JP2004113730A - Computer-aided diagnosis system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To organize an environment for easily preserving projection data and re-constituting an image, in order to use image data which is re-constituted by using an image re-constitution parameter most suitable for a CAD data processing, in a computer-aided diagnosis (CAD) system using medical image data. <P>SOLUTION: This CAD system includes: a large scale data storage device; and a high speed data processor, which are arranged on a network. The system also includes; a mechanism for storing projection data collected by a medical image diagnostic instrument in the large scale data storage device; a mechanism with which a user on the network reads projection data from the large scale data storage device, and transfers the data to the high speed data processor, a mechanism for re-constituting the image by using the image re-constituting parameter desired by the user with respect to the projection data; a mechanism for performing the CAD data processing in re-constituted image data through the use of the CAD data processing parameter desired by the user; and a mechanism for transferring the result to the user. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
[Technology to which the Invention belongs]
The present invention relates to a computer interpretation support system that supports interpretation of medical images such as X-ray CT images.
[0002]
[Prior art]
Chest X-ray images taken with an X-ray imaging device have been used for the diagnosis of breast malignant tumors. However, since the X-ray image is a projection image in which all information in the thickness direction is convoluted, the chest X-ray image It was difficult to find malignant tumors at an early stage by interpretation.
[0003]
An X-ray CT apparatus can obtain accurate cross-sectional image data representing physical properties of a human body. In the initial X-ray CT, the number of cross-sectional image data collected in one examination for one subject was 10 to 20, and the slice thickness was about 5 mm to 10 mm. The data collection time also required about 10 seconds per slice. The use of chest X-ray CT images collected by an X-ray CT apparatus for the diagnosis of malignant breast tumors has been studied, but the number of chest X-ray CT images taken is limited due to the restriction of the amount of X-ray exposure to the patient. Because of the fact that clear images cannot be obtained due to the effects of body movement, chest X-ray CT images have not been widely used for diagnosis of breast malignant tumors.
[0004]
In recent years, with the technical advancement of X-ray CT systems, helical scan X-ray CT systems and multi-row detector X-ray CT systems have been put to practical use. Image data can be acquired in a short time. In addition, due to the technical progress of the X-ray CT apparatus, the amount of X-ray exposure to a patient can be significantly reduced. Accordingly, the use of chest X-ray CT images for diagnosis of breast malignancy has been studied again.
[0005]
In order to detect a breast malignant tumor at an early stage using an X-ray CT image, it is necessary to detect an abnormal shadow as small as possible. For this purpose, it is desirable to use an X-ray CT image having a slice thickness as small as possible. On the other hand, it is necessary to completely cover the entire region of the chest with the X-ray CT image of the slice thickness, so the thinner the slice thickness, the more the number of X-ray CT images.
[0006]
Conventionally, as a method of reading an X-ray CT image, a method of printing an X-ray CT image on a film and attaching the film to a light box for reading the image has been performed. However, with the increase in the number of X-ray CT images, it has become difficult to print the X-ray CT images on a film and attach the film to a light box for interpretation. For this reason, a so-called electronic light box has been attempted in which a plurality of high-definition image display devices capable of displaying a large number of images are arranged, and the images collected by one inspection are sequentially displayed on the high-definition image display devices.
[0007]
However, even when a high-definition image display device is used, the detection of abnormal shadows depends on the ability of the doctor to interpret the images, as in the case of image interpretation using a film. Therefore, there is a risk that fine shadows may be overlooked. In order to reduce this oversight, an attempt has been made to detect abnormal shadow candidates by image processing, attach a mark to that portion, and alert a doctor when reading images. Such a system for assisting doctors in interpreting images is called a computer diagnosis support system (CAD).
[0008]
In CAD that assists a doctor in interpreting an X-ray CT image, it is common practice to apply a two-dimensional spatial filter to a cross-sectional image to detect a shadow candidate. Research has been conducted to obtain a two-dimensional spatial filter that is most suitable for various types of shadows by changing the shape of the two-dimensional spatial filter in various ways.
[0009]
As in the case where a doctor interprets an X-ray CT image with the naked eye, it is desirable to use an X-ray CT image with a slice thickness as thin as possible in order to detect a candidate for an abnormal shadow as small as possible even in CAD. . On the other hand, as the slice thickness is reduced, the number of X-ray CT images increases. In addition, the shadow candidate picked up by the CAD needs to be judged by the doctor reading all the candidate areas with the naked eye. For this reason, the number of X-ray CT images processed by CAD increases and the number of X-ray CT images interpreted by a doctor increases as the X-ray CT images having a smaller slice thickness are used.
[0010]
In recent years, helical scan X-ray CT devices and multi-row detector X-ray CT devices have been put into practical use due to technological advances in X-ray CT devices. It has become possible to acquire more precise image data in a short time, in which the distance between cross sections of the image data is narrow. In addition, due to the technical progress of the X-ray CT apparatus, the amount of X-ray exposure to a patient can be significantly reduced. On the other hand, the number of X-ray CT image data used for CAD is limited due to the limitations of CAD processing ability and doctor's interpretation ability.
[0011]
In an X-ray CT apparatus, projection data of a cross section of a subject is collected by the main body of the X-ray CT apparatus, and the projection data of the cross section is subjected to image reconstruction processing by a high-speed data processing device to obtain a plurality of cross sections. Image data is created, and the image data of the plurality of cross sections is stored in a storage medium such as a magnetic disk device, a magneto-optical disk device, or an optical disk device. The image data of a plurality of cross sections stored in the storage medium is read out, and the image data is displayed on an image observation device for observation. Further, image data of a plurality of transverse sections stored in the storage medium may be read, and an image processing device may perform image processing on the image data. For example, a three-dimensional image processing apparatus constructs a three-dimensional image from image data of a plurality of cross sections and observes the three-dimensional image.
