JP2010200920A - Image processing apparatus and method for controlling the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a mechanism for estimating a normal structure of a predetermined layer constituting an object with high accuracy. <P>SOLUTION: The image processing apparatus includes: an image acquisition unit 220 for acquiring a tomographic image of the object; an object information acquisition unit 210 for acquiring normal-time data of the predetermined layer that constitutes the object; a layer detection unit 251 for detecting the predetermined layer from the tomographic image obtained by the image acquisition unit 220; and a normal structure estimation unit 252 for estimating the normal structure of the predetermined layer, using the detection result of the predetermined layer detected by the layer detection unit 251 and the normal-time data of the predetermined layer obtained by the object information acquisition unit 210. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image processing apparatus for processing a tomographic image of a subject, a control method therefor, a program for causing a computer to execute the control method, and a computer-readable storage medium storing the program.

  For the purpose of early diagnosis of various diseases that occupy the top causes of lifestyle-related diseases and blindness, examinations using imaging of the eye as a subject have been widely performed. For example, an tomographic image acquisition device for an eye using an optical coherence tomography (hereinafter referred to as “OCT”) or the like can three-dimensionally observe the state inside the retinal layer of the eye. . For this reason, this tomographic image acquisition apparatus is expected to be useful for more accurately diagnosing diseases.

Here, a tomographic image (tomographic image) of the eye taken using OCT will be described.
FIG. 10 is a schematic diagram showing an example of the layer structure of the retina of the eye taken using OCT.

  Specifically, FIG. 10 shows an example of a layer structure in the macular portion of the retina. In general, a tomographic image of the eye by OCT is obtained three-dimensionally, but in order to simplify the explanation, one cross-section is shown two-dimensionally in FIG.

  In FIG. 10, reference numeral 1001 denotes a retinal pigment epithelium layer, 1002 denotes a nerve fiber layer, and 1003 denotes an inner limiting membrane. For example, when a tomographic image as shown in FIG. 10A is input, the thickness of the nerve fiber layer 1002 (T1 in FIG. 10A) and the thickness of the entire retinal layer (T2 in FIG. 10A) Can be measured quantitatively to diagnose the degree of progression of diseases such as glaucoma and the degree of recovery after treatment.

  Conventionally, in order to measure the thickness of these layers, a doctor or an engineer manually operates a retina on a two-dimensional tomographic image (B-scan image) obtained by cutting a cross section of interest from a tomographic image that is three-dimensionally imaged. It was necessary to specify the boundary of each layer. For example, in order to examine the thickness (T1) of the nerve fiber layer 1002 in FIG. 10 (a), as shown in FIG. 10 (b), the inner boundary membrane 1003, the nerve fiber layer 1002, and the lower layer thereof are separated. It was necessary to designate a boundary (hereinafter referred to as “nerve fiber layer boundary” 1004) on the tomographic image. Further, in order to examine the thickness of the entire retinal layer (T2 in FIG. 10A), the boundary between the retinal pigment epithelium layer 1001 and the lower layer (hereinafter referred to as “retinal pigment epithelium layer boundary” 1005). ) On the tomographic image.

  In order to obtain a three-dimensional distribution of each layer thickness, the three-dimensional tomographic image is regarded as a set of several two-dimensional tomographic images, and the boundary of the layer of interest is designated on each two-dimensional tomographic image. There was a need.

  However, the task of manually specifying the layer boundaries is a burden for doctors and engineers. In addition, it is difficult to manually quantify the layer boundaries because it causes variations due to differences in the workers and work dates and times. For the purpose of reducing the burden on workers and variations in work, a technique has been proposed in which a computer detects the boundary of each layer of the retina from a tomographic image and measures the thickness of each layer (for example, Patent Document 1 below). And Patent Document 2).

  Moreover, in diseases such as age-related macular degeneration, the shape of the retinal pigment epithelium layer is deformed into irregularities according to the disease state. Therefore, quantifying the degree of deformation is effective for understanding the medical condition.

  FIG. 10 (a) shows an example of the layer structure of the retina in a normal eye, and FIG. 10 (c) shows an example of the layer structure of the retina in age-related macular degeneration. The retinal pigment epithelium layer 1001 in the normal eye shown in FIG. 10A has a structure having a smooth curve. On the other hand, the retinal pigment epithelium layer 1001 in age-related macular degeneration shown in FIG. 10 (c) has a shape in which a part of the region is waved up and down.

  In the following description, the estimated position of the boundary where the retinal pigment epithelial layer boundary would have existed at the normal time before illness is referred to as “normal structure of the retinal pigment epithelial layer boundary” or simply “normal structure”. ". In addition, as a method for diagnosing age-related macular degeneration using OCT, an actual retinal pigment epithelial layer boundary that can be observed on a tomographic image (actually measured value: solid line 1005 in FIG. 10 (d)) and its on the tomographic image A normal structure (estimated value: broken line 1006 in FIG. 10D) is obtained. Then, for each cross section formed by the difference between the actual retinal pigment epithelium layer boundary (1005) observable on the tomographic image and the normal structure (1006) on the tomographic image (the hatched portion in FIG. 10 (d)). The state of the disease is quantified by the area and the total volume of the disease.

  However, conventionally, this work is also often performed manually by a doctor or engineer, and the burden and variations in the results are problematic. Therefore, it is required to automate the detection of the retinal pigment epithelium layer boundary from the tomographic image and the estimation of its normal structure.

  The above-described detection of the retinal pigment epithelium layer boundary can be performed by the techniques disclosed in Patent Document 1 and Patent Document 2. On the other hand, since the normal structure cannot be observed as an image feature on the tomographic image, it cannot be estimated by the techniques of Patent Document 1 and Patent Document 2. Therefore, conventionally, an elliptic curve fitted to minimize the square error with respect to the retinal pigment epithelium layer boundary detected from a two-dimensional tomographic image (B-scan image) is often used as its normal structure. .

  Since the retinal pigment epithelium layer 1001 is a very thin layer, the retinal pigment epithelium layer boundary 1005 is generally sometimes referred to as a “retinal pigment epithelium layer”. To follow that convention, in the following explanation, “retinal pigment epithelial layer boundary” is read as “retinal pigment epithelial layer boundary” and “normal structure of retinal pigment epithelial layer boundary” is read as “normal structure of retinal pigment epithelial layer”. May be.

JP 2008-73099 A JP 2007-325831 A

However, the normal structure estimated by fitting the elliptic curve by the conventional least square method described above often does not match the actual normal structure. Therefore, even if the area based on the difference from the detected actual layer boundary is obtained, the reliability may not be sufficient for use as a scale for quantifying the medical condition.
In addition, when there is a portion where the retinal pigment epithelium layer boundary 1005 cannot be clearly observed on a tomographic image due to the influence of blood vessels or vitiligo (image characteristics cannot be obtained), there is a problem that the detection cannot be performed. .

  The present invention has been made in view of such problems, and an object thereof is to provide a mechanism for accurately estimating the normal structure of a predetermined layer constituting a subject. It is another object of the present invention to accurately detect a predetermined layer even when a part of a predetermined layer boundary constituting the subject is difficult to observe on a tomographic image.

The image processing apparatus of the present invention includes an image acquisition unit that acquires a tomographic image of a subject, a subject information acquisition unit that acquires normal time data of a predetermined layer constituting the subject, and the predetermined layer from the tomographic image. A normal structure of the predetermined layer using a layer detection unit to detect, a detection result of the predetermined layer by the layer detection unit, and normal data of the predetermined layer acquired by the subject information acquisition unit And normal structure estimating means for estimating.
According to another aspect of the image processing apparatus of the present invention, an image acquisition unit that acquires a tomographic image of a subject, a subject information acquisition unit that acquires normal time data of a predetermined layer that constitutes the subject, and the subject information acquisition unit And layer detecting means for detecting the predetermined layer from the tomographic image using the normal data of the predetermined layer acquired in step (1).

