the cerebellum. By using this correlation, a threshold value for specifying the liver or a threshold value for specifying the kidney can be obtained from the average SUV value of the cerebellum.

  Note that the cerebellum has a higher FDG SUV value than other organs and is easy to identify from its shape, so that it is possible to identify an organ without using a threshold.

  In addition, although the case of the cerebellum is shown here, the lung may be used as an organ that determines the correlation. Since the lung can be identified from the CT image, the organ can be identified without using a threshold in the PET image. Therefore, the reference organ for obtaining a threshold value in this correlation is not limited to a cerebellum or lung organ as long as it can be identified without using a threshold value in a PET image.

  FIG. 7A schematically shows the relationship between the average SUV value of the cerebellum and the threshold value for specifying the organ. The solid lines in FIGS. 7A and 7B indicate the correlation with respect to the liver threshold, and the broken lines indicate the correlation with respect to the kidney threshold.

  In FIG. 7B, when the average SUV value of the cerebellum is “a”, the threshold value for identifying the liver is obtained from the position intersecting the solid line (SUV2), and the threshold value for identifying the kidney is a broken line. A threshold (SUV3) is obtained from the intersecting position.

  When the average SUV value of the cerebellum is “b”, the threshold for specifying the liver is obtained from the position intersecting the solid line (SUV4), and the threshold for identifying the kidney is determined from the position intersecting the broken line. (SUV5) is required.

  By using this correlation, even when the SUV value fluctuates due to differences in measurement conditions such as the subject, case, and imaging conditions, the threshold value of each organ can be obtained based on the SUV value of the cerebellum.

  As shown in FIG. 8, the configuration example of the organ region specifying apparatus 1 according to the first embodiment of the present invention includes a calculation unit 2 that sets a threshold and specifies a region of a target organ, a storage unit 3, and a specific organ setting unit. 4.

  The storage unit 3 stores original image data 3a such as a PET image and a CT image, and correlation data 3c representing a correlation between the average SUV value and a threshold, and also stores a specific region 3b of the target organ obtained by the calculation unit 2. Remember.

  The original image data 3a is composed of a plurality of pixels and pixel values (SUV values) given to the pixels, the specific area 3b is composed of pixels or pixel positions representing the position of the target organ, and the correlation data 3c is A relational expression indicating a relation between the average SUV value and the threshold value or a data table can be used.

  The specific organ setting unit 4 is a means for specifying a reference organ based on a correlation such as a cerebellum or a lung from the original image data 3a. When the cerebellum is used as a specific organ, the specific organ setting unit 4 extracts pixels having a high SUV value corresponding to the cerebellum from the original image data 3a, and compares the shape of the extracted region with the shape of the cerebellum. Identify the cerebellar organs.

  When the lung is used as the specific organ, the specific organ setting unit 4 specifies the lung organ by extracting pixels corresponding to the position and shape of the lung from the original image data 3a.

  The calculation unit 2 includes a threshold setting unit 2a that sets a threshold, a comparison unit 2b that compares the original image data 3a and the threshold set by the threshold setting unit 2a, and a region of the target organ based on the comparison result of the comparison unit 2b. The area specifying unit 2c for specifying the average SUV value and the average SUV value calculating unit 2d for calculating the average SUV value.

  The average SUV value calculation unit 2d, for pixels corresponding to specific organs such as the cerebellum and lungs set by the specific organ setting unit 4, stores the SUV values that are the pixel values assigned to the pixels in the original image in the storage unit 3. The average value is obtained by reading from the data 3a. This is to obtain an average SUV value necessary for reading out the threshold value of each target organ based on the correlation data stored in the storage unit 3.

  In the correlation data, when the SUV value of a specific organ is not the average SUV value but the maximum SUV value, the maximum SUV value calculation unit is used instead of the average SUV value calculation unit 2d.

  Hereinafter, a calculation operation example of the calculation unit 2 will be described with reference to a flowchart of FIG.

  The average SUV value calculation unit 2 d reads a PET image from the original image data 3 a of the storage unit 3, and for the pixel corresponding to the specific organ set by the specific organ setting unit 4, the SUV value (pixel value) assigned to the pixel ) Is read (S31), and the average of the read SUV values is obtained to calculate the average SUV value (S32).