[0012]
In the early X-ray CT apparatus, the movement of the tabletop on which the subject lies down and the collection of projection data of the subject by X-ray irradiation were performed alternately. X-ray irradiation is performed simultaneously with the movement of the tabletop to collect projection data. Such a method of collecting projection data is called a helical scan.
[0013]
In the early X-ray CT apparatus, the movement of the tabletop on which the subject lies down and the collection of projection data of the subject by X-ray irradiation were performed alternately. The cross section of the image data obtained by performing the above. That is, the cross section where the projection data was collected matches the cross section of the image data obtained by performing the image reconstruction, and the number of the cross sections from which the projection data was collected and the image data obtained by performing the image reconstruction. Are the same. The position of the cross section where the projection data was collected was discretely distributed in the body axis direction of the patient. On the other hand, in the helical scan method, projection data is collected by performing X-ray irradiation simultaneously with the movement of the tabletop on which the subject lies, and therefore there is no discrete projection data collection cross section. Generally, projection data of a virtual cross section is obtained by setting a virtual cross section and performing interpolation processing of projection data, and image reconstruction processing is performed using the projection data of the virtual cross section. I have. Since the virtual cross section of the projection data can be set quite freely, the number of image data and the interval between the cross sections, ie, the slice interval, obtained by the image reconstruction process depend on the setting of the virtual cross section of the projection data. You can choose quite freely.
[0014]
In early X-ray CT systems, X-ray detectors for detecting the intensity of X-rays transmitted through a subject are arranged in a line in the body axis direction of the subject, and collect projection data of one cross section of the subject. Was. In recent X-ray CT apparatuses, a plurality of rows of X-ray detectors are arranged in the body axis direction of the subject, and a multi-row detector that simultaneously collects projection data of a plurality of cross sections of the subject is used. By using the multi-row detector and the helical scan method at the same time, much more projection data can be acquired than in the early X-ray CT apparatus. Furthermore, in the multi-row detector / helical scan method that uses this multi-row detector and the helical scan method together, there is no discrete projection data acquisition cross section as in the case of the early X-ray CT apparatus. The projection data of the virtual cross section is obtained by performing the interpolation processing of the projection data by setting the cross section, and the image reconstruction processing is performed using the projection data of the virtual cross section. Since the virtual cross section of the projection data can be freely set, the number of image data obtained by the image reconstruction processing and the interval between the cross sections can be freely selected depending on the setting of the virtual cross section of the projection data. Can be.
[0015]
In an examination using an X-ray CT apparatus, projection data of a subject is collected by an X-ray CT apparatus, and image reconstruction processing is performed on the projection data by a high-speed data processing apparatus, and the result of the image reconstruction processing is obtained. Is stored in a storage medium such as a magnetic disk device, a magneto-optical disk device, or an optical disk device. An image obtained by performing image processing such as window processing on image data of a plurality of cross sections obtained as a result of image reconstruction processing is printed on a film. A method for observing a film, a method for displaying an image obtained by performing image processing such as window processing on image data of a cross section obtained as a result of the image reconstruction processing on an image observation device, and an image reconstructing method. There is a method of performing advanced image processing such as three-dimensional image processing with an image processing device on image data of a plurality of cross sections obtained as a result of the configuration processing, and observing the result. Although there is an optimum image reconstruction processing parameter for each of these utilization methods, it is actually performed to perform the image reconstruction processing again using the optimal image reconstruction processing parameter for each of the utilization methods. Absent.
[0016]
In the case of CAD using X-ray CT image data, image reconstruction is performed using image reconstruction processing parameters suitable for CAD, which are different from image reconstruction processing parameters suitable for directly observing a cross-sectional image. Although it is possible to improve the ability to detect shadow candidates by using image data obtained by performing the configuration, the data stored for a long period is generally only the image data after image reconstruction, and Since data is not stored for a long period of time, such image reconstruction processing suitable for CAD cannot be generally performed.
[0017]
When a three-dimensional image is constructed using image data of a plurality of cross sections and CAD is performed using the data, image reconstruction processing parameters suitable for directly observing a cross section image are different. It is often desirable to use image data of multiple cross sections obtained by performing image reconstruction using image reconstruction processing parameters suitable for three-dimensional image processing. Since only the later image data and the projection data are not stored for a long period of time, such image reconstruction processing for each purpose of use cannot be actually performed.
[0018]
[Problems to be solved by the invention]
In an examination using an X-ray CT apparatus, at present, projection data is collected by an X-ray CT apparatus, image reconstruction processing is performed on the projection data by a high-speed data processing apparatus, and as a result of the image reconstruction processing, The obtained image data of a plurality of cross sections are stored in a storage medium such as a magnetic disk device, a magneto-optical disk device, and an optical disk device. Although there are image reconstruction processing parameters suitable for each of the methods using the results of inspection by the X-ray CT apparatus, the data stored over a long period of time is image data after image reconstruction, Since is not stored over a long period of time, it is not actually performed to perform the image reconstruction processing with the optimal image reconstruction parameters for each of the usage methods. In particular, when constructing a three-dimensional image using multiple cross-sectional image data, it is suitable for three-dimensional image processing, which is different from image reconstruction processing parameters suitable for direct observation of cross-sectional images. It is often desired to use image data of a plurality of cross sections reconstructed using the image reconstruction processing parameters, but such image reconstruction processing specialized for three-dimensional image processing has not been performed. In the case of CAD, a cross section obtained by image reconstruction using an image reconstruction processing parameter suitable for CAD data processing, which is different from an image reconstruction processing parameter suitable for directly observing the image of the cross section. In many cases, it is desired to use the image data described above, but such image reconstruction processing for CAD has not been performed. The reasons for this are that an image reconstruction device that reconstructs projection data to create image data is expensive, that the data processing time required for image reconstruction processing is long, and that the amount of projection data is large, Saving both data and image-reconstructed image data may increase the required storage medium capacity.