An image processing apparatus control method according to the present invention is an image processing apparatus control method for processing a tomographic image of a subject, an image acquisition step of acquiring the tomographic image of the subject, and a predetermined layer constituting the subject Acquired in the subject information acquisition step for acquiring normal data, the layer detection step for detecting the predetermined layer from the tomographic image, the detection result of the predetermined layer by the layer detection step, and the subject information acquisition step A normal structure estimation step of estimating a normal structure of the predetermined layer using the normal data of the predetermined layer.
Another aspect of the control method for an image processing apparatus according to the present invention is a control method for an image processing apparatus that performs processing of a tomographic image of a subject. The image acquisition step for acquiring a tomographic image of a subject and the subject are configured. A subject information acquisition step for acquiring normal data of a predetermined layer, and a layer for detecting the predetermined layer from the tomographic image using the normal data of the predetermined layer acquired in the subject information acquisition step Detecting step.

A program according to the present invention is a program for causing a computer to execute a control method of an image processing apparatus that processes a tomographic image of a subject, and that configures the subject and an image acquisition step of acquiring the tomographic image of the subject. A subject information acquisition step for acquiring normal data of a predetermined layer; a layer detection step for detecting the predetermined layer from the tomographic image; a detection result of the predetermined layer by the layer detection step; and the subject information acquisition. The normal structure estimation step of estimating the normal structure of the predetermined layer using the normal time data of the predetermined layer acquired in the step is executed by a computer.
Another aspect of the program of the present invention is a program for causing a computer to execute a control method of an image processing apparatus that processes a tomographic image of a subject, an image acquisition step of acquiring a tomographic image of the subject, and the subject Subject information acquisition step for acquiring normal time data of a predetermined layer constituting the predetermined layer, and using the normal time data of the predetermined layer acquired in the subject information acquisition step, the predetermined layer is extracted from the tomographic image. This is for causing a computer to execute a layer detection step for detection.

  The computer-readable storage medium of the present invention stores the program.

  According to the present invention, it is possible to accurately estimate the normal structure of a predetermined layer constituting a subject. This makes it possible to quantify the degree of disease progression in the subject, the degree of recovery after treatment, and the like. Further, even when a part of the predetermined layer boundary constituting the subject is difficult to observe on the tomographic image, the predetermined layer can be detected with high accuracy.

1 is a schematic diagram illustrating an example of a schematic configuration of an image processing system according to a first embodiment of the present invention. It is a schematic diagram which shows an example of the hardware constitutions of the image processing apparatus which concerns on the 1st Embodiment of this invention. It is a schematic diagram which shows an example of a function structure of the image processing apparatus which concerns on the 1st Embodiment of this invention. 3 is a flowchart illustrating an example of a processing procedure of a control method of the image processing apparatus according to the first embodiment of the present invention. 5 is a flowchart showing an example of a detailed processing procedure of “normal processing of retinal pigment epithelial layer” in step S106 of FIG. FIG. 5 is a flowchart showing an example of a detailed processing procedure of “retinal pigment epithelial layer abnormality processing” in step S107 of FIG. 4; It is a schematic diagram which shows the 1st Embodiment of this invention and shows an example of the weight function utilized when estimating the normal structure of the layer contained in a tomographic image. It is a schematic diagram which shows an example of a function structure of the image processing apparatus which concerns on the 2nd Embodiment of this invention. It is a flowchart which shows an example of the process sequence of the control method of the image processing apparatus which concerns on the 2nd Embodiment of this invention. It is a schematic diagram which shows an example of the layer structure of the retina of the eye | photography part image | photographed using OCT.

  Hereinafter, embodiments (embodiments) for carrying out the present invention will be described with reference to the drawings. The embodiment described below is only an example, and the present invention is not limited to the illustrated configuration.

(First embodiment)
First, a first embodiment of the present invention will be described.
FIG. 1 is a schematic diagram illustrating an example of a schematic configuration of an image processing system 100 according to the first embodiment of the present invention.

  As shown in FIG. 1, the image processing system 100 includes an image processing device 110, a tomographic image capturing device 120, a data server 130, and a local area network (LAN) 140. That is, the image processing system 100 illustrated in FIG. 1 is configured such that the image processing apparatus 110 is connected to the tomographic image capturing apparatus 120 and the data server 130 via the LAN 140.

  In the example shown in FIG. 1, the image processing apparatus 110, the tomographic image capturing apparatus 120, and the data server 130 are connected via the LAN 140. However, in the present embodiment, the present invention is not limited to this. is not. For example, instead of the LAN 140, an interface such as USB or IEEE1394 may be used, or an external network such as the Internet may be used.

  The image processing apparatus 110 according to the present embodiment acquires, for example, a tomographic image of a subject (in this example, an eye part (macular region); also referred to as “eye to be examined”) imaged by the tomographic image imaging apparatus 120, From the image, the boundary of the layer constituting the eye is detected. Then, the image processing apparatus 110 performs an analysis process based on a predetermined disease in accordance with an instruction from an operator (engineer or doctor), and performs a quantification process of the disease. In the present embodiment, an example in which the eye to be examined (eye part) is applied as an example of the subject is shown. However, the present invention is not limited to this, and may be an imaging target as a tomographic image. If applicable.

  Here, the image processing apparatus 110 performs an analysis process for examination of glaucoma or the like when the operator instructs “normal processing of the retinal pigment epithelium layer”. In this case, for example, as shown in FIG. 10B, the image processing apparatus 110 detects the inner boundary membrane 1003, the nerve fiber layer boundary 1004, and the retinal pigment epithelium layer boundary 1005 from the tomographic image, and FIG. The thickness of the nerve fiber layer 1002 shown in FIG.

  On the other hand, the image processing apparatus 110 performs analysis processing for examination such as age-related macular degeneration when the operator instructs “processing when the retinal pigment epithelium layer is abnormal”. In this case, for example, as shown in FIG. 10D, the image processing apparatus 110 detects the inner boundary film 1003 and the retinal pigment epithelium layer boundary 1005 from the tomographic image, and further, the normal structure 1006 of the retinal pigment epithelium layer boundary. Is estimated. The image processing apparatus 110 measures the thickness of the entire retinal layer and quantifies the state of the retinal pigment epithelium layer 1001 (specifically, the disturbance of the retinal pigment epithelium layer 1001).

  The tomographic image capturing apparatus 120 is an apparatus that captures a tomographic image of an eye part or the like as a subject, for example, and includes, for example, a time domain type OCT or a Fourier domain type OCT. The tomographic image capturing apparatus 120 three-dimensionally captures a tomographic image of an eye (macular region) of a subject (patient) (not shown) according to an operation by an operator. Then, the tomographic image capturing apparatus 120 transmits the obtained tomographic image to the image processing apparatus 110.

  The data server 130 is a server that holds tomographic image data of a subject's eye or the like and analysis result data (layer boundary data, quantified numerical data, and the like). The data server 130 stores tomographic image data of the eye part and the like and the analysis result data output from the image processing apparatus 110 (or tomographic image capturing apparatus 120). Further, the data server 130 transmits past data related to the subject (eye to be examined) to the image processing apparatus 110 in response to a request from the image processing apparatus 110.

Next, a hardware configuration of the image processing apparatus 110 illustrated in FIG. 1 will be described.
FIG. 2 is a schematic diagram illustrating an example of a hardware configuration of the image processing apparatus 110 according to the first embodiment of the present invention.

  As shown in FIG. 2, the image processing apparatus 110 has hardware configurations of a CPU 111, a RAM 112, a ROM 113, an external storage device 114, a monitor 115, a keyboard 116, a mouse 117, a communication interface 118, and a bus 119. It is configured.

  The CPU 111 controls the image processing apparatus 110 as a whole using programs and data stored in the ROM 113 or the external storage device 114.

  The RAM 112 includes an area for temporarily storing programs and data loaded from the external storage device 114 (or the ROM 113), and a work area required for the CPU 111 to perform various processes.

  The ROM 113 generally stores a computer BIOS and setting data.

  The external storage device 114 is a device that functions as a large-capacity information storage device such as a hard disk drive, and stores, for example, an operating system, a program executed by the CPU 111, and the like. Also, various types of information and data known in the description of the present embodiment are stored in the external storage device 114 and loaded into the RAM 112 as necessary. In this example, the program executed by the CPU 111 is stored in the external storage device 114, but may be stored in the ROM 113, for example.

