JP7312672B2 - MEDICAL IMAGE PROCESSING APPARATUS AND MEDICAL IMAGE CORRECTION METHOD - Google Patents

Description

The present embodiment relates to a medical image processing apparatus and a medical image correction method.

2. Description of the Related Art In recent years, 3T (Tesla) MRI apparatuses, for example, have come into widespread use as magnetic resonance imaging (MRI) apparatuses have become higher magnetic fields. A 3T MRI apparatus has a higher resonance frequency and a shorter wavelength of RF (Radio Frequency) pulses to be transmitted than a 1.5T MRI apparatus. As a result, an RF pulse with a short wavelength is attenuated in the subject, resulting in non-uniformity. Along with this non-uniformity, echo signals also become non-uniform. This phenomenon is called inhomogeneity of the high-frequency magnetic field.

As a technique for reducing the non-uniformity of the high-frequency magnetic field, for example, a technique of performing brightness correction based on information on the sensitivity distribution of the coil has been proposed. However, if information on the sensitivity distribution of the coil is collected at a lower resolution than the imaging for generating the diagnostic image in order to speed up the imaging, the sharpness of the contour portion may be impaired as a result of the brightness correction processing, the influence of the interpolation processing, and the like.

Japanese Patent Application Laid-Open No. 2005-237702

An object of the present embodiment is to perform luminance correction without impairing the sharpness of an image.

A medical image correction method according to an embodiment obtains an MR image, divides the MR image into a plurality of groups using a histogram of pixel values of the MR image to generate information for grouping luminance values in the MR image, divides the MR image into regions using information indicating continuity between pixels in the MR image and information for grouping the luminance values in the MR image, calculates a distribution of signal non-uniformity for each of the divided regions, and calculates the distribution of signal non-uniformity. Correcting the MR image based on the distribution .

FIG. 1 is a block diagram showing the configuration of a medical image processing apparatus 1 according to an embodiment. FIG. 2 is a flow chart showing the flow of processing executed in the medical image correction method according to the embodiment. FIG. 3 is a histogram of luminance values of pixels of the base image. FIG. 4 is a diagram exemplifying a part of the base image whose histogram is divided into a plurality of groups in step S2 of FIG. FIG. 5 is a diagram showing a part of the base image from which edges are extracted by the edge extraction process in step S4 of FIG.

A medical image correction method and a medical image processing apparatus according to embodiments will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of a medical image processing apparatus 1 that executes a medical image correction method according to an embodiment.

The medical image processing apparatus 1 is, for example, a dedicated or general-purpose computer. Note that the medical image processing apparatus 1 may have functions 111 and 112, which will be described later. For example, the functions of the medical image processing apparatus 1 may be included in a medical image diagnostic apparatus such as an MRI apparatus connected via a network, a PC (workstation) that performs image processing on medical images, a medical image management apparatus (server) that stores and manages medical images, and the like.

A case where the medical image processing apparatus 1 is a dedicated or general-purpose computer will be described below as an example.

The medical image processing apparatus 1 includes a processing circuit 11 , an input circuit 12 , a display circuit 13 , a communication I/F section 14 and a memory circuit 15 .

The processing circuit 11 is a processor that reads a program from the storage circuit 15 and executes it to realize a function corresponding to each program. The processing circuit 11 has, for example, an image reconstruction function 111 and a brightness correction function 112 . The processing circuit 11 reads various control programs stored in the storage circuit 15 to implement the image reconstruction function 111 and the luminance correction function 112, and also controls the processing operations in the input circuit 12, the display circuit 13, the communication I/F unit 14, and the storage circuit 15. In other words, the processing circuit 11 in a state where each program is read has each function shown in the processing circuit 11 of FIG.

The image reconstruction function 111 is a function that performs reconstruction processing such as Fourier transform on k-space data to generate an MR image. The k-space data is data obtained by arranging the MR data acquired by the MRI apparatus according to the phase encoding amount and the frequency encoding amount applied by the gradient magnetic field.

The brightness correction function 112 is a function that executes the medical image correction method according to this embodiment, and is an example of a division unit, a calculation unit, and a correction unit. That is, the luminance correction function 112 divides the MR image into regions using information indicating continuity between pixels in the MR image and information for grouping luminance values in the MR image, calculates the distribution of signal non-uniformity for each divided region, and corrects the MR image based on the distribution of signal non-uniformity. The MR image corrected by the luminance correction function 112 may be acquired from an image server or an external magnetic resonance imaging apparatus via the communication I/F circuit 14 described later, or may be an MR image reconstructed by the image reconstruction function 111. Furthermore, MR images stored in the storage circuit 15 may be obtained.

