JP7323893B2 - Standard brain model generation system, standard brain model generation method, and standard brain model generation program - Google Patents

Description

The present invention relates to a standard brain model generation system, a standard brain model generation method, and a standard brain model generation program.

In brain tumor removal surgery, future predictive surgery has been proposed that predicts and presents the postoperative survival rate and postoperative complication probability according to the planned removal range, and supports the decision-making of the operator (for example, non-patent literature 1).

Ken Masamune, Jun Okamoto, Kyoichiro Nambu, Manabu Tamura, Yoshiyuki Konishi, Yuki Horise, Kaori Kusuda, Takashi Maruyama, Satoko Ikuta, Hiroshi Izeki, Yoshihiro Muragaki. Further development of SCOT.” MEDIX 67: 4-7. (2017). J. Ashburner, et al., "A fast diffeomorphic image registration algorithm", NeuroImage, vol.38(1):95-113, 2007. Y. Liu, et al., “An ITK implementation of a physics-based non-rigid registration method for brain deformation in image-guided neurosurgery”, Frontiers in neuroinformatics, vol.8, no.33, pp.1-10, 2014. ICBM 152 Nonlinear atlases version 2009: http://www.bic.mni.mcgill.ca/ServicesAtlases/ICBM152NLin2009

In order to realize future predictive surgery, it is necessary to analyze many past cases and build a database of information such as past cases that serves as a scientific basis. However, the accumulation of such information is still insufficient. In particular, the construction of databases by existing methods largely relied on manual data collection of brain function positions and the like and their statistical processing (for example, simple averaging). In order to construct a sufficient database for future predictive surgery, a method that integrates information such as the location of brain functions, which varies from patient to patient, into a standard brain is desired.

The present invention has been made in view of such circumstances, and its object is to generate a standard brain model in which brain function positions that differ depending on the patient are integrated into a standard brain.

In order to solve the above problems, a standard brain model generation system according to one aspect of the present invention is a standard brain model generation system that integrates brain function positions into a standard brain, comprising: a brain function position acquisition unit; A processing unit, an intraoperative MRI image transformation unit, and a standard brain model generation unit are provided.

Another aspect of the invention is a standard brain model generation method. This method is a computer-implemented standard brain model generation method that integrates brain function locations into a standard brain, comprising: a brain function location acquisition step; a non-rigid body deformation preprocessing step; an intraoperative MRI image deformation step; and a standard brain model generation step.

Yet another aspect of the invention is a standard brain model generation program. This program causes a computer to execute a functional brain location acquisition step, a non-rigid body deformation preprocessing step, an intraoperative MRI image deformation step, and a standard brain model generation step.

Any combination of the above constituent elements, or mutual replacement of the constituent elements and expressions of the present invention in a method, apparatus, program, temporary or non-temporary storage medium recording the program, system, etc. is also effective as an aspect of the present invention.

According to the present invention, different brain function positions in different patients can be integrated into the standard brain.

1 is a functional block diagram of a standard brain model generation system according to a first embodiment; FIG. 2 is a diagram showing a standard brain model generated by the standard brain model generation system of FIG. 1; FIG. FIG. 10 is a functional block diagram of a standard brain model generation system according to a second embodiment; FIG. FIG. 11 is a flow chart showing processing of a standard brain generation method according to a third embodiment; FIG.

BEST MODE FOR CARRYING OUT THE INVENTION The present invention will be described below based on preferred embodiments with reference to the drawings. In the embodiment and modified examples, the same or equivalent constituent elements and parts are denoted by the same reference numerals, and overlapping explanations are omitted as appropriate. In addition, the dimensions of parts in each drawing are appropriately enlarged or reduced for easy understanding. Also, in each drawing, some of the elements that are not important for explaining the embodiments are omitted. Also, although terms including ordinal numbers such as first, second, etc. are used to describe various components, these terms are used only for the purpose of distinguishing one component from other components, and the terms The constituent elements are not limited by