  The monitor 115 is composed of, for example, a liquid crystal display.

  The keyboard 116 and the mouse 117 constitute input devices, and the operator can give various instructions to the image processing apparatus 110 using these input devices.

  The communication interface 118 is for the image processing apparatus 110 to communicate various data with an external apparatus, and is configured by, for example, IEEE 1394, USB, Ethernet (registered trademark) port, or the like. Data acquired via the communication interface 118 is taken into the external storage device 114 and then loaded into the RAM 112 as necessary.

  The bus 119 connects the components (111 to 118) inside the image processing apparatus 110 so that they can communicate with each other.

Next, a functional configuration (software configuration) of the image processing apparatus 110 will be described.
FIG. 3 is a schematic diagram illustrating an example of a functional configuration of the image processing apparatus 110 according to the first embodiment of the present invention.

  As illustrated in FIG. 3, the image processing apparatus 110 includes a subject information acquisition unit 210, an image acquisition unit 220, an instruction acquisition unit 230, a storage unit 240, an image processing unit 250, a display unit 260, and a result output unit 270. It has a functional configuration.

  Here, in the present embodiment, for example, from the programs stored in the CPU 111 and the external storage device 114 shown in FIG. 2 and the input devices of the keyboard 116 and the mouse 117, the subject information acquisition unit 210 shown in FIG. An instruction acquisition unit 230 is configured. For example, the image acquisition unit 220 and the result output unit 270 illustrated in FIG. 3 are configured from the programs stored in the CPU 111 and the external storage device 114 illustrated in FIG. 2 and the communication interface 118. Further, for example, the RAM 112, the ROM 113, or the external storage device 114 shown in FIG. 2 constitutes the storage unit 240 shown in FIG. Further, for example, the image processing unit 250 shown in FIG. 3 is configured from the programs stored in the CPU 111 and the external storage device 114 shown in FIG. In addition, for example, the display unit 260 includes the program stored in the CPU 111 and the external storage device 114 shown in FIG.

  The subject information acquisition unit 210 acquires information for identifying a subject (eye to be examined) by an operator's input via the keyboard 116 and the input device of the mouse 117, for example. Here, the information for identifying a subject (eye to be examined) is, for example, an identification number assigned to each subject (eye to be examined). In addition to this, as the information for identifying the subject (eye to be examined), a subject identification number and an identifier indicating whether the examination target is the right eye or the left eye may be used in combination. Furthermore, the subject information acquisition unit 210 performs processing for acquiring information on the subject (eye to be examined) held by the data server 130 based on information for identifying the subject (eye to be examined).

  In this example, information for identifying a subject (eye to be examined) is acquired by an operator via an input device. However, the present invention is not limited to this form. For example, when the tomographic image capturing apparatus 120 holds information for identifying a subject (eye to be examined), the information may be acquired from the tomographic image capturing apparatus 120 together with the tomographic image.

  The image acquisition unit 220 requests the tomographic image capturing apparatus 120 to transmit a predetermined tomographic image, and performs processing for acquiring the tomographic image transmitted from the tomographic image capturing apparatus 120. In the following description, it is assumed that the tomographic image acquired by the image acquisition unit 220 is that of a subject (eye to be examined) identified by the subject information acquisition unit 210. In addition, it is assumed that the tomographic image acquired by the image acquisition unit 220 includes various parameters related to the imaging of the tomographic image as information.

  The instruction acquisition unit 230 acquires an instruction for processing that the operator inputs through the input device. For example, the instruction acquisition unit 230 acquires an instruction as to which of the above-described analysis processing (“normal processing of the retinal pigment epithelium layer” or “abnormal processing of the retinal pigment epithelium layer”) is to be executed. Also, the instruction acquisition unit 230 acquires an instruction to interrupt or end the analysis process, whether to save the analysis result, an instruction for a storage location, or the like. Then, the instruction acquisition unit 230 transmits the content of the acquired instruction to the image processing unit 250 and the result output unit 270 as necessary.

  The storage unit 240 temporarily stores and holds information about the subject (eye to be examined) acquired by the subject information acquisition unit 210. The storage unit 240 temporarily stores and holds the tomographic image of the subject (eye to be examined) acquired by the image acquisition unit 220. Further, the storage unit 240 temporarily stores and holds a tomographic image analysis result by the image processing unit 250 described later. These data stored in the storage unit 240 are transmitted to the image processing unit 250, the display unit 260, and the result output unit 270 as necessary.

  The image processing unit 250 inputs a tomographic image from the storage unit 240, and executes analysis processing on the tomographic image based on a predetermined disease in accordance with the instruction acquired by the instruction acquisition unit 230. Here, as shown in FIG. 3, the image processing unit 250 includes a layer detection unit 251, a normal structure estimation unit 252, and a quantification unit 253.

  The layer detection unit 251 determines a predetermined layer (specifically, in this example, the inner boundary film 1003 and the retinal pigment epithelial layer boundary shown in FIG. 10) constituting the subject (eye to be examined) from the tomographic image acquired from the storage unit 240. 1005) is detected. Further, when “normal processing of the retinal pigment epithelium layer” is instructed, the layer detection unit 251 further detects the nerve fiber layer boundary 1004 shown in FIG. Details of specific processing executed by the layer detection unit 251 will be described later.

  When the “retinal pigment epithelium layer abnormality processing” is instructed, the normal structure estimation unit 252 has a normal structure of a predetermined layer detected by the layer detection unit 251 (in this example, the retinal pigment epithelium layer (boundary ) Is estimated (1006 in FIG. 10 (d)) On the other hand, the normal structure estimation unit 252 is instructed to execute “normal processing of the retinal pigment epithelium layer”. The processing is not executed, and details of specific processing executed by the normal structure estimation unit 252 will be described later.

  The quantification unit 253 quantifies the state of the subject (eye to be examined) based on the detection result of the predetermined layer by the layer detection unit 251 and the normal structure of the predetermined layer estimated by the normal structure estimation unit 252. I do.

  For example, the quantification unit 253 quantifies the thickness of the entire retinal layer. Further, when “normal processing of the retinal pigment epithelium layer” is instructed, the quantification unit 253 quantifies the thickness of the nerve fiber layer 1002 illustrated in FIG. In addition, when “the abnormal processing of the retinal pigment epithelium layer” is instructed, the quantification unit 253 determines the difference between the detected retinal pigment epithelium layer boundary and its normal structure (that is, the disturbance of the retinal pigment epithelium layer). ). Details of specific processing executed by the quantification unit 253 will be described later.

  The display unit 260 superimposes information on the layer boundaries and the normal structure obtained by the image processing unit 250 on the tomographic image and displays the information on the monitor 115. The display unit 260 displays various numerical data quantified by the quantification unit 253 on the monitor 115.

  The result output unit 270 associates the examination date and time, information for identifying the subject (eye to be examined), the tomographic image of the subject (eye to be examined), and the analysis result obtained by the image processing unit 250, and information to be stored To the data server 130.

  Next, a specific processing procedure executed by the image processing apparatus 110 according to the present embodiment will be described with reference to FIG.

  FIG. 4 is a flowchart illustrating an example of a processing procedure of the control method of the image processing apparatus 110 according to the first embodiment of the present invention. Note that each functional configuration shown in FIG. 3 of the image processing apparatus 110 according to the present embodiment is realized by the CPU 111 executing a program that realizes each functional configuration and controlling the entire computer. Further, it is assumed that the program code according to the flowchart shown in FIG. 4 is already loaded from, for example, the external storage device 114 to the RAM 112 before the following processing is performed.

<Step S101>
First, in step S101, the subject information acquisition unit 210 acquires information for identifying a subject (eye to be examined) from the outside. Specifically, the subject information acquisition unit 210 acquires information for identifying a subject (eye to be examined) by an operator's input via the keyboard 116 and the input device of the mouse 117, for example. Then, the subject information acquisition unit 210 performs processing for acquiring information on the subject (eye to be examined) held by the data server 130 based on the information for identifying the acquired subject (eye to be examined). For example, the subject information acquisition unit 210 acquires a tomographic image of the eye to be inspected in the past and an analysis result thereof (such as a layer boundary and quantified numerical data). In particular, the subject information acquisition unit 210 acquires data from the data server 130 when the data server 130 holds normal data of the retinal pigment epithelium layer boundary related to the eye to be examined. Then, the subject information acquisition unit 210 transmits the information acquired in this way to the storage unit 240.