  The threshold value setting unit 2a reads out and reads out the threshold value corresponding to the average SUV value using the average SUV value calculated by the average SUV value calculation unit 2d and the average SUV value stored in the storage 3b and the correlation data 3c of the threshold value. A threshold value is set as a threshold value for each target organ (S33).

  The comparison unit 2b compares the original image data 3a read from the storage unit 3 with the threshold set by the threshold setting unit 2a, and extracts pixels having an SUV value equal to or greater than the threshold (S34).

  The region specifying unit 2c specifies the region of the target organ based on the pixels that are compared and extracted by the comparison unit 2b. Thus, the shape of the target organ can be obtained (S35).

  The specified organ region is stored in the specific region 3b of the organ in the storage unit 3. The organ area stored in the specific area 3b is used for subsequent diagnosis. In addition to storing the specified organ region in the specific region 3b, it may be configured to be attached to the original image data 3a and recorded, or transmitted to another device.

  FIGS. 10 and 11 show an example of region specification when setting using the correlation between the average SUV value of the cerebellum and the threshold, and FIG. 10 shows the case where the average SUV value in FIG. 7B is “a”. FIG. 11 shows a case where the average SUV value in FIG. 7B is “b”.

  10 and 11 show a case where the pixel arrangement is 6 pixels in the vertical direction and 7 pixels in the horizontal direction, and the number in each pixel represents the SUV value. Further, this SUV value is provisionally determined for convenience of explanation, and does not indicate an actual numerical value.

  The example shown in FIG. 10 shows a case where the SUV values are 1 to 3. In the correlation between the average SUV value of the cerebellum and the threshold value shown in FIG. 7B, the SUV value “3” is shown as the threshold value in the kidney and the SUV value as the threshold value in the liver with respect to the average SUV value “a” of the cerebellum. “2” is shown.

  Here, when the SUV value “3” is set as the threshold value for specifying the kidney, by extracting pixels having an SUV value of “3” or more in FIG. 10A, the region of FIG. Identified. This specific area corresponds to the area of the kidney.

  Further, when SUV values “2” and “2” are set as threshold values for specifying the liver, by extracting pixels having an SUV value of “2” or more and less than “3” in FIG. The area of FIG. 10D is specified. This specific area corresponds to the liver area.

  The example shown in FIG. 11 shows a case where the SUV value is 2-4. In the correlation between the average SUV value of the cerebellum and the threshold value shown in FIG. 7B, the SUV value “5” is shown as the threshold value in the kidney and the SUV value as the threshold value in the liver with respect to the average SUV value “b” of the cerebellum. “4” is shown.

  Here, when the SUV value “5” is set as the threshold value for specifying the kidney, by extracting pixels having an SUV value of “5” or more in FIG. 11A, the region of FIG. Identified. This specific area corresponds to the area of the kidney.

  Further, when the SUV values “4” and “5” are set as the threshold values for specifying the liver, by extracting the pixels whose SUV values are “4” or more and less than “5” in FIG. 11D is specified. This specific area corresponds to the liver area.

  Next, a 2nd form is demonstrated. FIG. 12 is a diagram showing a histogram showing the relationship between the SUV value (pixel value) given to a pixel and the number of pixels to which the SUV value (pixel value) is given, and FIG. 13 shows the second embodiment of the present invention. FIG. 14 is a flowchart for explaining an organ region specifying operation according to the second embodiment of the present invention, and FIGS. FIG. 17 is a diagram for explaining the operation of organ region identification according to the second embodiment of the invention, and FIG. 17 is a flowchart for explaining the operation of organ region identification according to another example of the second embodiment of the present invention; 18 and 19 are diagrams for explaining an organ region specifying operation according to another example of the second mode of the present invention.

  In the histogram shown in FIG. 12, the solid line in FIG. 12C shows the relationship between the SUV value (pixel value) given to the pixel corresponding to the liver and the number of pixels to which the SUV value (pixel value) is given. The broken line in FIG. 12D shows the relationship between the SUV value (pixel value) given to the pixel corresponding to the kidney and the number of pixels to which the SUV value (pixel value) is given. The alternate long and short dash line in b) shows the relationship between the SUV value (pixel value) assigned to the pixels corresponding to other organs and the number of pixels to which the SUV value (pixel value) is assigned, as shown in FIG. The alternate long and short dash line indicates the relationship between the SUV value (pixel value) assigned to pixels corresponding to other organs and the number of pixels to which the SUV value (pixel value) is assigned, and is thick in FIG. Dashed lines indicate pixels in the original image data. Granted SUV value (pixel value) and its SUV value (pixel value) indicates the relationship between the number of pixels that are granted to. Here, it is assumed that the original image data includes image data of the liver, kidney, and other organs.