In FIG. 1, the processing circuit 11, which is a single processor, has been described as realizing the processing functions performed by the image reconstruction function 111 and the luminance correction function 112. However, a processing circuit may be configured by combining a plurality of independent processors, and the functions may be realized by each processor executing a program. 1, a single memory circuit 15 stores a program corresponding to each processing function, but a plurality of memory circuits 15 may be distributed and the processing circuit 11 may read out the corresponding program from each memory circuit 15.

The term "processor" used in the above description includes, for example, a CPU (Central Processing Unit), a GPU (Graphical Processing Unit), an Application Specific Integrated Circuit (ASIC), a programmable logic device (for example, a Simple Programmable Logic Device). mmable Logic Device (SPLD), Complex Programmable Logic Device (CPLD), and Field Programmable Gate Array (FPGA)). The processor implements its functions by reading and executing programs stored in the storage circuit 15 . Instead of storing the program in the memory circuit 15, the program may be directly incorporated into the circuit of the processor. In this case, the processor realizes its function by reading and executing the program embedded in the circuit.

The input circuit 12 is a circuit for inputting signals from an input device such as a pointing device (mouse, etc.) or a keyboard that can be operated by an operator. Here, the input device itself is also included in the input circuit 12. When the operator operates the input device, the input circuit 12 generates an input signal according to the operation and outputs it to the processing circuit 11 . Note that the medical image processing apparatus 1 may include a touch panel in which an input device is integrated with the display circuit 13 .

The input circuit 12 is realized by a trackball, a switch button, a mouse, a keyboard, etc. for setting a region of interest (ROI), a touch pad for performing input operations by touching the operation surface, a touch pad for performing input operations by touching the operation surface, a touch screen in which a display screen and a touch pad are integrated, a touch panel display in which a display screen and a touch pad are integrated, a non-contact input circuit using an optical sensor, an audio input circuit, a display screen and a touch pad, and the like.

It should be noted that the input circuit 12 is not limited to having physical operation parts such as a mouse and keyboard. For example, the input circuit 12 also includes an electrical signal processing circuit that receives an electrical signal corresponding to an input operation from an external input device provided separately from the device and outputs the electrical signal to the control circuit.

The display circuit 13 is a display for displaying images, and is configured by an LCD (Liquid Crystal Display) or the like. The display circuit 13 displays various operation screens and various display information such as image data on the LCD according to instructions from the processing circuit 11 .

A communication I/F (interface) circuit 14 performs a communication operation with an external device according to a predetermined communication standard. When medical image processing apparatus 1 is provided on a network, communication I/F circuit 14 transmits and receives information to and from external devices on the network. For example, the communication I/F circuit 14 receives data obtained by imaging from a medical image diagnostic apparatus such as an MRI apparatus or a medical image management apparatus.

The storage circuit 15 is composed of a semiconductor memory device such as a RAM (Random Access Memory), a flash memory, a hard disk, an optical disk, or the like. The storage circuit 15 may be configured by portable media such as USB (Universal Serial Bus) memory and DVD (Digital Video Disk).

The storage circuit 15 includes various processing programs (application programs, OS (Operating System), etc.) used in the processing circuit 11, and stores data necessary for executing the programs, volume data, and medical images. In addition, the OS may include a GUI (Graphical User Interface) in which graphics are often used to display information on the display circuit 13 for the operator, and basic operations can be performed by the input circuit 12. I can.

Next, a medical image correction method according to this embodiment will be described.

FIG. 2 is a flow chart showing the flow of processing executed in the medical image correction method according to this embodiment.

First, the processing circuit 11 acquires an MR image using the brightness correction function 112 (step S1). The MR image acquired in step S1 is hereinafter referred to as a "base image".

The procedures of steps S2, S3, and S4 that follow correspond to pre-processing of the fitting process that is performed in step S5, which will be described later. In step S5, the unevenness of luminance values in the base image is estimated by dividing the base image into several regions, and the regions are determined in steps S2, S3, and S4.

The processing circuit 11 uses the luminance correction function 112 to generate a histogram of the luminance values of the pixels of the base image, and uses the generated histogram to group the pixels of the base image according to the luminance values (step S2).

FIG. 3 is a histogram of luminance values of pixels of the base image. The luminance correction function 112 of the processing circuit 11 uses the histogram to divide the luminance values into four groups a, b, c, and d, and labels each pixel of the base image with one of the four groups. Note that the number of divisions of the histogram shown in FIG. 3, that is, the number of groups, is merely an example.

Further, before creating a histogram in step S2, a filtering process may be performed on the base image for the purpose of suppressing deterioration of the accuracy of grouping.