<Step S102>
Subsequently, in step S <b> 102, the image acquisition unit 220 requests the tomographic image capturing apparatus 120 to transmit a predetermined tomographic image, and performs processing for acquiring a tomographic image transmitted from the tomographic image capturing apparatus 120. Then, the image acquisition unit 220 transmits the acquired tomographic image to the storage unit 240.

<Step S103>
Subsequently, in step S103, the display unit 260 performs processing for displaying the tomographic image acquired in step S102 on the monitor 115. Here, for example, an image as schematically shown in FIGS. 10A and 10C is displayed on the monitor 115. Here, since the tomographic image is three-dimensional data, what is actually displayed on the monitor 115 is a two-dimensional tomographic image obtained by cutting out any cross section of interest. Note that it is desirable that the displayed cross section can be arbitrarily selected via a GUI or the like. Moreover, the structure which can arrange | position and display the various data (tomographic image, its analysis result, etc.) in the past time acquired by step S101 may be sufficient.

<Step S104>
Subsequently, in step S <b> 104, the instruction acquisition unit 230 acquires an instruction for processing that the operator inputs via an input device such as the keyboard 116 or the mouse 117. Specifically, here, the instruction acquisition unit 230 acquires an instruction to execute either “normal processing of the retinal pigment epithelium layer” or “abnormal processing of the retinal pigment epithelium layer”. At this time, the operator may determine which analysis process is to be executed by the image processing apparatus 110 and input the determination result to the image processing apparatus 110 (select any process). The processing instruction obtained in step S104 is transmitted to the image processing unit 250.

<Step S105>
Subsequently, in step S105, for example, the image processing unit 250 determines whether or not the processing instruction obtained in step S104 is “normal processing of the retinal pigment epithelium layer”.

  As a result of the determination in step S105, when the processing instruction obtained in step S104 is “normal processing of the retinal pigment epithelium layer”, the process proceeds to step S106. On the other hand, as a result of the determination in step S105, when the processing instruction obtained in step S104 is not “normal processing of the retinal pigment epithelial layer” (that is, “processing when retinal pigment epithelial layer is abnormal”). Advances to step S107.

<Step S106>
In step S106, the image processing unit 250 executes analysis processing related to “normal processing of the retinal pigment epithelium layer”. In addition, the display unit 260 performs processing for displaying the result of the analysis processing by the image processing unit 250. Detailed processing in step S106 will be described later with reference to the flowchart shown in FIG. Thereafter, the process proceeds to step S108.

<Step S107>
On the other hand, when the processing proceeds to step S107, the image processing unit 250 executes an analysis process related to the “retinal pigment epithelium layer abnormality process”. In addition, the display unit 260 performs processing for displaying the result of the analysis processing by the image processing unit 250. Detailed processing in step S107 will be described later with reference to the flowchart shown in FIG. Thereafter, the process proceeds to step S108.

<Step S108>
In step S108, first, the instruction acquisition unit 230 determines whether or not to store, in the data server 130, the current processing result regarding the subject (eye to be examined) input from the operator via the keyboard 116 or the mouse 117, for example. The process which acquires the instruction | indication is performed. Then, for example, the result output unit 270 determines whether or not to save the current processing result regarding the subject (eye to be examined) in the data server 130 in accordance with the instruction acquired by the instruction acquisition unit 230.

  As a result of the determination in step S108, when the current processing result regarding the subject (eye to be examined) is stored in the data server 130, the process proceeds to step S109. On the other hand, as a result of the determination in step S108, if the current processing result related to the subject (eye to be examined) is not stored in the data server 130, the process proceeds to step S110.

<Step S109>
In step S109, the result output unit 270 associates the examination date and time, information for identifying the subject (eye to be examined), the tomographic image of the subject (eye to be examined), and the analysis result obtained by the image processing unit 250. And transmitted to the data server 130 as information to be stored. Then, it progresses to step S110.

<Step S110>
In step S110, the instruction acquisition unit 230 acquires an instruction whether to end the tomographic image analysis processing by the image processing apparatus 110, which is input from the operator via the keyboard 116 or the mouse 117, for example. I do. Then, the image processing apparatus 110 determines whether to end the tomographic image analysis process according to the instruction acquired by the instruction acquisition unit 230.

  If the tomographic image analysis process is not terminated as a result of the determination in step S110, the process returns to step S101, and the process for the next subject (eye to be examined) (or reprocessing for the same eye to be examined) is executed. On the other hand, as a result of the determination in step S110, if the tomographic image analysis process is to be terminated, the process in the flowchart shown in FIG. 4 is terminated.

  Next, with reference to FIG. 5, a detailed processing procedure of the processing executed in step S106 of FIG. 4 will be described.

  FIG. 5 is a flowchart showing an example of a detailed processing procedure of “normal processing of the retinal pigment epithelium layer” in step S106 of FIG.

<Step S201>
First, in step S <b> 201, the layer detection unit 251 acquires a tomographic image of the subject (eye to be examined) from the storage unit 240 and performs a process of detecting a predetermined layer constituting the subject (eye to be examined) from the tomographic image. . Specifically, the layer detection unit 251 detects an inner boundary film 1003, a nerve fiber layer boundary 1004, and a retinal pigment epithelium layer boundary 1005 shown in FIG. 10 from a tomographic image of a subject (eye to be examined). Furthermore, the layer detection unit 251 also detects the coordinates of the fovea, which is the center of the macula. Then, the layer detection unit 251 outputs these detection results to the storage unit 240.

Hereinafter, a specific processing method for detecting a layer boundary by the layer detection unit 251 will be described.
Here, the three-dimensional tomographic image to be processed is considered as a set of two-dimensional tomographic images (B-scan images), and the following two-dimensional image processing is executed on each two-dimensional tomographic image.

  First, the layer detection unit 251 performs a smoothing filter process on the focused two-dimensional tomographic image to remove noise components. Then, the layer detection unit 251 detects edge components from the tomographic image, and extracts some line segments as layer boundary candidates based on the connectivity. Then, the layer detecting unit 251 selects the uppermost line segment as the inner boundary film 1003 from these candidates. In addition, the layer detection unit 251 selects a line segment immediately below the inner boundary film 1003 as the nerve fiber layer boundary 1004. Further, the layer detection unit 251 selects the lowermost line segment as the retinal pigment epithelium layer boundary 1005.

  Furthermore, the detection accuracy may be improved by applying a dynamic contour method such as a Snake or a level set method using these line segments as initial values. Further, a layer boundary may be detected using a technique such as graph cut.

  Note that the boundary detection method using the dynamic contour method or the graph cut may be performed three-dimensionally for a three-dimensional tomographic image, or the three-dimensional tomographic image to be processed is set as a set of two-dimensional tomographic images. It is possible to apply two-dimensionally to each two-dimensional tomographic image. The method for detecting the layer boundary is not limited to these methods, and any method may be used as long as the layer boundary can be detected from the tomographic image of the eye.

  The layer detection unit 251 further detects coordinates of the fovea using the inner boundary film 1003. Specifically, the layer detection unit 251 sets the point where the z coordinate of the detected inner boundary film 1003 is maximum near the center of the tomographic image as the fovea. In the following processing, all coordinates are converted (translated) so that the fovea is the origin.

<Step S202>
Subsequently, in step S202, the quantification unit 253 quantifies necessary information by measuring the thickness of the nerve fiber layer 1002 and the thickness of the entire retinal layer based on the layer boundary detected in step S201. Do.