Before describing specific embodiments, basic knowledge will be described. In brain tumor resection surgery, maximal resection is ideal because there is a positive correlation between resection rate and survival time. On the other hand, in tumor resection adjacent to the Eloquent region, which is involved in language and motor functions, excessively aggressive resection may cause postoperative complications such as paralysis and aphasia. Therefore, maximal tumor resection and minimal postoperative complications are desirable when considering the patient's prognosis. For this purpose, it is necessary to accurately confirm the location and size of the tumor, which varies from patient to patient, as well as the location of brain structures and functions. An intelligent operating room has been built as a facility to realize this. An intelligent operating room is an operating room equipped with advanced medical equipment such as "open MRI", "surgical navigation system", and "operating microscope". By transmitting the data acquired by these medical devices to the server via the LAN in the operating room, information during surgery can be recorded. That is, it is possible to accumulate motions and moving images of surgical instruments in past surgeries. The advanced medical equipment installed in the intelligent operating room will be described below.

[Open MRI]
Open MRI is an open, low-field MRI imager. In the conventional brain tumor removal surgery, the procedure was taken to confirm the craniotomy site based on the MRI image taken before the surgery, and to confirm the remaining tumor by imaging the MRI again after the tumor removal. However, when the brain is craniotomyed to remove a brain tumor, a phenomenon called brain shift occurs, in which the shape of the brain changes due to leakage of cerebrospinal fluid and changes in pressure. Since the shape of the brain changes due to the brain shift, there arises a problem that the shape of the brain in the MRI imaged before surgery differs from the shape of the brain to be treated. However, open MRI, which is an open MRI imaging device, is capable of imaging the deformed brain after craniotomy during surgery. This allows confirmation of the shape of the brain after deformation and residual tumor after tumor excision.

[Surgical navigation system]
A surgical navigation system is a system capable of presenting the relative positional relationship between the surgical tool and the brain by displaying the three-dimensional position of the surgical tool on intraoperative MRI. Current mainstream surgical navigation systems acquire the three-dimensional positions of surgical instruments using infrared stereo cameras and reflecting spheres. By displaying the three-dimensional position obtained here on intraoperative MRI imaged by open MRI, it is possible to visualize the relative positional relationship between the brain and surgical instruments during surgery.

[Operating microscope]
A surgical microscope is a microscope dedicated to surgery that clearly displays the operative field in detail. In neurosurgery, since delicate procedures are required, the surgical field is often magnified and observed, and surgical microscopes are used in many brain tumor resections. Due to the precision required, surgical microscopes have high resolution and are equipped with auxiliary systems to clarify the field of view. The images taken with the operating microscope are recorded as moving images. Some surgical microscopes have a laser irradiation function used for photodynamic therapy.

[Awake Brain Tumor Extraction Surgery]
Next, awake brain tumor removal surgery performed in brain tumor removal surgery will be described. Awake brain tumor resection is an operation in which the patient is awakened from anesthesia after craniotomy, and a brain tumor is removed while confirming the brain function position by applying electrical stimulation to the brain. As noted above, tumorectomy adjacent to the Eloquent area can result in postoperative complications due to damage to normal tissue. To reduce the incidence of postoperative complications, the surgeon needs to obtain intraoperative functional brain localization of the patient. For this reason, in intelligent surgery, a technique called awake craniotomy has been introduced. In this method, after the brain is craniotomy, the patient is woken up from anesthesia during the operation, and the operation is performed while giving the patient a task or moving the limbs. As an example, in an operation involving a tumor in the language field, a conversation with an expert doctor or an illustration (horse, pencil, kanji characters, hiragana characters, verb generation) is displayed on a monitor that the patient can see, and the patient answers what is shown. I have a language task to do. By applying electrical stimulation during these tasks, localization of the patient's brain function can be investigated. In other words, by performing electrical stimulation during a language task, the locations where the patient's wrong answers or negative reactions such as aphasia occur are recorded as brain function localization. In this way, by waking the patient during surgery and applying electrical stimulation, the brain function of each patient can be investigated, and tumor resection can be realized in consideration of the brain function.