  The histogram indicated by the thick broken line shown in FIG. 12A corresponds to a superposition of the histograms of the kidney, liver and other organs, and only this histogram can be obtained from the original image data. The histograms of the liver and other organs cannot be acquired because the region of each organ is not specified.

  In the second form, in this histogram, paying attention to the fact that the pixel position obtained in the histogram of the whole organ and the pixel position obtained in the histogram of each organ appear at almost the same position, and specify each organ from this peak position. The threshold value to be obtained is obtained.

  By using the peak position of this histogram, even if the SUV value fluctuates due to differences in measurement conditions such as the subject, case, and imaging conditions, the threshold value of each organ can be obtained corresponding to the original image data. it can.

  As shown in FIG. 13, the configuration example of the organ region specifying device 1 according to the second embodiment of the present invention includes a calculation unit 2 that sets a threshold and specifies a region of a target organ, and a storage unit 3.

  The storage unit 3 stores original image data 3a such as a PET image and a CT image, and a specific region 3b of the target organ obtained by the calculation unit 2.

  As in the configuration example shown in FIG. 8, the original image data 3a is composed of a plurality of pixels and pixel values (SUV values) assigned to the pixels, and the specific region 3b is a pixel or pixel representing the position of the target organ. Can be configured by position.

  As in the configuration example shown in FIG. 8, the calculation unit 2 includes a threshold setting unit 2a that sets a threshold, a comparison unit 2b that compares the original image data 3a and the threshold set by the threshold setting unit 2a, and a comparison unit 2b. In addition to the region specifying unit 2c for specifying the region of the target organ based on the comparison result, a histogram forming unit 2e for forming a histogram and a peak detecting unit 2f for detecting a peak position from the histogram are provided.

  The histogram forming unit 2e forms a histogram of the number of pixels with respect to the SUV value (pixel value) using the SUV value (pixel value) assigned to each pixel of the original image data. The peak detection unit 2f obtains an SUV value in which the number of pixels shows a peak in the formed histogram.

  Hereinafter, a calculation operation example of the calculation unit 2 will be described with reference to a flowchart of FIG. The histogram forming unit 2e reads a PET image from the original image data 3a in the storage unit 3, and the SUV value (pixel value) is assigned to each SUV value (pixel value) assigned to each pixel of the original image data. A histogram is formed by counting the number of pixels (S41).

  The peak detector 2f obtains an SUV value in which the number of pixels shows a peak in the formed histogram (S42). The threshold value setting unit 2a sets the SUV value corresponding to the peak position detected by the peak detection unit 2f as the organ threshold value corresponding to the peak position (S43).

  The comparison unit 2b compares the original image data 3a read from the storage unit 3 with the threshold set by the threshold setting unit 2a, and extracts a pixel having an SUV value equal to or greater than the threshold (S44).

  The region specifying unit 2c specifies the region of the target organ based on the pixels extracted by the comparison by the comparison unit 2b. Thereby, the shape of the target organ can be obtained (S45).

  The specified organ region is stored in the specific region 3b of the organ in the storage unit 3. The organ area stored in the specific area 3b is used for subsequent diagnosis. In addition to storing the specified organ region in the specific region 3b, it may be configured to be attached to the original image data 3a and recorded, or transmitted to another device.

  FIG. 15 and FIG. 16 show an example of specifying a region by setting a threshold value from the peak position of the histogram using the histogram. FIG. 15 shows the threshold value from the peak having the highest SUV value among the three peaks of the histogram. FIG. 16 shows a case where the threshold is set from the second peak among the three peaks.

  In FIGS. 15 and 16, the numbers in the pixels represent SUV values of 1 to 6, but these SUV values are provisionally determined for convenience of explanation and do not indicate actual numerical values.