  First, the quantifying unit 253 obtains the z-coordinate difference between the nerve fiber layer boundary 1004 and the inner boundary film 1003 at each coordinate point on the xy plane, thereby determining the thickness of the nerve fiber layer 1002 (FIG. 10A). Of T1). Similarly, the quantification unit 253 measures the thickness of the entire retinal layer (T2 in FIG. 10A) by obtaining a difference in z-coordinate between the retinal pigment epithelium layer boundary 1005 and the inner boundary film 1003. Furthermore, the quantification unit 253 measures the area of each layer (the nerve fiber layer 1002 and the entire retinal layer) in each cross section by adding the layer thickness at each coordinate point in the x-axis direction for each y coordinate. To do. Further, the quantification unit 253 measures the volume of each layer by adding the obtained areas in the y-axis direction. Then, the quantification unit 253 outputs data related to these measurement results to the storage unit 240.

<Step S203>
Subsequently, in step S203, the display unit 260 performs a process of superimposing the layer boundary detection result in step S101 on the tomographic image and displaying it on the monitor 115. Here, as shown in FIG. 10B, when the layer boundaries are indicated by lines, it is desirable to use lines of a predetermined color for each boundary. For example, the display unit 260 displays (presents) the inner boundary film 1003 using a red line, the nerve fiber layer boundary 1004 using a yellow line, and the retinal pigment epithelium layer boundary 1005 using a green line. Further, for example, the display unit 260 may display (present) a semi-transparent color in a layer region without explicitly showing the layer boundary. For example, the display unit 260 may display (present) a region representing the entire retinal layer other than the nerve fiber layer 1002 in green and a region representing the nerve fiber layer 1002 in red. In addition, when performing these displays, it is desirable to have a configuration in which a cross section of interest can be selected using a GUI or the like. Further, it may be displayed three-dimensionally using a known volume rendering technique.

  Further, the display unit 260 performs processing for displaying information on the layer thickness measured by quantification in step S202 on the monitor 115. This display may be displayed as a layer thickness distribution map for the entire three-dimensional tomographic image (xy plane), or the area of each layer in the cross section of interest may be displayed in conjunction with the display of the detection result described above. Further, the entire volume may be displayed, or the volume in a region designated by the operator on the xy plane may be calculated and displayed.

<Step S204>
Subsequently, in step S204, the instruction acquisition unit 230 first stores the coordinate group representing the retinal pigment epithelium layer boundary 1005 detected in step S201 as normal data of the retinal pigment epithelium layer boundary of the subject (examined eye). The process which acquires the instruction | indication of whether to do is performed. This instruction is input by the operator via the keyboard 116 or the mouse 117, for example. Then, for example, the result output unit 270 uses the coordinate group representing the retinal pigment epithelium layer boundary 1005 detected in step S201 in accordance with the instruction acquired by the instruction acquisition unit 230 as the retinal pigment epithelium layer boundary of the subject (examined eye). It is determined whether or not to save as normal data.

  As a result of the determination in step S204, when the coordinate group representing the retinal pigment epithelium layer boundary 1005 detected in step S201 is not stored as normal data of the retinal pigment epithelium layer boundary of the subject (examined eye), FIG. The process of the flowchart ends. On the other hand, if the coordinate group representing the retinal pigment epithelium layer boundary 1005 detected in step S201 is stored as normal data of the retinal pigment epithelium layer boundary of the subject (examined eye) as a result of the determination in step S204, The process proceeds to S205.

<Step S205>
In step S205, the result output unit 270 uses the retinal pigment epithelium layer boundary data (coordinate group representing the layer boundary) detected in step S201 as normal data of the retinal pigment epithelium layer boundary of the subject (eye to be examined). To the data server 130. Thereby, the data server 130 stores the data acquired from the result output unit 270 in association with the subject (eye to be examined) as normal data of the retinal pigment epithelium layer boundary. Thereafter, the process in the flowchart shown in FIG. 5 (that is, the process of S106 in FIG. 4) is completed.

  In the present embodiment, only the thicknesses of the nerve fiber layer 1002 and the entire retinal layer are measured in the process of step S106. Of course, other layers such as the photoreceptor layer of the eye and the outer boundary membrane are also measured. The structure which performs an analysis may be sufficient.

  Next, with reference to FIG. 6, the detailed process procedure of the process performed by step S107 of FIG. 4 is demonstrated.

  FIG. 6 is a flowchart showing an example of a detailed processing procedure of the “retinal pigment epithelium layer abnormality process” in step S107 of FIG.

<Step S301>
First, in step S301, the layer detection unit 251 acquires a tomographic image of the subject (eye to be examined) from the storage unit 240, and performs a process of detecting a predetermined layer constituting the subject (eye to be examined) from the tomographic image. . Specifically, the layer detection unit 251 detects the inner boundary film 1003 and the retinal pigment epithelium layer boundary 1005 shown in FIG. 10 from the tomographic image of the subject (eye to be examined). Furthermore, the layer detection unit 251 also detects the coordinates of the fovea, which is the center of the macula. Then, the layer detection unit 251 outputs these detection results to the storage unit 240. Here, since the specific method of the layer boundary detection by the layer detection unit 251 is the same as that in step S201, detailed description thereof is omitted. In the following processing, all coordinates are converted (translated) so that the fovea is the origin.

<Step S302>
Subsequently, in step S302, the normal structure estimation unit 252 normalizes the retinal pigment epithelium layer boundary of the subject (eye to be examined) based on the information regarding the subject (eye to be examined) acquired by the subject information acquisition unit 210 in step S101. Determine whether hourly data is available. More specifically, whether or not the information on the subject (eye to be examined) acquired by the subject information acquisition unit 210 in step S101 includes normal data of the retinal pigment epithelium layer boundary of the subject (eye to be examined). Based on the above, it is determined whether or not the normal data is available.

  As a result of the determination in step S302, when normal data of the retinal pigment epithelium layer boundary of the subject (eye to be examined) is available, after the normal structure estimation unit 252 acquires the normal data from the storage unit 240, Proceed to step S303. On the other hand, as a result of the determination in step S302, when normal data of the retinal pigment epithelium layer boundary of the subject (eye to be examined) is not available, the display unit 260 displays that the normal data cannot be used, Proceed to step S304.

<Step S303>
In step S303, the normal structure estimation unit 252 performs processing for applying the normal data of the retinal pigment epithelial layer boundary acquired in step S302 to the tomographic image, and determines the normal structure of the retinal pigment epithelial layer boundary on the tomographic image. Performs estimation processing. Then, when the process of step S303 ends, the process proceeds to step S305.

  Hereinafter, a specific processing method for estimating the normal structure of the retinal pigment epithelium layer boundary in step S303 will be described.

  The normal data is expressed in a coordinate system with the fovea as the origin on the tomographic image when the data is acquired. On the other hand, the tomographic image (current tomographic image) acquired in step S102 is also expressed in a coordinate system (current coordinate system) with the fovea as the origin. Therefore, the normal structure estimation unit 252 superimposes the normal data on the current coordinate system so that these coordinate systems match. Then, the normal structure estimation unit 252 sets the normal data superimposed in this way as the initial value of the normal structure of the retinal pigment epithelium layer boundary on the current tomographic image (that is, performs initial alignment). Then, the normal structure estimation unit 252 updates the normal structure by the fitting process to the retinal pigment epithelium layer boundary 1005 detected on the current tomographic image. In other words, the normal structure estimation unit 252 defines the initial value of the conversion matrix A for converting normal data into a normal structure as a unit matrix, and estimates the conversion matrix A so that the normal structure is appropriate. . In this embodiment, the transformation matrix A is defined as a matrix representing a three-dimensional rigid transformation, but it can also be a matrix including a uniform scale deformation, or can be a matrix that allows affine transformation in general. is there.

In the present embodiment, in order to perform the above-described fitting process, the difference ε _i between the detection result and the normal structure is defined as the following equation (1).

Xorg_i in the expression (1) represents the coordinates of the i-th point of normal data described as a coordinate group. In addition, Xnormal_i in the expression (1) represents the coordinates of a point after coordinate transformation by the transformation matrix A (that is, a point on the normal structure). Further, Xdetected_i in the equation (1) represents the coordinates of a point on the retinal pigment epithelium layer boundary 1005 corresponding to each Xnormal_i. This association is performed, for example, by selecting a point on the retinal pigment epithelium layer boundary 1005 having the same x, y coordinates for each Xnormal_i. Alternatively, a point on the retinal pigment epithelium layer boundary 1005 closest to each Xnormal_i is selected. In addition, each coordinate and coordinate transformation matrix in equation (1) are assumed to be expressed in a homogeneous coordinate system.