A variety of information can be acquired in an awake brain tumor removal surgery performed in an intelligent operating room. These include information such as intraoperative/preoperative MRI images obtained from open MRI, surgical tool log information obtained from a surgical navigation system, electrical stimulation time, and the like. In addition, information such as the patient's task and electrical stimulation intensity is included in the surgical records performed by the surgeon.

The standard brain is a standardized brain that eliminates individual differences in morphology by transforming the brains of different individuals in shape and size into the same shape and size on the screen.

[First embodiment]
FIG. 1 shows a functional block diagram of a standard brain model generation system 100 according to the first embodiment. The standard brain model generation system 100 includes a functional brain position acquisition unit 10 , a non-rigid deformation preprocessing unit 20 , an intraoperative MRI image deformation unit 30 , and a standard brain model generation unit 40 .

A brain function location acquisition unit 10 acquires brain function location information from intraoperative/preoperative MRI images, surgical tool log information, intraoperative information including electrical stimulation time, and surgical record information including patient tasks and electrical stimulation intensity. . The brain function position acquisition unit 10 plots the acquired brain function positions in the image space of the same coordinate system as the intraoperative MRI image. As a result, information such as brain function positions is imaged. The brain function position acquisition unit 10 outputs the plotted brain function positions to the non-rigid body deformation preprocessing unit 20 .

The non-rigid body deformation preprocessing unit 20 performs preprocessing necessary for subsequent processing on the plotted image received from the brain function location acquisition unit 10 . Specifically, the non-rigid body deformation preprocessing unit 20 performs the following processing.
・Bias correction of intraoperative and preoperative MRI images ・Noise removal of intraoperative MRI images ・Brain to determine non-rigid deformation regions The region segmentation non-rigid body deformation preprocessing unit 20 transmits the preprocessed image data to the intraoperative MRI image deformation unit 30 .

The intraoperative MRI image transforming unit 30 calculates the displacement from the intraoperative MRI image received from the non-rigid deformation preprocessing unit 20 to the standard brain, and non-rigidly transforms it. The non-rigid body deformation may use a known technique, for example, DARTEL (Diffeomorphic Anatomic Registration using Exponentialized Lie algebra) (see, for example, Non-Patent Document 2). The intraoperative MRI image transformation unit 30 transmits the non-rigid body transformed MRI image to the standard brain model generation unit 40 .

The standard brain model generating unit 40 calculates the displacement of the standard brain from the deformed MRI image received from the intraoperative MRI image transforming unit 30, non-rigidly deforms it, and integrates it into the standard brain to generate the standard brain model. . A known technique may be used for the deformation, and for example, a non-rigid body deformation method based on the finite element method can be used (see, for example, Non-Patent Document 3). Standard brains can use existing models. Since the shape and size of the brain differ according to race, multiple standard brain models have been proposed. For example, in the case of a Japanese standard brain, an image (JP50NMI_template) obtained by averaging 50 healthy Japanese brains using SPM (Statistical Particle Mapping) can be used. In the case of Europeans and Americans, images obtained by averaging healthy brains can also be used (see, for example, Non-Patent Document 4). Thus, this embodiment is applicable to brains of various races. The standard brain model generation unit 40 outputs the generated standard brain model.

FIG. 2 shows a standard brain model generated by the standard brain model generation system 100. As shown in FIG. In this example, the anterior language area, the posterior language area and the motor area are integrated into the normal brain. Each point shown in FIG. 2 is mapped on the standard brain as a brain functional position related to cases such as cessation of speech, inhibition of speech, and movement of each part of the body. As described above, according to the present embodiment, brain function positions that differ from patient to patient can be integrated into the standard brain.

[Second embodiment]
FIG. 3 shows a functional block diagram of the standard brain model generation system 110 according to the second embodiment. The standard brain model generation system 110 includes a functional brain location acquisition unit 10 , a non-rigid body deformation preprocessing unit 20 , an intraoperative MRI image transformation unit 30 , a standard brain model generation unit 40 and an individual brain projection unit 50 . That is, the standard brain model generation system 110 includes a personal brain projection unit 50 in addition to the configuration of the standard brain model generation system 100 of FIG.