  In the examples shown in FIGS. 15 and 16, the histogram shows peaks in the vicinity of SUV values of 2, 4, and 6. If the data included in the original image data is a range that includes the kidney, liver, and other organs, this histogram is equivalent to the overlay of the histograms of the kidney, liver, and other organs, and peaks. The position corresponds to the peak position of each organ.

  Here, in the peak position of each organ, the peak position with the lowest SUV value corresponds to the peak position based on the histogram of the other organs, the peak position with the intermediate SUV value corresponds to the peak position based on the histogram of the liver, and the SUV value The highest peak position corresponds to the peak position by the kidney histogram.

  For the original image data in FIG. 15B, the SUV value “6” corresponding to the peak position with the highest SUV value in the histogram of FIG. By doing so, the region of FIG. 15C is specified. This specific area corresponds to the area of the kidney.

  FIG. 16B shows image data obtained by removing the kidney region shown in FIG. 15C from the original image data in FIG. For this image data, the SUV value “4” corresponding to the second peak position of the SUV value in the histogram of FIG. 16A is used as a threshold value, and pixels having the threshold value “4” or more are extracted. The area of 16 (c) is specified. This specific area corresponds to the liver area.

  Next, another example in which a region is specified by setting a threshold value from the peak position of the histogram will be described with reference to FIGS. Note that this example can have the same configuration as the configuration example shown in FIG.

  This will be described with reference to the flowchart of FIG. The histogram forming unit 2e reads a PET image from the original image data 3a in the storage unit 3, and the SUV value (pixel value) is assigned to each SUV value (pixel value) assigned to each pixel of the original image data. A histogram is formed by counting the number of pixels (S51).

  The peak detector 2f obtains an SUV value in which the number of pixels shows a peak in the formed histogram (S52). The threshold setting unit 2a sets a range of a predetermined width before and after the SUV value corresponding to the peak position detected by the peak detection unit 2f, and sets the SUV values above and below this range above and below the organ corresponding to the peak position. The threshold is set (S53).

  The comparison unit 2b compares the original image data 3a read from the storage unit 3 with the upper and lower threshold values set by the threshold setting unit 2a, and extracts pixels having SUV values sandwiched between the upper and lower threshold values (S54).

  The region specifying unit 2c specifies the region of the target organ based on the pixels extracted by the comparison by the comparison unit 2b. Thereby, the shape of the target organ can be obtained (S55).

  The specified organ region is stored in the specific region 3b of the organ in the storage unit 3. The organ area stored in the specific area 3b is used for subsequent diagnosis. In addition to storing the specified organ region in the specific region 3b, it may be configured to be attached to the original image data 3a and recorded, or transmitted to another device.

  18 and 19 are examples in which a histogram is used to specify a region by setting upper and lower threshold values from the peak position of the histogram. FIG. 18 shows a peak having the highest SUV value among the three peaks included in the histogram. FIG. 19 shows a case where the upper and lower threshold values are set from the second peak among the three peaks.

  In FIGS. 18 and 19, the numbers in the pixels represent SUV values of 1 to 7, but these SUV values are provisionally determined for convenience of explanation and do not indicate actual numerical values.

  It is estimated that the range having a vertical width with respect to the peak position substantially corresponds to the expansion of the histogram of the organ corresponding to the peak position. In this example, an organ region is specified by setting a threshold value in the vertical direction corresponding to the spread of the histogram.

  In the examples shown in FIGS. 18 and 19, the histogram shows peaks in the vicinity of SUV values of 2, 4, and 6. If the data included in the original image data is a range that includes the kidney, liver, and other organs, this histogram is equivalent to the overlay of the histograms of the kidney, liver, and other organs, and peaks. The position corresponds to the peak position of each organ.

  Here, in the peak position of each organ, the peak position with the lowest SUV value corresponds to the peak position based on the histogram of the other organs, the peak position with the intermediate SUV value corresponds to the peak position based on the histogram of the liver, and the SUV value The highest peak position corresponds to the peak position by the kidney histogram.

  For the original image data in FIG. 18B, threshold values are set up and down across the SUV value “6” corresponding to the peak position having the highest SUV value in the histogram of FIG. Here, the SUV value “5” is set as the lower threshold, and the SUV value “7” is set as the upper threshold. By extracting pixels having SUV values of the upper and lower threshold values “5” to “7”, the region of FIG. 18C is specified. This specific area corresponds to the area of the kidney.