FIG. 4 is a diagram exemplifying how each pixel is grouped in step S2 for a part of the base image. In FIG. 4, the symbols a, b, c, and d indicate to which group each pixel is classified. In the base image, information indicating to which of the plurality of groups divided according to the luminance value each pixel belongs is hereinafter referred to as group information according to the luminance value.

In parallel with step S2, the processing circuit 11 uses the brightness correction function 112 to perform edge extraction processing on the base image, and acquires information indicating continuity between pixels in the base image (step S3).

FIG. 5 is a diagram exemplifying a part of the base image subjected to edge extraction processing in step S3. In FIG. 5, the group information corresponding to the luminance value is shown superimposed, but it is only for convenience of explanation. The edge extraction processing in step S3 does not include a procedure for classifying each pixel into one of groups a, b, c, and d.

Continuity between pixels is detected by edge extraction processing in step S3, and a boundary L is detected as shown in FIG. 5, for example.

Next, the processing circuit 11 uses the luminance correction function 112 to divide the base image into a plurality of regions for unevenness estimation using the group information corresponding to the luminance value generated in step S2 and the information indicating the continuity between pixels obtained in step S3 (step S4).

In FIG. 4, a pixel group Pb made up of six pixels classified into group b locally exists in an area where many pixels classified into group a are distributed. If the group of pixels Pb and its neighboring pixels represent signals from different tissues, contrast must be preserved. On the other hand, if the pixel group Pb and its neighboring pixels show signals from the same tissue, it is considered that luminance unevenness occurs, and thus luminance correction is necessary.

Similarly, in FIG. 4, a pixel group Pc consisting of 6 pixels classified into group c locally exists in an area where many pixels classified into group d are distributed. If the group of pixels Pc and its neighboring pixels represent signals from different tissues, contrast must be preserved. On the other hand, if the pixel group Pc and its neighboring pixels show signals from the same tissue, it is considered that luminance unevenness occurs, and thus luminance correction is required.

The processing circuit 11 uses the luminance correction function 112 to divide the base image so that groups of pixels having continuity between pixels are in the same region. For example, in the group information according to the luminance value shown in FIG. 4, the pixel group Pb belongs to a group different from the surrounding pixels, but referring to the information indicating the continuity between pixels shown in FIG. 5, it is not surrounded by edges. Therefore, the classification of each pixel of pixel group Pb is changed from group b to group a. Similarly, the classification of each pixel in pixel group Pc is changed from group c to group d. The above procedure completes the division of the base image for unevenness estimation. Group information is assigned to each pixel of the base image divided into a plurality of regions, but since edge information is considered in step S4, the group information assigned to each pixel may be changed from the point of step S2.

Next, the processing circuit 11 uses the luminance correction function 112 to perform function fitting of changes in the luminance value of the base image for each region divided in step S4 (step S5). Any fitting process may be used, but as a typical example, a technique using low-order equations such as second-order and third-order equations can be adopted. Here, the information such as the parameters defining the function obtained by the fitting process is hereinafter referred to as the fitting result.

Next, the processing circuit 11 uses the luminance correction function 112 to fit the change in luminance value over the entire base image with a function using the fitting result obtained for each region in step S5. Any fitting process may be used, but as a typical example, a technique using low-order equations such as second-order and third-order equations can be employed. Based on the result of fitting to the entire base image obtained in step S6, the luminance correction function 112 generates a luminance correction map used to equalize the luminance of the base image (step S6). Note that the processing in step S6 can also be regarded as processing for suppressing unnatural edge-like pixel value changes that may occur when the fitting result obtained in step S5 is used for brightness correction for each divided region.

Next, the processing circuit 11 uses the brightness correction map generated by the brightness correction function 112 to correct the brightness of the base image acquired in step S1 (step S7).

The display circuit 13 outputs the luminance-corrected base image as a display image (step S8).

As described above, the medical image correction method according to the present embodiment acquires an MR image as a base image, and divides the base image into regions using group information corresponding to luminance values and information indicating continuity between pixels. In addition, based on the results of functional fitting of changes in luminance value for each divided region, information on non-uniformity in luminance value of the entire base image is obtained. Then, based on the acquired information about unevenness of luminance values, a luminance correction map for equalizing the luminance values of the base image is created, and the luminance values of the base image are corrected.

The group information according to the luminance values in the MR image as the base image is generated by classifying the pixels included in the base image into a plurality of groups using, for example, a histogram of luminance values. Information indicating the continuity between pixels of the MR image is generated, for example, by performing edge extraction processing on the base image. The group information corresponding to the luminance value, the information indicating the continuity between pixels, and the luminance correction map generated using these information are all generated from the base image. Therefore, the brightness correction map has the same resolution as the base image.