[0019]
In CAD that assists a doctor in interpreting an X-ray CT image, it is common practice to apply a two-dimensional spatial filter to a cross-sectional image to detect a shadow candidate. Research has been conducted to find the most suitable two-dimensional spatial filter for various shades by changing the shape of the two-dimensional spatial filter in various ways. Similar to the case where a doctor interprets an X-ray CT image with the naked eye, it is desirable to use an X-ray CT image having a slice thickness as thin as possible in order to detect an abnormal shadow as small as possible even in CAD. On the other hand, as the slice thickness is reduced, the number of X-ray CT images increases. In addition, the shadow candidate picked up by the CAD needs to be judged by the doctor reading all the candidate areas with the naked eye. For this reason, the number of X-ray CT images processed by CAD increases and the number of X-ray CT images interpreted by a doctor increases as the X-ray CT images having a smaller slice thickness are used. The number of X-ray CT images used for CAD is limited due to restrictions on CAD processing ability and doctor's interpretation ability.
[0020]
[Means for Solving the Problems]
The present invention provides a system for supporting interpretation of an X-ray CT image devised to solve the above-mentioned problem, a mechanism for storing projection data collected by a main body of the X-ray CT apparatus in a large-capacity storage device, A mechanism for reconstructing projection data stored in the apparatus using CAD or image reconstruction processing parameters based on a doctor's request, and a mechanism for detecting shadows using the reconstructed image data A mechanism for performing image processing. This makes it possible to change the slice thickness and the number of pieces of image data used by CAD between the case of screening and the case of precise image reading.
[0021]
When the present invention is used, for example, at the time of screening, shadow candidates are detected by using image data of 100 slices having a slice thickness of 2 mm obtained by reconstructing projection data, and then the shadow detected from this image data is detected. It is possible to precisely inspect a region of interest including a candidate using image data obtained by performing image reconstruction of projection data with a slice thickness of 0.2 mm.
[0022]
In the conventional CAD apparatus, since the CAD data processing, that is, the data processing for detecting a shadow candidate is performed using the image data reconstructed by the X-ray CT apparatus, different spatial resolutions are used for screening and detailed inspection. Could not be used with image data. When high spatial resolution image data suitable for precision inspection is used as image data, the number of image data increases. For example, when image data reconstructed from the projection data with a slice thickness of 0.2 mm is used, about 1000 images are obtained in the lung region. Further, if only the image data having a slice thickness of 2 mm and the number of sheets of 100 is used, it is suitable for screening, but the spatial resolution is insufficient for reexamining the shadow candidate precisely.
[0023]
In the CAD system according to the present invention, a mechanism for storing projection data collected by the main body of the X-ray CT apparatus in a large-capacity storage device, and a method for converting projection data stored in the large-capacity storage device into an image based on CAD or a doctor's request Since a mechanism for performing image reconstruction using the reconstruction processing parameters and a mechanism for performing data processing for shadow detection using the image data obtained by the reconstruction are provided, an image used by CAD is provided. The slice thickness and the number of data slices can be changed between screening and precision reading, and the most suitable image data for each purpose can be reconstructed from projection data and used.
[0024]
Based on an instruction from a user on the network, the projection data desired by the user is read from the large-scale data storage device and transferred to the high-speed data processing device, and the image reconstruction desired by the user is performed on the projection data. The image reconstruction processing is performed using the processing parameters, the CAD data processing is performed on the image data that has been subjected to the image reconstruction processing using the CAD data processing parameters desired by the user, and the result of the CAD data processing is performed. Was transferred to the user.
[0025]
Conventionally, a user who performs CAD data processing installs a computer device such as a workstation capable of high-speed data processing, stores a large amount of image data in the computer device, and uses the stored image data. Image processing is being performed. According to the present invention, a high-speed data processing device is installed near a network of a projection data storage device on a network, the data processing device performs image reconstruction processing and CAD data processing, and only the result of the CAD data processing is used by the user. Since the data is transmitted to the terminal device, the user does not need to prepare a computer device capable of processing data at high speed. Since the user's terminal device only instructs the selection of the projection data, the image reconstruction processing parameter and the CAD data processing parameter, and only displays the result of the CAD data processing sent from the high-speed data processing device, With a general personal computer such as a laptop computer or a notebook personal computer, it has become possible to obtain a result equal to or higher than that of a workstation in a conventional system.
[0026]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, a computer interpretation support system according to the present invention will be described. FIG. 1 schematically shows a computer interpretation support system according to the present invention. Reference numerals 101, 102, 103, etc. denote large-scale networks such as the Internet. Reference numeral 201 denotes a projection data storage / processing center; 211, a projection data storage device; and 212, a high-speed data processor for image reconstruction processing. Reference numeral 251 denotes a data processing device for CAD. Reference numeral 301 denotes an X-ray CT apparatus connected to a network, and 302 denotes a console thereof. Reference numeral 401 denotes a terminal device for CAD, and 402 denotes a terminal device such as a personal computer connected to a network.