  For the original image data in FIG. 19B, threshold values are set up and down across the SUV value “4” corresponding to the second peak position of the SUV value in the histogram of FIG. 19A. Here, the SUV value “3” is set as the lower threshold, and the SUV value “5” is set as the upper threshold. By extracting the pixels having the SUV values from the upper and lower threshold values “3” to “5”, the region of FIG. 19C is specified. This specific area corresponds to the liver area.

  Next, a third embodiment of the present invention will be described.

  In the third embodiment, the distribution of pixel values of the target organ is obtained by applying the profile pattern of each organ to the histogram formed from the original image data, the variance is obtained from the distribution of pixel values of the profile pattern, and the peak value is calculated. The upper and lower pixel values determined by the sandwiching are used as upper and lower threshold values, and a threshold value for specifying the target organ is determined by extracting pixels existing between the upper and lower threshold values, and is set as a threshold value for specifying the target organ.

  FIG. 20 is a diagram for explaining a configuration example of an organ region specifying apparatus according to the third embodiment of the present invention, and FIG. 21 is a flowchart for explaining an operation of organ region specifying according to the third embodiment of the present invention. FIG. 22 and FIG. 23 are diagrams for explaining the organ region specifying operation according to the third embodiment of the present invention.

  By using the profile pattern of this histogram, the threshold value of each organ can be obtained corresponding to the original image data even when the SUV value fluctuates due to differences in measurement conditions such as the subject, case, and imaging conditions. it can.

  As shown in FIG. 20, the configuration example of the organ region specifying device 1 according to the third embodiment of the present invention includes a calculation unit 2 that sets a threshold and specifies a region of a target organ, and a storage unit 3.

  The storage unit 3 stores original image data 3a such as a PET image or a CT image, a specific region 3b of the target organ obtained by the calculation unit 2, and a histogram pattern 3d of each organ. This histogram pattern is a profile pattern represented by a histogram when a histogram is formed for each organ. This profile pattern is presumed that the shape of the histogram is the same even when the SUV value fluctuates and the peak value or spread width changes due to differences in measurement conditions such as the subject, case, and imaging conditions. . The size of the histogram pattern 3d is standardized and stored.

  In the third embodiment, the size of the profile pattern is determined by matching peak values, the range of SUV values (pixel values) is determined by using the same variance value, and the upper and lower thresholds are determined using this range.

  As in the configuration example shown in FIG. 8, the original image data 3a is composed of a plurality of pixels and pixel values (SUV values) assigned to the pixels, and the specific region 3b is a pixel or pixel representing the position of the target organ. Can be configured by position.

  As in the configuration example shown in FIG. 8, the calculation unit 2 includes a threshold setting unit 2a that sets a threshold, a comparison unit 2b that compares the original image data 3a and the threshold set by the threshold setting unit 2a, and a comparison unit 2b. In addition to the region specifying unit 2c that specifies the region of the target organ based on the comparison result, a histogram forming unit 2g that forms a histogram and a variance calculation unit 2h that distributes each target organ are provided.

  The histogram forming unit 2e forms a histogram of the number of pixels with respect to the SUV value (pixel value) using the SUV value (pixel value) assigned to each pixel of the original image data. In the formed histogram, the variance calculation unit 2h obtains a variance value representing the distribution state of each target organ by matching the profile pattern with the histogram at the peak position. This variance value is used as a value that determines the upper and lower threshold values of the pixel value (SUV value) together with the peak position of the histogram.

  Next, with reference to the flowchart of FIG. 21 and the operation explanatory diagrams of FIGS. 22 and 23, an operation example of specifying a region by setting a threshold value from a profile pattern of a histogram will be described.

  The histogram forming unit 2g reads a PET image from the original image data 3a of the storage unit 3, and the SUV value (pixel value) is assigned to each SUV value (pixel value) assigned to each pixel of the original image data. A histogram is formed by counting the number of pixels (S61).

  In the formed histogram, the variance calculation unit 2h obtains an SUV value indicating the peak number of pixels (S62), reads the histogram pattern 3d of the target organ from the storage unit 3, and matches the read histogram pattern 3d to the peak value. Zoom in and out.