  Then, the normal structure estimation unit 252 obtains a transformation matrix A that minimizes the evaluation formula shown in the following formula (2).

In equation (2), ρ (x) is a weight function.
FIG. 7 is a schematic diagram illustrating an example of a weight function used in estimating a normal structure of a layer included in a tomographic image according to the first embodiment of this invention.

7A, 7B, and 7C, the horizontal axis is x (0 near the center), and the vertical axis is the weighting function ρ (x). Note that the weighting function used when estimating the normal structure of the layer included in the tomographic image is not all shown in FIG. 7, and any function may be set.
Further, minimizing the expression (2) corresponds to obtaining the transformation matrix A using the least square method or M estimation.

<Step S304>
On the other hand, when the process proceeds to step S304, the normal structure estimation unit 252 performs a process of applying a curve described by a polynomial to the retinal pigment epithelium layer boundary 1005 for each two-dimensional cross section of the tomographic image, and the normal structure thereof. The process which estimates is performed. The estimation processing method is the same as the method described in the background art. Thereafter, the process proceeds to step S305.

<Step S305>
In step S305, the quantification unit 253 measures the thickness of the entire retinal layer based on the layer boundary detected in step S301, and quantifies it. Further, the quantification unit 253 determines the disturbance of the retinal pigment epithelium layer 1001 based on the difference between the retinal pigment epithelium layer boundary 1005 detected in step S301 and the normal structure 1006 estimated in step S303 or step S304. Perform the process of quantification. Then, the quantification unit 253 outputs these measurement results to the storage unit 240.

  Note that the quantification of the thickness of the entire retinal layer is the same as the processing in step S202, and thus detailed description thereof is omitted. Further, regarding the quantification of the disturbance of the retinal pigment epithelium layer 1001, the quantification of the thickness, area, and volume of the difference is the same as the quantification of the layer thickness, and detailed description thereof is omitted.

  The feature amount representing the disturbance of the retinal pigment epithelium layer 1001 is any one that quantifies the difference between the retinal pigment epithelium layer boundary 1005 detected in step S301 and its normal structure 1006. Also good. For example, the distribution of the obtained difference (thickness), the maximum value, the average value, the median value, the variance, the standard deviation, the number of points greater than or equal to the threshold, the ratio thereof, and the like may be obtained as the characteristics of the regions that are made different. Further, density features such as a density histogram, an average density value, a density variance value, and contrast may be obtained as the characteristics of the different areas.

<Step S306>
Subsequently, in step S306, the display unit 260 superimposes and displays the retinal pigment epithelium layer boundary 1005 detected in step S301 and the normal structure 1006 estimated in step S303 or step S304 on the tomographic image. I do. In accordance with this, the display unit 260 performs a process of displaying the data measured by quantification in step S305. Since these display processes are the same as the processes in step S203, detailed description thereof is omitted.

  When the process of step S306 ends, the process in the flowchart of FIG. 6 (that is, the process of S107 of FIG. 4) ends.

  According to the first embodiment described above, since the process of fitting the retinal pigment epithelium layer boundary when the subject (test eye) itself is normal is performed, the retinal pigment epithelium layer constituting the subject on the tomographic image is processed. The normal structure can be estimated with high accuracy. Therefore, the state (disturbance of shape) of the retinal pigment epithelium layer can be accurately quantified. This makes it possible to quantitatively analyze, for example, the degree of progression of diseases such as age-related macular degeneration and the degree of recovery after treatment. Further, even when a part of the predetermined layer boundary constituting the subject is difficult to observe on the tomographic image, the predetermined layer can be detected with high accuracy.

(Second Embodiment)
Next, a second embodiment of the present invention will be described.
The image processing apparatus 110 according to the first embodiment acquires a tomographic image of a subject (eye to be examined) imaged by the tomographic image imaging apparatus 120, and performs an analysis process based on a predetermined disease according to an operator's instruction. To quantitate the disease. In contrast, the image processing apparatus according to the second embodiment is the same as the first embodiment in that analysis processing is performed on the acquired tomographic image of the subject (eye to be examined). However, in the image processing apparatus according to the second embodiment, an instruction from the operator determines whether to perform “normal processing of the retinal pigment epithelium layer” or “abnormal processing of the retinal pigment epithelium layer”. This is different from the first embodiment in that the determination is automatically performed inside the apparatus.

  The schematic configuration of the image processing system according to the second embodiment is the same as the schematic configuration of the image processing system 100 according to the first embodiment shown in FIG. The hardware configuration of the image processing apparatus according to the second embodiment is the same as the hardware configuration of the image processing apparatus 110 according to the first embodiment shown in FIG.

  FIG. 8 is a schematic diagram illustrating an example of a functional configuration of the image processing apparatus 110 according to the second embodiment of the present invention. Here, in FIG. 8, the same components as those in FIG. 3 are denoted by the same reference numerals. As shown in FIG. 8, the image processing apparatus 110 according to the second embodiment is configured by adding a state determination unit 254 to the inside of the image processing unit 250, and thus the first implementation shown in FIG. 3. This is different from the functional configuration of the image processing apparatus 110 according to the embodiment. In the following description, differences from the first embodiment will be described.

  The state determination unit 254 determines whether the retinal pigment epithelium layer is normal or abnormal based on the retinal pigment epithelium layer boundary 1005 detected by the layer detection unit 251. Details of specific processing executed by the state determination unit 254 will be described later.

  Next, a specific processing procedure executed by the image processing apparatus 110 according to the present embodiment will be described with reference to FIG.

  FIG. 9 is a flowchart illustrating an example of a processing procedure of a control method of the image processing apparatus 110 according to the second embodiment of the present invention. In FIG. 9, processes similar to those in FIG. 4 are denoted by the same step numbers and are the same as those in the first embodiment, and thus detailed description thereof is omitted.

  First, in the processing of the image processing apparatus 110 according to the second embodiment, the processing of steps S101 to S103 shown in FIG. 4 is performed.

<Step S401>
Subsequently, in step S401, the layer detection unit 251 obtains a tomographic image of the subject (eye to be examined) from the storage unit 240, and detects a predetermined layer constituting the subject (eye to be examined) from the tomographic image. Do. Specifically, the layer detection unit 251 detects an inner boundary film 1003, a nerve fiber layer boundary 1004, and a retinal pigment epithelium layer boundary 1005 shown in FIG. 10 from a tomographic image of a subject (eye to be examined). Then, the layer detection unit 251 outputs these detection results to the storage unit 240. Since this step S401 is the same as step S201 of FIG. 5 in the first embodiment, detailed description thereof is omitted.

<Step S402>
Subsequently, in step S402, the state determination unit 254 determines whether or not the retinal pigment epithelium layer is normal based on the retinal pigment epithelium layer boundary 1005 detected in step S401. For example, the state determination unit 254 determines that the retinal pigment epithelium layer is abnormal when the maximum value of the curvature of the retinal pigment epithelium layer boundary 1005 detected in step S401 is equal to or greater than a predetermined threshold. Then, the state determination unit 254 outputs the determination result to the storage unit 240.

  Note that the determination of whether or not the retinal pigment epithelium layer is normal in step S402 may be performed by any other method. For example, the determination as to whether or not the retinal pigment epithelium layer is normal in step S402 may be performed based on whether or not the number of maximum points and minimum points of the retinal pigment epithelium layer boundary 1005 is equal to or greater than a threshold value. Further, it may be determined by the number of inflection points, or may be determined by the dispersion of curvature. Further, for example, in the determination of whether or not the retinal pigment epithelial layer is normal in step S402, a principal component analysis is performed on the three-dimensional coordinate group representing the retinal pigment epithelial layer boundary 1005, and a third representing the variation in the z-axis direction. You may determine by the magnitude | size of a main component being more than a threshold value. Furthermore, frequency analysis by Fourier transform may be performed, and determination may be made based on a predetermined frequency region intensity. Note that these processes may be performed only on the retinal pigment epithelium layer boundary 1005 in the vicinity of the fovea where the abnormality is likely to occur.