The individual brain projection unit 50 projects the standard brain model generated by the standard brain model generation unit 40 onto an MRI image of the patient's individual brain using non-rigid deformation. The MRI image of the patient's individual brain may be an intraoperative MRI image or a preoperative MRI image. According to the present embodiment, brain function localization for each individual patient can be visualized from the brain function location information integrated into the standard brain model.

[Third Embodiment]
FIG. 4 is a flow chart showing the processing of the standard brain generation method according to the third embodiment. The method comprises the steps of acquiring functional brain location information S10, performing non-rigid deformation preprocessing S20, deforming intraoperative MRI images S30, and generating a standard brain model S40.

At step S10, the method obtains brain function location information from intraoperative information including intraoperative and preoperative MRI images, surgical tool log information and electrical stimulation time, and surgical record information including patient task and electrical stimulation intensity. Furthermore, the acquired brain function positions are plotted in the image space of the same coordinate system as the intraoperative MRI image. As a result, information such as brain function positions is imaged.

In step S20, the method performs preprocessing necessary for the processing after step S30 on the image plotted in step S10. Specifically, the following processes are executed in step S20.
・Bias correction of intraoperative and preoperative MRI images ・Noise removal of intraoperative MRI images ・Brain to determine non-rigid deformation regions Region segmentation

At step S30, the method computes the displacement from the intra-operative MRI image preprocessed at step S20 to the normal brain and non-rigidly deforms it.

In step S40, the method calculates the displacement of the standard brain from the MRI image deformed in step S30, non-rigidly deforms it and integrates it into the standard brain to generate and output a standard brain model.
According to this embodiment, brain function positions that differ from patient to patient can be integrated into the standard brain.

[Fourth embodiment]
A computer program according to the third embodiment causes a computer to execute the flow shown in FIG. That is, the present program performs a step S10 of acquiring brain function position information, a step S20 of performing non-rigid body deformation preprocessing, a step S30 of deforming an intraoperative MRI image, and a step S40 of generating a standard brain model on a computer. let it run.
According to the present embodiment, a program that integrates brain function positions that differ from patient to patient into a standard brain can be implemented in software, so that a standard brain model can be generated with high accuracy using a computer.

The present invention has been described above based on the embodiments. It should be understood by those skilled in the art that the embodiments are examples, and that various modifications can be made to combinations of each component and each treatment process, and that such modifications are within the scope of the present invention. .

The processing performed using non-rigid deformation in the first and second embodiments may be performed using machine learning instead of non-rigid deformation.

REFERENCE SIGNS LIST 100 Standard brain model generation system 110 Standard brain model generation system 10 Brain function position acquisition unit 20 Non-rigid transformation preprocessing unit 30 Intraoperative MRI image deformation unit 40 Standard brain model generation unit 50 … Personal Brain Projector.

Claims (15)