[0027]
The projection data collected by the X-ray CT apparatus 301 is transferred to the projection data storage and processing center 201 via a network composed of networks 101, 102, 103,. The projection data storage / processing center 201 stores the projection data in the projection data storage device 211. A user who uses image data for CAD makes an inquiry from the CAD terminal device 401 to the projection data storage / processing center 201, specifies necessary projection data, and transmits parameters for image reconstruction processing. The projection data storage / processing center 201 reads the requested projection data from the data storage device 211 and transfers it to the image reconstruction processing device 212. The image reconstruction processing device 212 performs image reconstruction processing on the projection data based on the image reconstruction processing parameters sent from the CAD terminal device 401, and converts the image data obtained by the image reconstruction processing into CAD data via a network. The data is transferred to the processing device 251. The CAD data processing device 251 performs CAD data processing on the projection data based on the CAD data processing parameters sent from the CAD terminal device 401, and transfers the data resulting from the CAD data processing to the CAD terminal device 401 via the network. The user displays the result data on a terminal device such as the CAD terminal device 401 and the personal computer 402 and examines the result data.
[0028]
Since the CAD system according to the present invention stores projection data in the projection data storage device in this way, the user of the CAD system can specify desired projection data, specify reconstructed image processing parameters and CAD data processing. By specifying the parameters and transmitting them, the results of the image reconstruction processing and the CAD data processing can be received by the CAD terminal device 401 or the personal computer 402.
[0029]
The user of the CAD system who wants to use the image data reconstructed using the new image reconstruction processing parameters can specify the desired projection data and the new image reconstruction processing parameters using the CAD terminal device 401. And sends them to the projection data storage and processing center 201. In addition, CAD data processing parameters are specified and transmitted to the CAD data processing device. The projection data storage / processing center reads the projection data requested by the user from the data storage device and transfers it to the image reconstruction processing device. The image reconstruction processing device performs image reconstruction processing of the projection data based on the image reconstruction processing parameters sent from the user, and sends the image data obtained by the image reconstruction to the CAD data processing device via the network. Forward. The user specifies CAD data processing parameters and sends them to the CAD data processing device. The CAD data processing device performs CAD data processing based on CAD data processing parameters specified by the user, and transmits data obtained as a result of the CAD data processing to the user. The user can receive the result data of the CAD data processing executed using the new image reconstruction processing parameters and the new CAD data processing parameters only by issuing an instruction.
[0030]
For example, in order to perform CAD data processing using image data reconstructed with image reconstruction processing parameters most suitable for a screening inspection, the user designates desired projection data on the CAD terminal device 401 and specifies a desired image. Reconstruction processing parameters are specified and transmitted to the projection data storage and processing center 201. The projection data storage / processing center reads the requested projection data from the projection data storage device 211 and transfers it to the image reconstruction processing device 212. The image reconstruction processing device performs image reconstruction processing of the projection data based on the image reconstruction processing parameters sent from the user, and sends the obtained image data to the CAD data processing device 251 via the network. The user specifies CAD data processing parameters on the CAD terminal device 401 and transmits the parameters to the CAD data processing device. The CAD data processing device 251 performs CAD data processing based on the CAD data processing parameters specified by the user, and transmits the result data to the user. The user examines the result of the CAD data processing performed by the CAD data processing device 251. Since the CAD data processing device 251 lists, for example, candidates for abnormal shadows as a result of the CAD data processing, the user examines the candidates for abnormal shadows.
[0031]
In order to precisely inspect the area including the abnormal shadow candidate, a user who wishes to perform CAD data processing using image data reconstructed using image reconstruction processing parameters suitable for the detailed examination is CAD. The terminal device 401 specifies a new image reconstruction area to be used for image reconstruction of projection data and a new image reconstruction processing parameter, and sends them to the projection data accumulation / processing center 201. The image reconstruction processing device 212 of the projection data storage and processing center performs image reconstruction processing on the projection data using the new image reconstruction area and the new image reconstruction processing parameters sent from the user, and obtains the image reconstruction result. The image data is sent to the CAD data processing device 251 via the network. The user designates a new CAD data processing parameter with the CAD terminal device 401 and transmits the new CAD data processing parameter to the CAD data processing device. The CAD data processing device 251 performs CAD data processing based on the CAD data processing parameters specified by the user, and transmits the result data to the user. The user examines the result of the CAD data processing performed by the CAD data processing device 251.
[0032]
As described above, in the CAD system according to the present invention, since the projection data is stored in the projection data storage device, the screening CAD examination is performed using the image data reconstructed using the image reconstruction parameters suitable for the screening examination. Is performed, and a region including the found abnormal shadow candidate can be subjected to a precise CAD inspection using image data reconstructed using image reconstruction parameters suitable for the detailed inspection.
[0033]
FIG. 2 is a block diagram schematically showing a data flow of the CAD system according to the present invention. 201 is a projection data storage and processing center, 211 is a projection data storage device, and 212 is a model of a high-speed data processing device for image reconstruction processing. Reference numeral 251 schematically represents a CAD data processing device that performs CAD data processing using image data reconstructed by the image reconstructing device.
[0034]
Reference numeral 301 denotes an X-ray CT apparatus, and reference numeral 311 denotes a scanner unit which is a main body of the X-ray CT apparatus. The output data 321 of the X-ray detector collected by the scanner unit 311 of the X-ray CT apparatus is sent to a preprocessing apparatus 312, where processing such as noise removal and calibration of the data is performed. The projection data 322 preprocessed by the preprocessing device 312 is sent to the image reconstruction device 313. The image reconstructing device 313 performs an image reconstructing process on the projection data 322, and the reconstructed image data 324 is sent to the console 302 and displayed after image processing such as window processing.