  As described above, the histogram is obtained by superimposing the histograms of the target organs assumed in advance, and the peak position of the histogram substantially corresponds to the peak position of the histogram of each target organ.

  The present invention matches the peak position of the histogram pattern of the organ corresponding to the peak position with respect to each peak position of the histogram, and enlarges / reduces the histogram pattern 3d so that the peak values match, thereby the target organ Is estimated (S64).

  In the third embodiment, pixels within a predetermined range of the target organ histogram are extracted as pixels of the target organ. This predetermined spread is determined using the variance σ of the histogram.

  The variance σ is calculated from the histogram of each target organ formed in S64. The threshold setting unit 2a sets the range of nσ as the range of pixel values (SUV values) from which pixels are extracted, sets the value obtained by subtracting nσ from the pixel value (SUV value) of the peak value as the lower threshold, and sets the pixel value of the peak value A value obtained by adding nσ to (SUV value) is set as the upper threshold value. Note that “n” can be arbitrarily set in consideration of the overlapping of histograms of each target organ (S65).

  The comparison unit 2b compares the original image data 3a read from the storage unit 3 with the upper and lower threshold values set by the threshold setting unit 2a, and extracts pixels having SUV values sandwiched between the upper and lower threshold values (S66).

  The region specifying unit 2c specifies the region of the target organ based on the pixels extracted by the comparison by the comparison unit 2b. Thereby, the shape of the target organ can be obtained (S67).

  The specified organ region is stored in the specific region 3b of the organ in the storage unit 3. The organ area stored in the specific area 3b is used for subsequent diagnosis. In addition to storing the specified organ region in the specific region 3b, it may be configured to be attached to the original image data 3a and recorded, or transmitted to another device.

  FIGS. 22 and 23 are examples in which a histogram is used to specify a region by setting a threshold value from a profile pattern of the histogram. FIG. 22 shows the top and bottom of the peak having the highest SUV value among the three peaks of the histogram. FIG. 23 shows a case where the upper and lower threshold values are set from the second peak among the three peaks.

  22 and 23, the numbers in the pixels represent SUV values of 1 to 7, but these SUV values are provisionally determined for convenience of explanation, and do not represent actual numerical values.

  It is estimated that the range having a vertical width with respect to the peak position substantially corresponds to the expansion of the histogram of the organ corresponding to the peak position. In this example, an organ region is specified by setting a threshold value in the vertical direction corresponding to the spread of the histogram.

  In the examples shown in FIGS. 22 and 23, the histogram shows peaks in the vicinity of SUV values of 2, 4, and 6. If the data included in the original image data is a range that includes the kidney, liver, and other organs, this histogram is equivalent to the overlay of the histograms of the kidney, liver, and other organs, and peaks. The position corresponds to the peak position of each organ.

  Here, in the peak position of each organ, the peak position with the lowest SUV value corresponds to the peak position based on the histogram of the other organs, the peak position with the intermediate SUV value corresponds to the peak position based on the histogram of the liver, and the SUV value The highest peak position corresponds to the peak position by the kidney histogram.

  For the original image data in FIG. 22B, the kidney histogram pattern is applied to the SUV value “6” corresponding to the peak position having the highest SUV value in the histogram in FIG. Adjust the size of the histogram pattern. Using the variance σ of the formed histogram pattern, a threshold value is set in the range of nσ width above and below the peak position. Here, the SUV value “5” is set as the lower threshold value, and the SUV value “6” is set as the upper threshold value. By extracting pixels having SUV values from “5” to “6” above and below the upper and lower thresholds, the region of FIG. 22C is specified. This specific area corresponds to the area of the kidney.

  For the original image data in FIG. 23B, threshold values are set up and down across the SUV value “4” corresponding to the second peak position of the SUV value in the histogram of FIG.

  In this case, the liver histogram pattern is applied to the SUV value “4” corresponding to the second peak position of the SUV value in the histogram of FIG. 23A, and the size of the histogram pattern is set so that the peak values match. The thresholds are set in the range of the width of nσ above and below the peak position using the variance σ of the formed histogram pattern.