A standard brain model generation system that integrates brain function positions into a standard brain,
a brain function location acquisition unit;
a non-rigid deformation pretreatment unit;
an intraoperative MRI image deformation unit;
a standard brain model generator ;
The brain function location acquisition unit obtains brain function location information from intraoperative MRI images, preoperative MRI images, surgical instrument log information, intraoperative information including electrical stimulation time, and surgical record information including patient tasks and electrical stimulation intensity. Acquired,
1. A standard brain model generation system characterized by imaging information on brain function positions by plotting acquired brain function positions in an image space of the same coordinate system as an intraoperative MRI image. The non-rigid body deformation preprocessing unit includes:
registration to the standard brain by rigid body transformation for the intraoperative and preoperative MRI images;
bias correction of intraoperative and preoperative MRI images;
denoising intraoperative MRI images;
segmentation of brain regions to determine deformation regions for non-rigid deformation;
2. The standard brain model generation system according to claim 1, wherein: 3. The standard brain according to claim 2, wherein the intraoperative MRI image transforming unit calculates the displacement of the standard brain from the intraoperative MRI image received from the non-rigid deformation preprocessing unit, and non-rigidly transforms it. model generation system. The standard brain model generating unit calculates the displacement of the standard brain from the deformed MRI image received from the intraoperative MRI image transforming unit, non-rigidly deforms it, integrates it into the standard brain, and generates the standard brain model. 4. The standard brain model generation system according to claim 3, characterized by: further equipped with a personal brain projection unit,
5. The standard brain according to claim 4, wherein the individual brain projection unit projects the standard brain model generated by the standard brain model generation unit onto an MRI image of the patient's individual brain using non-rigid deformation. model generation system. A computer-implemented standard brain model generation method for integrating brain function locations into a standard brain, comprising:
a brain function location acquisition step;
a non-rigid deformation pretreatment step;
an intraoperative MRI image deformation step;
a standard brain model generation step ;
The brain function position acquisition step obtains brain function position information from intraoperative information including intraoperative MRI images, preoperative MRI images, surgical tool log information and electrical stimulation time, and surgical record information including patient tasks and electrical stimulation intensity. Acquired,
A standard brain model generation method characterized by plotting acquired brain function positions in an image space of the same coordinate system as an intraoperative MRI image . The non-rigid deformation pretreatment step includes:
registration to the standard brain by rigid body transformation for the intraoperative and preoperative MRI images;
bias correction of intraoperative and preoperative MRI images;
denoising intraoperative MRI images;
segmentation of brain regions to determine deformation regions for non-rigid deformation;
7. The standard brain model generation method according to claim 6, wherein: 8. The method according to claim 7, wherein the intraoperative MRI image deformation step calculates a displacement from the intraoperative MRI image preprocessed in the non-rigid deformation preprocessing step to a standard brain, and non-rigid deformation is performed on the displacement. Standard brain model generation method as described. The standard brain model generation step calculates the displacement of the standard brain from the deformed MRI image received from the intraoperative MRI image deformation step, non-rigidly deforms it, and integrates it into the standard brain to generate the standard brain model. 9. The standard brain model generation method according to claim 8, characterized by: Further equipped with a personal brain projection step,
10. The standard brain according to claim 9, wherein the individual brain projection step projects the standard brain model generated by the standard brain model generation step onto an MRI image of the patient's individual brain using non-rigid body deformation. Model generation method. A standard brain model generation program for integrating brain function positions into a standard brain,
a brain function location acquisition step;
a non-rigid deformation pretreatment step;
an intraoperative MRI image deformation step;
causing a computer to perform a standard brain model generation step;
The brain function position acquisition step obtains brain function position information from intraoperative information including intraoperative MRI images, preoperative MRI images, surgical tool log information and electrical stimulation time, and surgical record information including patient tasks and electrical stimulation intensity. Acquired,
A standard brain model generation program characterized by plotting acquired brain function positions in an image space of the same coordinate system as an intraoperative MRI image . The non-rigid deformation pretreatment step includes :
registration to the standard brain by rigid body transformation for the intraoperative and preoperative MRI images;
bias correction of intraoperative and preoperative MRI images;
denoising intraoperative MRI images;
segmentation of brain regions to determine deformation regions for non-rigid deformation;
12. The standard brain model generation program according to claim 11, characterized by performing 13. The standard brain according to claim 12, wherein the intraoperative MRI image deformation step calculates a displacement from the intraoperative MRI image received from the non-rigid body deformation preprocessing step to the standard brain, and non-rigid body deformation is applied thereto. Model generation program. The standard brain model generation step calculates the displacement of the standard brain from the deformed MRI image received from the intraoperative MRI image deformation step, non-rigidly deforms it, and integrates it into the standard brain to generate the standard brain model. 14. The standard brain model generation program according to claim 13, characterized by: Further equipped with a personal brain projection step,
15. The standard brain according to claim 14, wherein the individual brain projection step projects the standard brain model generated by the standard brain model generation step onto an MRI image of the patient's individual brain using non-rigid deformation. Model generation program.

Images