[0035]
The projection data preprocessed by the preprocessing device 312 is sent to the image reconstruction device and also sent to the projection data storage and processing center. Reference numeral 323 denotes projection data sent to the projection data storage / processing center. The projection data 323 sent to the projection data storage / processing center 201 is stored in the projection data storage device 211. The projection data storage device 211 registers the attribute of the received projection data in the database device of the projection data storage device, and stores the projection data in the mass storage device.
[0036]
The CAD terminal device 401 transmits an instruction 421 for specifying projection data to be used for CAD inspection to the projection data storage device 211 of the projection data storage / processing center 201. The projection data storage device 211 reads out the specified projection data based on the instruction 421 from the CAD terminal device 401 and transfers it to the image reconstruction device 212.
The CAD terminal device 401 transmits an instruction 422 for specifying an image reconstruction parameter to the image reconstruction device 212 of the projection data accumulation / processing center 201. The image reconstruction device 212 performs an image reconstruction process on the projection data 221 sent from the projection data storage device using the image reconstruction parameter 422 instructed from the CAD terminal device 401, and converts the obtained image data 222 into CAD data. The data is transferred to the data processing device 251.
[0037]
The CAD terminal device 401 transmits a parameter 423 relating to CAD data processing to the CAD data processing device 251. The CAD data processing device 251 performs CAD data processing on the image data 222 by using the CAD data processing parameters 423 transmitted from the CAD terminal device 401. The CAD data processing device 251 transmits the result 261 of the CAD data processing to the CAD terminal device 401.
[0038]
In the CAD data processing result 261 processed by the CAD data processing device, for example, a region of interest including a candidate for a malignant tumor is displayed as a circular region on a CT image, for example. In some cases, a three-dimensional image is constructed from image data obtained by reconstructing projection data, and a region of interest including a malignant tumor candidate is displayed on the three-dimensional image, for example, as a spherical region.
[0039]
The user uses the CAD terminal device 401 to study the result 261 of the CAD data processing in detail, and narrows down, for example, malignant tumor candidates.
[0040]
The user uses the CAD terminal device 401 to reconstruct an image including the narrowed malignant tumor candidate at a fine slice interval in order to further examine the narrowed malignant tumor candidate. To send a new image reconstruction processing parameter 422 to the.
[0041]
Based on the received new image reconstruction parameter 422, the image reconstruction device 212 reconstructs an image of the region including the malignant tumor candidate at a fine slice interval, and transfers the obtained image data 222 to the CAD data processing device 251. .
[0042]
The CAD terminal device 401 transmits the new CAD data processing parameter 423 to the CAD data processing device 251. The CAD data processing device 251 performs CAD data processing on the image data 222 using the new CAD data processing parameter 423 transmitted from the CAD terminal device 401, and sends the CAD data processing result 261 to the CAD terminal device 401. Send.
[0043]
Step 1: The user performs the following using the CAD terminal device 401.
(1) Projection data is designated and transmitted as 421 to the projection data storage device.
(2) Designate image reconstruction parameters and send them to the image reconstruction processing device 212 as 422. For example, it instructs to reconstruct a range of 20 cm in the body axis direction of the chest at 1 mm intervals and create 200 images.
(3) CAD parameters are designated and transmitted as 423 to the CAD data processing device 251. For example, a spatial filter that detects an abnormal shadow of 1 mm is specified.
Step 2: The projection data storage device 211 reads out the specified projection data based on the instruction 421 from the CAD terminal device 401, and transfers it to the image reconstruction processing device 212.
Step 3: The image reconstruction processing device 212 reconstructs an image of the projection data 221 transferred from the projection data storage device using the image reconstruction parameter 422 instructed from the CAD terminal device 401, and reconstructs the reconstructed image. The data 222 is transferred to the CAD data processing device 251.
Step 4: The CAD data processing device 251 performs CAD data processing on the image data 222 transferred from the image reconstruction device using the CAD data processing parameter 423 specified by the CAD terminal device 401, and detects the detected malignant tumor. Mark the region of interest where the candidate is located. The result 261 of the CAD data processing is transferred to the CAD terminal device 401.
Step 5: Using the CAD terminal device 401, the user examines the result 261 of the CAD data processing including the detected region of interest transmitted from the CAD data processing device 251.
Step 6: The user picks up a region of interest deemed necessary to be re-examined among the regions of interest detected by the CAD data processing device 251 and executes the following using the CAD terminal device 401.
(1) A new image reconstruction area and a new image reconstruction parameter are designated, and these are transmitted to the image reconstruction processing device 212 as 422. For example, a range of 50 mm in diameter and 10 mm in the body axis direction centered on the abnormal shadow candidate is reconstructed at 10 intervals at 0.1 mm intervals, and 1000 images are created.
(2) Designate a new CAD parameter and transmit it to the CAD data processing device 251 as 423. For example, a filter for detecting an abnormal shadow of 0.1 mm is specified.
Step 7: The image reconstruction device 212 performs image reconstruction of the projection data using the new image reconstruction region and the new image reconstruction parameter 422 specified by the CAD terminal device 401, and obtains the obtained image data 222. Is transferred to the CAD data processing device 251.
Step 8: The CAD data processing device 251 performs CAD data processing on the image data 222 transferred from the image reconstruction device 212 using the new CAD data processing parameter 423 specified by the CAD terminal device 401, and detects The region of interest where the malignant tumor candidate exists is marked. The result 261 of the CAD data processing is transferred to the CAD terminal device 401.