  Here, the SUV value “3” is set as the lower threshold, and the SUV value “5” is set as the upper threshold. By extracting the pixels having the SUV values from the upper and lower threshold values “3” to “5”, the region of FIG. 23C is specified. This specific area corresponds to the liver area.

  In the above-described embodiment, an example in which organ regions of the liver and the kidney are specified is shown, but the present invention can be applied not only to the liver and the kidney but also to other organs.

  In the first embodiment, the cerebellum and lung organs are shown as the reference organs, but the threshold values can be set based on other organs that can be identified from features such as shapes regardless of the threshold values. Further, another organ threshold value can be set on the basis of the organ specified in the second and third embodiments.

  The technique related to organ region identification according to the present invention is not limited to the medical field, but can be applied to a field for performing a diagnostic operation including a series of operations for performing an estimation operation based on a predetermined determination method and determination criteria.

It is a figure which shows the grade of the accumulation | aggregation of FDG of a PET image, and the grade of the transmission of the radiation of CT phase. It is a flowchart for demonstrating the diagnostic procedure using a PET-CT image. It is a flowchart for demonstrating the process of specifying an organ area | region from original image data. It is a flowchart for demonstrating the procedure which specifies a candidate area | region from an extraction area | region. It is explanatory drawing for demonstrating the procedure which specifies a candidate area | region from an extraction area | region. It is a figure which shows the relationship between a pixel value (SUV value) and the threshold value for specifying an organ. It is a figure for demonstrating the procedure which calculates | requires the threshold value which specifies an organ from a pixel value (SUV value). It is a figure for demonstrating the structural example of the organ area | region identification apparatus by the 1st form of this invention. It is a flowchart for demonstrating the operation | movement of the organ area | region specification by the 1st form of this invention. It is a figure for demonstrating the operation | movement of organ region specification by the 1st form of this invention. It is a figure for demonstrating the operation | movement of organ region specification by the 1st form of this invention. It is a figure which shows the histogram showing the relationship between SUV value (pixel value) and the number of pixels. It is a figure for demonstrating the structural example of the organ area | region identification apparatus by the 2nd form of this invention. It is a flowchart for demonstrating operation | movement of the organ area | region specification by the 2nd form of invention. It is a figure for demonstrating the operation | movement of the organ area | region specification by the 2nd form of this invention. It is a figure for demonstrating the operation | movement of the organ area | region specification by the 2nd form of this invention. It is a flowchart for demonstrating the operation | movement of the organ area | region identification by another example of the 2nd form of this invention. It is a figure for demonstrating the operation | movement of the organ area | region identification by another example of the 2nd form of this invention. It is a figure for demonstrating the operation | movement of the organ area | region identification by another example of the 2nd form of this invention. It is a figure for demonstrating the structural example of the organ area | region identification apparatus by the 3rd form of this invention. It is a flowchart for demonstrating the operation | movement of the organ area | region specification by the 3rd form of this invention. It is a figure for demonstrating the operation | movement of the organ area | region specification by the 3rd form of this invention. It is a figure for demonstrating the operation | movement of the organ area | region specification by the 3rd form of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Organ area | region identification apparatus 2 Calculation part 2a Threshold setting part 2b Comparison part 2c Area | region identification part 2d SUV average value calculation part 2e Histogram formation part 2f Peak detection part 2g Histogram formation part 2h Variance calculation part 3 Memory | storage part 3a Original image data 3b specification Area 3c Correlation data 3d Histogram pattern 4 Specific part setting part

Claims (12)