As a comparative example, there is a brightness correction method using a coil sensitivity map for non-uniformity correction of transmission RF and non-uniformity correction caused by a receiving coil. The coil sensitivity map used in such luminance correction generally has a smaller matrix size and a lower resolution than when capturing a diagnostic image, so the contour may not be correctly estimated due to the influence of interpolation processing and the like. In other words, when luminance correction is applied to a portion of an image where the pixel value changes sharply, the change in pixel value may be smoothed out if the correction is performed using a sensitivity map with a small matrix size.

On the other hand, the medical image correction method according to the present embodiment uses a luminance correction map having the same resolution as that of the diagnostic image, so that it is possible to prevent deterioration of the sharpness of the diagnostic image when performing luminance correction.

The group information corresponding to the luminance value in the MR image is acquired by grouping using the histogram of the luminance value, so that it is information in which pixels with similar luminance are grouped together. Information indicating the continuity between pixels in the MR image is obtained by edge extraction processing, and thus is information that ensures tissue continuity. Therefore, the brightness correction map generated using the information indicating the continuity between pixels and the group information corresponding to the brightness value makes it possible to perform the brightness correction while ensuring the continuity of the tissue while not impairing the original contrast of the MR image due to the group information.

The medical image correction method according to the present embodiment can perform luminance correction without receiving parameters such as correction strength from the user. That is, since it is not necessary for the user to consider changing the correction parameter for each imaging sequence, the burden on the user can be reduced.

In addition, in the medical image correction method according to the present embodiment, after estimating the change in luminance value for each divided region by function fitting, based on the estimation result, the change in luminance value is further estimated by function fitting for the entire base image. Although it is possible to improve the accuracy of luminance correction even with the fitting results obtained for each divided region, by further performing function fitting on the entire base image, it is possible to suppress the possibility of unnatural luminance value changes occurring at the boundaries of group division in the image subjected to luminance correction.

(Modification 1)
In the medical image correction method according to the above embodiment, an MR image that has not been subjected to specific luminance correction is acquired in step S1 and used as a base image for luminance correction. On the other hand, the MR image acquired in step S1 may be subjected to conventional luminance correction such as transmission RF non-uniformity correction and reception coil non-uniformity correction, and the medical image correction method according to the present embodiment may be performed using that image as a base image.

According to the configuration according to Modification 1, it is possible to improve image unevenness that could not be corrected by the existing luminance correction, and to improve the visibility of the image.

According to at least one embodiment described above, luminance correction can be performed without impairing the sharpness of an image.

Also, while several embodiments of the invention have been described, these embodiments have been presented by way of example and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and modifications can be made without departing from the scope of the invention. These embodiments and their modifications are included in the scope and spirit of the invention, as well as the scope of the invention described in the claims and equivalents thereof.

1 medical image processing apparatus 11 processing circuit 12 input circuit 13 display circuit 14 communication I/F circuit 15 storage circuit 111 image reconstruction function 112 luminance correction function

Claims (8)

acquire an MR image;
generating information for grouping luminance values in the MR image by dividing the MR image into a plurality of groups using a histogram of pixel values of the MR image;
segmenting the MR image into regions using information indicating continuity between pixels in the MR image and information for grouping the luminance values in the MR image;
calculating a distribution of signal non-uniformity for each of the divided regions;
correcting the MR image based on the distribution of the signal non-uniformity;
Medical image correction method. edge detection processing on the MR image to generate information indicating continuity between the pixels;
The medical image correction method according to claim 1. the segmentation is performed on the MR image corrected for non-uniformity of transmitted high frequencies;
The medical image correction method according to claim 1 or 2 . The segmentation is performed on the MR image corrected for non-uniformity due to the sensitivity distribution of the receiving coil.
The medical image correction method according to any one of claims 1 to 3 . performing a fitting process for each of the divided regions on the distribution of the signal non-uniformity calculated for each of the divided regions;
The medical image correction method according to any one of claims 1 to 4 . After performing the fitting process for each of the divided regions, performing the fitting process on the entire MR image;
The medical image correction method according to claim 5 . segmenting the MR image includes segmenting the MR image by changing information for grouping the luminance values based on information indicating continuity between pixels;
The medical image correction method according to claim 1. a dividing unit that divides the MR image into a plurality of groups by using a histogram of acquired pixel values of the MR image to generate information for grouping luminance values in the MR image, and divides the MR image into areas using information indicating continuity between pixels in the MR image and information for grouping the luminance values in the MR image ;
a calculation unit that calculates a distribution of signal non-uniformity for each of the divided regions;
a correction unit that corrects the MR image based on the distribution of the signal non-uniformity;
A medical image processing device with

Images