Step 9: Using the CAD terminal device 401, the user examines the result 261 of the CAD data processing including the detected region of interest transmitted from the CAD data processing device 251.
Step 10: If there is an area to be further inspected, steps 6 to 9 are repeated.
[0044]
A CAD system according to the present invention includes a projection data storage device that stores projection data collected by an X-ray CT device. First, image data of medium spatial resolution is reconstructed using the projection data stored in the projection data storage device. CAD data processing is performed using the image data obtained by this image reconstruction, and a region of interest including a shadow of a malignant tumor candidate is detected. Next, from the projection data, an image including a region of interest including the shadow of the malignant tumor candidate is reconstructed with high spatial resolution. CAD data processing is performed using the image data reconstructed at the high spatial resolution, and a region of interest including a shadow of a malignant tumor candidate is detected with high accuracy.
[0045]
In this way, the projection data is used to initially reconstruct an image of the entire chest region with a moderate spatial resolution, and the obtained image data is used to screen CAD data of a region of interest including a shadow of a malignant tumor candidate. Processing is performed to detect a region of interest including a shadow of a malignant tumor candidate. Next, using the projection data, a spatial region including the region of interest detected by the screening CAD data processing is reconstructed with high spatial resolution, and the obtained image data is used to perform a fine CAD data processing. Then, the shadow of the malignant tumor candidate is confirmed. The image reconstruction of the entire chest region has a medium spatial resolution, and the image reconstruction of the spatial region including the region of interest is performed at a high spatial resolution, thereby reducing the total number of image data to be reconstructed and increasing the required region. It has become possible to obtain image data with a spatial resolution, thereby improving the accuracy of CAD data processing.
[0046]
When the user who observes the displayed CAD data processing result performs an operation of designating a new image reconstruction processing parameter again, the CAD terminal device 401 sends an image reconstruction processing request 422 including the new image reconstruction processing parameter to the image reconstruction processing. The image reconstruction apparatus 212 performs image reconstruction processing based on the new image reconstruction processing parameters, and transfers the created image data 222 to the CAD data processing apparatus 251. The CAD data processing device 251 having received this data performs CAD data processing based on the new CAD data processing parameter 423 sent from the user, and transmits the result to the CAD terminal device 401. By repeating this series of flows, the user can interactively perform image reconstruction processing and CAD data processing.
[0047]
In image reconstruction, the image reconstruction area in the tomographic plane of the subject and the axial image reconstruction area, that is, the reconstruction center and spatial resolution in the tomographic plane, and the axial reconstruction center and spatial resolution can be freely set. , It is possible to perform image reconstruction. This makes it possible to use image data with appropriate reconstruction area and spatial resolution for CAD data processing for screening and CAD for fine examination focusing on the region of interest pointed out by screening. Can be.
[0048]
The CAD data processing performed by the CAD data processing device 251 includes two-dimensional spatial filter processing using cross-sectional image data and three-dimensional spatial filter processing using a three-dimensional image constructed from image data. By changing the spatial domain for image reconstruction and the spatial resolution of the image data and the size of the two-dimensional and three-dimensional spatial filters at the screening stage and the detailed inspection stage, the increase in It has become possible to detect various shadow candidates.
[0049]
【Example】
The description so far has described a projection data storage device for storing projection data, an image reconstruction processing device for reconstructing image data of projection data, and a CAD data processing device for performing CAD data processing on image reconstructed image data. A system arranged on a network has been described. A projection data storage device for storing projection data, an image reconstruction processing device for reconstructing image data of projection data, and a CAD for performing CAD data processing on image reconstructed image data Systems integrating two or all of the data processing devices are likewise included.
[0050]
The description so far has described a projection data storage device for storing projection data, an image reconstruction processing device for reconstructing image data of projection data, and a CAD data processing device for performing CAD data processing on image reconstructed image data. Although an example in which one device is arranged on the network has been described, the number is not limited to one, and a plurality of devices can be distributed and installed on the network.
[0051]
In the description so far, as the CAD data processing performed by the CAD data processing apparatus, a three-dimensional image is constructed by using a system for performing two-dimensional filter processing on image reconstructed image data and image data, and a tertiary image is constructed. Although the system for performing the original spatial filter processing has been described, these are examples of CAD data processing and include all image processing performed on image data obtained by image reconstruction.
[0052]
As the projection data to be transmitted to the projection data storage device, compressed data obtained by performing various data compressions to reduce the data amount can be used. The projection data storage device can store the compressed projection data.
[0053]
In the description so far, the projection data stored in the projection data storage device is read out, and the image reconstruction is performed using the image reconstruction processing parameter and the CAD data processing parameter designated by one user for the projection data. Although a case has been described where the processing and CAD data processing are performed and the result is transmitted to one user, the number of users is not limited to one. A plurality of users participate on the network in the same manner as in a video conference, and perform image reconstruction processing and CAD data processing using image reconstruction processing parameters and CAD data processing parameters specified by one user for projection data. It is also possible to perform processing and transmit the result to a plurality of users, and to perform image reconstruction using image reconstruction processing parameters and CAD data processing parameters specified by a plurality of users for the projection data. It is also possible to perform the configuration processing and the CAD data processing and transmit the result to a plurality of users. A plurality of users can share the process of image reconstruction processing and CAD data processing and the results.
[0054]
Although the description so far has focused on the chest, this is merely an example, and includes parts of the whole body where CAD is effective.