A method for identifying each region where a plurality of target organs exist from the original image data of a medical image obtained by a diagnostic device,
The original image data includes a plurality of pixels and a pixel value assigned to each pixel;
Setting a threshold value corresponding to each target organ;
Comparing the pixel value with the threshold to extract a pixel;
Identifying a region of the target organ based on the extracted pixels,
The step of setting the threshold value sets the threshold value for each target organ from the pixel value assigned to each pixel of the original image data for each original image data. A correlation between a pixel value given to a predetermined organ other than the target organ and a threshold value for specifying the target organ is obtained in advance,
The step of setting the threshold includes
Obtain a pixel value given to a pixel of a predetermined organ from the original image data,
Obtaining a threshold value for a pixel value of a predetermined organ based on the correlation;
2. The organ region specifying method according to claim 1, wherein the obtained threshold is set as a threshold for specifying the target organ.   The predetermined organ is a cerebellum or a lung, and the pixel value given to the predetermined organ in the correlation is an average value of pixel values of pixels corresponding to the cerebellum or the lung. Method for organ region identification. The step of setting the threshold includes
Based on the original image data, a histogram representing a relationship between a pixel value assigned to a pixel and the number of pixels to which the pixel value is assigned is formed.
Obtaining a pixel value corresponding to the peak of the number of pixels in the histogram;
2. The organ region specifying method according to claim 1, wherein the obtained pixel value is set as a threshold value for specifying the target organ. For a target organ, a profile pattern of a histogram representing a relationship between a pixel value assigned to a pixel and the number of pixels to which the pixel value is assigned is obtained in advance.
The step of setting the threshold includes
Based on the original image data, a histogram representing a relationship between a pixel value assigned to a pixel and the number of pixels to which the pixel value is assigned is formed.
By obtaining the distribution of pixel values of the target organ by applying the profile pattern to the histogram formed from the original image data, obtaining a threshold from the distribution of the distribution of the pixel values,
2. The organ region specifying method according to claim 1, wherein a threshold value for specifying a target organ is obtained as the obtained threshold value.   6. The organ region specifying method according to claim 1, wherein the original image data is PET image data, and a value given to the pixel is an SUV value. A device that identifies each region where a plurality of target organs exist from the original image data of a medical image obtained by a diagnostic device,
The original image data includes a plurality of pixels and a pixel value assigned to each pixel;
An original image data of a medical image obtained by a diagnostic device, and a storage unit for storing a region of a target organ obtained by specifying an organ from the original image data;
A calculation unit that obtains a region of the target organ and stores it in the storage unit,
The computing unit is
A threshold setting unit for setting a threshold corresponding to each target organ;
A comparison / extraction unit that compares the pixel value with the threshold value to extract a pixel;
An area specifying unit for specifying the area of the target organ based on the extracted pixels;
The threshold value setting unit sets a threshold value for each target organ from a pixel value assigned to each pixel of the original image data for each original image data. The storage unit stores a correlation between a pixel value given to a predetermined organ other than the target organ and a threshold value for specifying the target organ,
The threshold setting unit includes:
Read the pixel value assigned to the pixel of the predetermined organ from the original image data,
Read the correlation from the storage unit,
8. The organ region specifying apparatus according to claim 7, wherein a threshold value for a pixel value of a predetermined organ is obtained based on the read correlation, and the obtained threshold value is set as a threshold value for specifying the target organ. The predetermined organ is a cerebellum or a lung;
9. The organ region identification according to claim 8, wherein in the correlation stored in the storage unit, a pixel value given to a predetermined organ is an average value of pixel values corresponding to a cerebellum or a lung. apparatus. A histogram representing the relationship between the pixel value assigned to the pixel and the number of pixels to which the pixel value is assigned by reading the pixel value assigned to the pixel corresponding to the predetermined organ from the original image data stored in the storage unit A histogram forming unit for forming
A peak detection unit for obtaining a pixel value corresponding to a peak of the number of pixels in the histogram,
The threshold setting unit includes:
8. The organ region specifying apparatus according to claim 7, wherein the pixel value obtained by the peak detection unit is set as a threshold value for specifying a target organ. The storage unit stores a profile pattern of a histogram representing a relationship between a pixel value assigned to a pixel and the number of pixels to which the pixel value is assigned for a target organ,
A histogram representing the relationship between the pixel value assigned to the pixel and the number of pixels to which the pixel value is assigned by reading the pixel value assigned to the pixel corresponding to the predetermined organ from the original image data stored in the storage unit A histogram forming unit for forming
A distribution calculation unit for obtaining a distribution of pixel values of a target organ by applying a profile pattern stored in a storage unit to a histogram formed by the histogram formation unit,
The threshold setting unit includes:
8. The organ region specifying apparatus according to claim 7, wherein a threshold value is obtained from the distribution of pixel value distributions calculated by the distribution calculating unit, and a threshold value for specifying a target organ is obtained from the obtained threshold value.   The organ area specifying device according to any one of claims 7 to 11, wherein the original image data is PET image data, and a value given to the pixel is an SUV value.