<Step S403>
Subsequently, in step S403, for example, the image processing unit 250 determines whether or not to perform “normal processing of the retinal pigment epithelium layer” based on the determination result obtained in the processing of step S402.

  As a result of the determination in step S403, when “normal processing of the retinal pigment epithelium layer” is performed (that is, when the determination result obtained in the processing of step S402 is that the retinal pigment epithelium layer is normal), step S404 is performed. Proceed to On the other hand, as a result of the determination in step S403, when “normal processing of the retinal pigment epithelium layer” is not performed (that is, the determination result obtained in the processing of step S402 indicates that the retinal pigment epithelium layer is not normal (abnormal). In the case), the process proceeds to step S405.

<Step S404>
In step S404, the image processing unit 250 executes analysis processing related to “normal processing of the retinal pigment epithelium layer”. The detailed process of step S404 is basically the same as the detailed process (FIG. 5) of step S106 in the first embodiment. However, in this embodiment, the layer boundary detection process in step S201 of FIG. 5 is not executed.

<Step S405>
On the other hand, when the processing proceeds to step S405, the image processing unit 250 executes an analysis process related to the “retinal pigment epithelium layer abnormality process”. The detailed process of step S405 is basically the same as the detailed process (FIG. 6) of step S107 in the first embodiment. However, in this embodiment, the layer boundary detection process in step S301 in FIG. 6 is not executed.

  Then, when the process of step S404 or step S405 ends, the process proceeds to step S108, and the processes after step S108 of FIG. 4 in the first embodiment are performed.

  According to the second embodiment described above, since the image processing device 110 determines whether or not the retinal pigment epithelium layer of the subject (eye to be examined) is normal, in addition to the effects in the first embodiment, If it is normal, the data can be automatically saved as normal data. As a result, when the retinal pigment epithelium layer is abnormal, it is possible to estimate the normal structure based on the stored normal data and quantitatively evaluate the state (shape disturbance).

(Third embodiment)
Next, a third embodiment of the present invention will be described.
The third embodiment is a form according to a modified example of the first and second embodiments described above. Below, each modification is demonstrated.

-Modification 1-
In each of the above-described embodiments, the normal structure of a predetermined layer is estimated using the least square method or M estimation in the process of step S303 in FIG. However, the method for estimating the normal structure is not limited to this. For example, the number of points (inliers) at which ε _{i in} Equation (1) is equal to or less than a predetermined threshold may be used as an evaluation criterion, and a transformation matrix that maximizes the number may be used. In the process of step S303 in FIG. 6, the normal structure is estimated using the detected retinal pigment epithelium layer boundary 1005 and all the normal data points, but the points to be used are thinned out at appropriate intervals. Also good. Any known technique for performing a fitting process between three-dimensional point groups may be used.

  In addition, the normal structure estimation unit 252 classifies the retinal pigment epithelium layer boundary 1005 detected on the current tomographic image into a normal range and an abnormal range, and sets the retinal pigment epithelium layer boundary 1005 in the normal range excluding the abnormal range. Only the above-described fitting process may be performed. At this time, the detection of the abnormal range in the normal structure estimation unit 252 is performed based on, for example, the local curvature of the retinal pigment epithelium layer boundary 1005. In other words, the normal structure estimation unit 252 calculates the local curvature of the retinal pigment epithelium layer boundary 1005, determines that the portion where the curvature change rate is greater than or equal to the threshold and the portion sandwiched between them is an abnormal range. Do.

  Note that the coordinate transformation applied to the normal data is not limited to the uniform transformation for the entire shape as expressed by the equation (1), and transformation with local shape deformation may be allowed. . For example, a penalty term is set for a shape different from normal data (in other words, a shape different from normal data is allowed within a certain range), and then freeform deformation is applied. It is good also as a structure which performs a fitting process.

-Modification 2-
In each of the embodiments described above, normal data of the retinal pigment epithelium layer boundary is represented by a set of three-dimensional coordinates representing the boundary. However, the normal data description method is not limited to this. For example, the z-coordinates of boundaries at several sampling points on the xy plane may be held as normal data. Further, the set of three-dimensional coordinates may be approximated by a function representing a surface, and the parameters of the function may be held as normal data. In this case, before performing the fitting process in step S303 in FIG. 6, a set of three-dimensional coordinates representing normal data may be generated by substituting the coordinates of the xy plane into the function. Or you may perform the said fitting process using any well-known method of estimating the coordinate transformation for fitting a point cloud to a function.

-Modification 3-
In each of the above-described embodiments, in step S204 in FIG. 5, an instruction as to whether or not to save the detected retinal pigment epithelium layer boundary 1005 as normal data is acquired from the outside. Then, according to the instruction, the detection result is stored as normal data of the subject (eye to be examined). However, the method for determining whether or not to save the detection result as normal data is not limited to this. For example, when normal data of the subject (eye to be examined) is not registered in the data server 130, the detection result may be stored as normal data without depending on an instruction. Further, the latest detection result may be always stored as normal data regardless of whether normal data of the subject (eye to be examined) is registered in the data server 130 or not.

  Alternatively, a plurality of normal time data registered in the past may be held together with the shooting date and time, and normal time data at the date and time of use may be generated from the information. Further, for example, the normal time data may be generated by performing three-dimensional alignment on the normal time data and performing extrapolation processing based on the imaging date and time and the use date and time.

-Modification 4-
In each of the above-described embodiments, the initial alignment between the tomographic image and the normal data is performed using the fovea of the macular portion. However, initial alignment may be performed using other methods. For example, a rough alignment may be performed by obtaining the respective centroids of the detected retinal pigment epithelium layer boundary 1005 and the normal data so that the centroids coincide with each other.

-Modification 5-
In each of the above-described embodiments, analysis processing is performed on the entire acquired three-dimensional tomographic image. However, a configuration in which a notable cross section is selected from a three-dimensional tomographic image and processing is performed on the selected two-dimensional tomographic image may be employed. For example, the process may be performed on a cross section including a predetermined specific part (for example, fovea) of the fundus. In this case, the detected layer boundary, normal structure, normal data, and the like are all two-dimensional data on the cross section. And what is necessary is just to estimate the 3 * 3 matrix showing the two-dimensional rigid body transformation in the said cross section as the transformation matrix A in the process of step S303 of FIG.

-Modification 6
In each of the above-described embodiments, the disorder of the shape of the retinal pigment epithelium layer is quantified by detecting the boundary between the retinal pigment epithelium layer 1001 and its lower layer and estimating its normal structure. However, the configuration may be such that the boundary between the retinal pigment epithelium layer 1001 and the upper layer thereof is detected and the normal structure is estimated. Further, the entire retinal pigment epithelium layer 1001 may be the target of processing. Further, the layer for quantifying the disorder of shape is not limited to the retinal pigment epithelium layer 1001, and a peripheral layer that behaves similarly to a disease may be the target of processing. That is, as long as it is a layer whose shape changes to unevenness due to a disease such as age-related macular degeneration, any layer or its boundary may be detected to estimate its normal structure and quantify the shape disturbance.

-Modification 7-
In each of the embodiments described above, when the retinal pigment epithelium layer boundary 1005 has been deformed due to a disease, in order to estimate the normal structure, the normal time measured in the past with respect to the subject (eye to be examined) Data is used. However, if there is a portion where the retinal pigment epithelium layer boundary 1005 cannot be clearly observed on a tomographic image (image characteristics cannot be obtained) due to the influence of blood vessels, vitiligo, etc., the subject (eye to be examined) for the purpose of assisting the detection It is also possible to use normal data.

  For example, consider a case where there is a missing portion in the retinal pigment epithelium layer boundary 1005 detected by the layer detection unit 251. At this time, the layer boundary of the missing portion can be estimated by performing normal data fitting processing (step S303 in FIG. 6) on the retinal pigment epithelium layer boundary 1005 having the missing portion.

  Alternatively, the detection process itself of the retinal pigment epithelium layer boundary 1005 by the layer detection unit 251 may be performed by a normal time data fitting process for a tomographic image. Further, the detection process of the retinal pigment epithelium layer boundary 1005 may be performed in consideration of the constraint that the shape of the retinal pigment epithelium layer boundary 1005 is similar to the normal data. For example, the detection by the dynamic contour method may be performed in a form in which a penalty term for the shape of the detected retinal pigment epithelium layer boundary 1005 being far from normal data is added.