[0055]
In the description so far, a set of projection data stored in the projection data storage device is read out, and image reconstruction processing and CAD data processing parameters specified for the projection data are used to perform image reconstruction processing. The case where the CAD data processing is performed and the result is transmitted to the user has been described. However, a plurality of sets of projection data stored in the projection data storage device are read, and an image designated for each of the plurality of sets of projection data is read. Image reconstruction processing is performed using the reconstruction processing parameters, CAD data processing is performed on the obtained plural sets of image data using the designated CAD data processing parameters, and the result is transmitted to the user. The case is also included. This may be the case, for example, when comparing the same patient's set of projection data without contrast agent and the set of projection data with contrast agent, or the same patient's set of projection data from the current examination and the previous examination. And comparing temporal changes using a set of projection data. Since the projection data collected by such a long-term follow-up examination of the patient can be stored in the projection data storage device for a long period of time, it is easy to detect a change over time in the symptoms.
[0056]
In the description so far, the projection data collected by the X-ray CT apparatus is stored in the projection data storage apparatus, and when the CAD inspection is performed, the stored projection data is read out to perform image reconstruction processing and CAD data processing. Has been described, but not only the X-ray CT device but also the MR device and other medical image diagnostic devices are used to store collected data in a large-capacity data storage device and perform CAD inspection. A CAD system for all medical images, which performs the image reconstruction processing and the CAD data processing by reading out the stored collected data, is also included.
[0057]
This includes not only collected data collected by a plurality of sets of the same modality but also collected data collected by a plurality of sets of different modalities.
[0058]
【The invention's effect】
A computer interpretation support system according to the present invention includes a mechanism for accumulating projection data collected by an X-ray CT apparatus in a large-scale data storage device via a network, and a method for storing projection data desired by a user read from the large-scale data storage device. , A mechanism for performing image reconstruction processing using image reconstruction parameters desired by a user, and a mechanism for performing CAD data processing using CAD data processing parameters desired by a user on image reconstructed image data And a mechanism for transferring the results to users on the network.
[0059]
Conventionally, projection data collected by an X-ray CT apparatus is rarely stored for a long period of time, and is discarded soon after image reconstruction processing using projection data is completed. For this reason, even when the user wants to use image data reconstructed with image reconstruction parameters suitable for CAD, new image reconstruction cannot be performed because projection data is not stored. This is because it is physically and economically difficult to store projection data in individual X-ray CT apparatuses for a long period of time. Even if projection data is stored for a long period of time, it can be used to reconstruct an image on demand. Performing the configuration depends on the ability of the image reconstructing apparatus to affect the routine X-ray CT examination, and is difficult in terms of time, physical, and economy. In the present invention, a mechanism for storing collected data in a large-scale data storage device via a network, and image reconstruction using projection reconstruction data stored in the large-scale data storage device using image reconstruction parameters desired by a user. These problems could be solved by providing a processing mechanism.
[0060]
In the present invention, a mechanism for storing collected data in a large-scale data storage device via a network, and image reconstruction using projection reconstruction data stored in the large-scale data storage device using image reconstruction parameters desired by a user. The provision of the mechanism for performing the processing makes it possible for the user to perform the CAD inspection while freely changing the image processing parameters and the CAD data processing parameters using the desired projection data. In this way, the CAD data processing for screening and the CAD for detailed examination centering on the region of interest pointed out in the screening, the spatial resolution in the tomographic plane and the spatial resolution in the axial direction of the subject, that is, It has become possible to freely set the image reconstruction area in the tomographic plane and the image reconstruction area in the axial direction, and perform image reconstruction. As a result, it has become possible to use optimal image data for CAD data processing for screening and CAD for fine examination focusing on a region of interest pointed out by screening.
[0061]
Conventionally, when a user on a network uses projection data stored in a large-scale data storage device, the projection data is transferred from the large-scale data storage device to a computer device such as a workstation owned by the user. It had to be transferred over the network. In the case of large amounts of data such as projection data, this data transfer takes a considerable amount of time, and this data transfer increases network traffic. According to the present invention, a user on a network stores the projection data stored in the large-scale data storage device in the large-scale data storage device by processing the projection data in the high-speed data processing device installed near the large-scale data storage device. It has made it possible to suppress an increase in network traffic when using projected data.
[0062]
Multiple users participate on the network to perform image reconstruction processing on projection data selected by one or more users using image reconstruction processing parameters specified by one or more users. It is possible to perform CAD data processing on image data obtained by using CAD data processing parameters designated by one or a plurality of users, and transmit the result to a plurality of users. This allows a plurality of users to share the process of image reconstruction processing and CAD data processing and the results.
[0063]
Since the data processing parameters used for the data processing can be stored in a large-scale data storage device on the network, other users can restore the data processing using these parameters.
[0064]
A large-capacity data storage device that stores projection data and a high-speed data processing device that performs image reconstruction processing of the stored projection data and performs CAD data processing on the image-reconstructed image data are installed on a network. An environment has been constructed that can store projection data, perform image reconstruction processing on the projection data, and perform CAD data processing on the network. This eliminates the need for each user to store projection data. Although an image reconstructing apparatus for reconstructing projection data is expensive, the present invention has made it possible for a plurality of users to jointly use the same on a network, thereby reducing the cost per user. .
[0065]
Further, a user on a network that performs image processing such as CAD data processing performs image reconstruction processing using reconstruction processing parameters desired by the user, and performs image reconstruction processing on the image data that has been subjected to the image reconstruction processing. As a result, image processing using image processing parameters desired by the user can be performed, so that it is possible to use optimal image data for image processing such as CAD data processing.
[0066]
Conventionally, when processing large-capacity data, it was necessary for each user to prepare a computer device having a high data processing capability such as a workstation. Since the device performs data processing, the user's terminal device only instructs projection data selection, image reconstruction parameters, and CAD data processing parameters, and only displays the processing results sent. With a simple personal computer, it has become possible to obtain results equal to or higher than those of the workstation of the conventional system.
[0067]
Conventionally, to process data stored on a network, each user had to prepare a high-speed data processing device such as a workstation. The use efficiency of workstations owned by each user is not very high. According to the present invention, since the high-speed data processing device on the network performs data processing in response to image processing requests of users on a plurality of networks, the utilization efficiency of the high-speed data processing device on the network can be increased. Therefore, it is also excellent in cost efficiency.
[0068]
A plurality of sets of projection data stored in the projection data storage device are read, and image reconstruction processing is performed on the plurality of sets of projection data using the designated image reconstruction processing parameters. CAD data processing can be performed using the designated CAD data processing parameters for the image data. It is possible to compare using a set of projection data without contrast agent and a set of projection data using contrast agent of the same patient, and to compare the set of projection data of the same patient with the projection data of the previous examination. Comparing changes over time using sets is now easier to perform.
[0069]
The projection data collected by the long-term follow-up examination of the patient can be stored in the projection data storage device for a long period of time, so that it becomes easy to detect the change over time of the symptoms. This is a great effect of this CAD system.
[0070]
Since a large-capacity data storage device and an image reconstruction device are installed on the network, the collected data collected by all medical image diagnostic devices such as MR devices as well as X-ray CT devices is stored in the large-capacity data storage device. In addition, it has become easy to construct a CAD system for all medical images for which image accumulation processing and CAD data processing are read out by reading out the accumulated data when performing a CAD examination.
[0071]
This facilitates construction of a CAD system that uses not only collected data collected by a plurality of sets of the same modality but also collected data collected by a plurality of sets of different modalities.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a computer diagnosis support system in a network environment according to the present invention.
FIG. 2 is a block diagram showing a data flow of a computer diagnosis support system in a network environment according to the present invention.
[Explanation of symbols]
101 Network
102 Network
103 Network
201 Projection Data Storage and Processing Center
211 Projection data storage device
212 image reconstruction processing device
221 Projection data
222 Reconstructed image data
251 CAD data processing device
261 CAD data processing result data
301 X-ray CT system
302 Console of X-ray CT system
311 X-ray CT system body
312 Pretreatment device for X-ray CT system
313 Image reconstruction device of X-ray CT system
321 Projection data obtained by the scanner unit of the X-ray CT system
322 Projection data pre-processed by the pre-processing device of the X-ray CT system
323 Projection data transferred to the projection data storage device preprocessed by the preprocessing device of the X-ray CT apparatus
324 Image data reconstructed
401 CAD terminal device
402 Personal Computer
421 Projection data specification
422 Image reconstruction processing parameter indication
423 CAD data processing parameter indication

Claims (6)

Medical image diagnostic devices such as X-ray CT devices and MR devices connected to the network, large-scale data storage devices, high-speed data processing devices for image reconstruction and CAD data processing, and X-ray CT devices and MR devices A mechanism for storing collected data, such as projection data collected by medical image diagnostic equipment, in a large-scale data storage device via a network, and reading out collected data desired by the user from the large-scale data storage device and transferring it to a high-speed data processing device A mechanism for performing image reconstruction processing on the acquired data using the image reconstruction processing parameters desired by the user, and a mechanism for performing computer interpretation desired by the user on the image reconstructed image data A mechanism for performing CAD data processing using support data processing (hereinafter referred to as CAD data processing) parameters. Computer interpretation support system in a network environment; and a mechanism to transfer the results of Tsu to the user. 2. A mechanism for performing image processing such as three-dimensional image reconstruction on an image reconstructed image data by using an image processing parameter desired by a user, and using the image reconstructed image data on the image reconstructed image data. A computer reading support system in a network environment, comprising: a mechanism for performing CAD data processing using a CAD data processing parameter desired by a user; and a mechanism for transferring a result of the CAD data processing to a user on a network. 2. A method according to claim 1, further comprising: sequentially narrowing down a spatial area for reconstructing the image of the projection data based on a result of the CAD data processing, and using an image reconstruction processing parameter suitable for the new spatial area. A mechanism for performing reconstruction, a mechanism for performing image processing such as three-dimensional image reconstruction on image-reconstructed image data using image processing parameters suitable for the new spatial region, and a mechanism for performing image processing on the image-processed image data. A computer-readable reading support system comprising a mechanism for performing CAD data processing using CAD data processing parameters suitable for a new spatial area, and capable of interactively changing the spatial area and spatial resolution. In claim 1, claim 2 or claim 3, a spatial domain for reconstructing projection data and an image reconstruction so that image data having a spatial resolution most suitable for a spatial filter used for CAD data processing can be obtained. A computer interpretation support system including a mechanism for determining a processing parameter. 5. The high-speed computer according to claim 1, wherein the large-capacity data stored in the large-scale data storage device is stored in a network such as a super-high-speed fiber channel or the like. By processing with a high-speed data processing device connected by an input / output bus, it is possible to suppress an increase in network traffic when users on a network use projection data stored in a large-scale data storage device. Computer reading support system in a closed network environment. A medical image diagnostic apparatus of a plurality of modalities such as an X-ray CT apparatus and an MR apparatus connected to a network and medical images of the plurality of modalities according to the first, second, third, fourth, and fifth aspects. A computer interpretation support system that comprehensively uses image data of a plurality of modalities corresponding to collected data collected by a diagnostic device.

Images