-Modification 8-
In each of the above-described embodiments, the tomographic imaging apparatus 120 does not necessarily have to be connected to the image processing apparatus 110. For example, a configuration may be adopted in which a tomographic image to be processed is already captured in the data server 130 and is read and processed. In this case, the image acquisition unit 220 requests the data server 130 to transmit a tomographic image of the subject (eye to be examined), and acquires the tomographic image transmitted from the data server 130. Then, the layer detection unit 251 performs layer boundary detection, and the quantification unit 253 performs quantification processing.

  On the other hand, the data server 130 does not necessarily have to be connected to the image processing apparatus 110, and the external storage device 114 of the image processing apparatus 110 may play a role.

(Other embodiments)
3 and FIG. 8 constituting the image processing apparatus 110 according to each embodiment of the present invention described above, and each step of FIG. 4 to FIG. 6 and FIG. Can be realized by executing a program stored in the external storage device (114). This program and a computer-readable recording medium storing the program are included in the present invention.

  In addition, the present invention can be implemented as, for example, a system, apparatus, method, program, storage medium, or the like. Specifically, the present invention may be applied to a system including a plurality of devices. You may apply to the apparatus which consists of one apparatus.

  In the present invention, a software program for realizing the functions of the above-described embodiments (in the embodiment, a program corresponding to the flowcharts shown in FIGS. 4 to 6 and 9) is directly or remotely transmitted to a system or apparatus. Includes supplies. The present invention also includes a case where the system or the computer of the apparatus is achieved by reading and executing the supplied program code.

  Accordingly, since the functions of the present invention are implemented by computer, the program code installed in the computer also implements the present invention. In other words, the present invention includes a computer program itself for realizing the functional processing of the present invention.

  In that case, as long as it has the function of a program, it may be in the form of object code, a program executed by an interpreter, script data supplied to the OS, and the like.

  Examples of the recording medium for supplying the program include a flexible disk, hard disk, optical disk, magneto-optical disk, MO, CD-ROM, CD-R, and CD-RW. In addition, there are magnetic tape, nonvolatile memory card, ROM, DVD (DVD-ROM, DVD-R), and the like.

  As another program supply method, a browser on a client computer is used to connect to an Internet home page. The computer program itself of the present invention or a compressed file including an automatic installation function can be downloaded from the homepage by downloading it to a recording medium such as a hard disk.

  It can also be realized by dividing the program code constituting the program of the present invention into a plurality of files and downloading each file from a different homepage. That is, the present invention includes a WWW server that allows a plurality of users to download a program file for realizing the functional processing of the present invention on a computer.

  In addition, the program of the present invention is encrypted, stored in a storage medium such as a CD-ROM, distributed to users, and key information for decryption is downloaded from a homepage via the Internet to users who have cleared predetermined conditions. Let It is also possible to execute the encrypted program by using the downloaded key information and install the program on a computer.

  Further, the functions of the above-described embodiments are realized by the computer executing the read program. In addition, based on the instructions of the program, an OS or the like running on the computer performs part or all of the actual processing, and the functions of the above-described embodiments can also be realized by the processing.

  Further, the program read from the recording medium is written in a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer. Thereafter, the CPU of the function expansion board or function expansion unit performs part or all of the actual processing based on the instructions of the program, and the functions of the above-described embodiments are realized by the processing.

  Note that each of the above-described embodiments is merely a specific example for carrying out the present invention, and the technical scope of the present invention should not be construed as being limited thereto. . That is, the present invention can be implemented in various forms without departing from the technical idea or the main features thereof.

110 Image processing apparatus 110
210 Subject information acquisition unit 220 Image acquisition unit 230 Instruction acquisition unit 240 Storage unit 250 Image processing unit 251 Layer detection unit 252 Normal structure estimation unit 253 Quantification unit 260 Display unit 270 Result output unit

Claims (14)

Image acquisition means for acquiring a tomographic image of the subject;
Subject information acquisition means for acquiring normal time data of a predetermined layer constituting the subject;
Layer detecting means for detecting the predetermined layer from the tomographic image;
Normal structure estimation means for estimating the normal structure of the predetermined layer using the detection result of the predetermined layer by the layer detection means and normal data of the predetermined layer acquired by the subject information acquisition means And an image processing apparatus. Based on the detection result of the predetermined layer by the layer detection unit, further includes a state determination unit that determines whether or not the state of the predetermined layer is normal,
The image according to claim 1, wherein the normal structure estimation unit estimates a normal structure of the predetermined layer when the state determination unit determines that the state of the predetermined layer is not normal. Processing equipment.   When the state determining unit determines that the state of the predetermined layer is normal, the apparatus further includes an output unit that outputs the detection result of the predetermined layer by the layer detecting unit as normal data of the subject. The image processing apparatus according to claim 2. Further comprising instruction acquisition means for acquiring an instruction as to whether or not to store the detection result of the predetermined layer by the layer detection means as normal data of the subject;
The output unit outputs the detection result of the predetermined layer by the layer detection unit as normal data of the subject when the instruction acquisition unit acquires the instruction to save. An image processing apparatus according to 1.   Quantifying means for quantifying the state of the predetermined layer based on the difference between the detection result of the predetermined layer by the layer detecting means and the normal structure of the predetermined layer estimated by the normal structure estimating means The image processing apparatus according to claim 1, further comprising:   The image processing apparatus according to claim 5, wherein the quantification unit quantifies the state of the predetermined layer based on any one of thickness distribution, area, volume, and concentration characteristics based on the difference. . Image acquisition means for acquiring a tomographic image of the subject;
Subject information acquisition means for acquiring normal time data of a predetermined layer constituting the subject;
An image processing apparatus comprising: layer detection means for detecting the predetermined layer from the tomographic image using normal data of the predetermined layer acquired by the subject information acquisition means.   The image processing apparatus according to claim 1, wherein the subject is an eye to be examined.   The predetermined layer is any one of a retinal pigment epithelium layer, a boundary between the retinal pigment epithelium layer and a lower layer thereof, and a boundary between the retinal pigment epithelium layer and an upper layer thereof. The image processing apparatus according to any one of 1 to 8. A control method for an image processing apparatus for processing a tomographic image of a subject,
An image acquisition step of acquiring a tomographic image of the subject;
Subject information obtaining step for obtaining normal data of a predetermined layer constituting the subject;
A layer detection step of detecting the predetermined layer from the tomographic image;
Normal structure estimation step for estimating the normal structure of the predetermined layer using the detection result of the predetermined layer by the layer detection step and normal data of the predetermined layer acquired by the subject information acquisition step And a control method for an image processing apparatus. A control method for an image processing apparatus for processing a tomographic image of a subject,
An image acquisition step of acquiring a tomographic image of the subject;
Subject information obtaining step for obtaining normal data of a predetermined layer constituting the subject;
And a layer detection step of detecting the predetermined layer from the tomographic image using normal data of the predetermined layer acquired in the subject information acquisition step. . A program for causing a computer to execute a control method of an image processing apparatus for processing a tomographic image of a subject,
An image acquisition step of acquiring a tomographic image of the subject;
Subject information obtaining step for obtaining normal data of a predetermined layer constituting the subject;
A layer detection step of detecting the predetermined layer from the tomographic image;
Normal structure estimation step for estimating the normal structure of the predetermined layer using the detection result of the predetermined layer by the layer detection step and normal data of the predetermined layer acquired by the subject information acquisition step A program that causes a computer to execute and. A program for causing a computer to execute a control method of an image processing apparatus for processing a tomographic image of a subject,
An image acquisition step of acquiring a tomographic image of the subject;
Subject information obtaining step for obtaining normal data of a predetermined layer constituting the subject;
A program for causing a computer to execute a layer detection step of detecting the predetermined layer from the tomographic image using normal data of the predetermined layer acquired in the subject information acquisition step.   A computer-readable storage medium storing the program according to claim 12 or 13.

Images