Brain injury identification method and identification system

Abstract

The application discloses a brain injury identification method and an identification system. The identification system identifies the injury of a patient according to the identification method. The identification method comprises the following steps: firstly, the electrical impedance tomography (EIT) of the brain of the patient is performed to obtain the EIT image of the brain of the patient; then, the region with abnormal resistivity distribution in the EIT image of the brain is taken as a reference for scanning of a near-infrared spectrometer, and a scanning path of the near-infrared spectrometer is planned; finally, the near-infrared spectrometer is scanned along the scanning path, and the lesion range and the lesion type are determined according to the obtained near-infrared spectral data. The electrodes need not be arranged on the whole head of the patient, and the automatic and rapid identification of the lesion range and the nature of the brain injury is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of near-infrared spectroscopy detection technology, and particularly relates to a brain injury identification method and an identification system. BACKGROUND

[0002] The information of the range and nature of brain injury is of great value to improve the accuracy of injury classification and guide the first-aid treatment. In the events of earthquakes, major traffic accidents and wars, the mortality and disability rate of brain injury is extremely high, so it is necessary to reach the scene at the first time and use the appropriate technology on the spot to quickly define the range and nature of intracranial brain injury lesions. Large imaging equipment such as CT and MRI and their vehicle-mounted forms cannot quickly reach the scene due to long distances or traffic congestion. Therefore, there is an urgent need for a technology that can quickly identify brain injury.

[0003] There is a near-infrared spectroscopy identification technology in the prior art, and the device has portability. The basic principle of the near-infrared spectroscopy technology for detecting intracranial lesions is that there is a difference in the content of hemoglobin between lesion tissue and normal brain tissue, especially the hematoma formed by brain hemorrhage, which can cause a significant change in near-infrared reflection signal. The lesion type and lesion range are identified by the change of near-infrared spectroscopy data.

[0004] The patent with the application number 201611034981.7 discloses a non-invasive intracranial hematoma detection method based on multi-channel near-infrared spectroscopy imaging. The method needs to search the position of intracranial hematoma on the full scalp in a coarse scale, and then search the boundary of the hematoma in a fine scale. However, the technical solution needs to arrange electrodes on the whole brain, the detection time is long, and the detection process lacks target guidance, so the detection efficiency is low. SUMMARY

[0005] In view of the deficiencies in the prior art, the present application provides a brain injury identification method and an identification system, which can quickly identify the range and type of brain injury lesions. The specific technical solutions are as follows:

[0006] In the first aspect, a brain injury identification method is provided, comprising:

[0007] Obtaining a cranial EIT image, and planning a scanning path of a near-infrared spectrometer through an abnormal resistivity distribution area in the cranial EIT image;

[0008] According to the planned scanning path, the near-infrared spectrometer is guided to scan, and the near-infrared spectroscopy data obtained by the near-infrared spectrometer scanning are used to determine the lesion range and lesion type.

[0009] In combination with the first aspect, in the first implementation manner of the first aspect, the obtaining of the cranial EIT image comprises:

[0010] acquire a top view image of the head, and determine a brain boundary from the top view image of the head;

[0011] acquire brain EIT data, and reconstruct a brain EIT image with a real boundary from the brain EIT data, taking the brain boundary as an imaging region.

[0012] In combination with the first aspect, in a second implementation manner of the first aspect, planning the scan path of the near-infrared spectrometer comprises:

[0013] constructing a two-dimensional coordinate space with a midpoint of a curve segment projected on the surface of the brain by a left-right symmetry axis of the brain as an origin;

[0014] mapping the brain EIT image into the two-dimensional coordinate space to determine an initial boundary position of an abnormal resistivity distribution area in the brain EIT image;

[0015] planning a scan path of the near-infrared spectrometer according to the initial boundary position of the abnormal resistivity distribution area.

[0016] In combination with the second implementation manner of the first aspect, in a third implementation manner of the first aspect, further comprising:

[0017] planning a boundary expansion scan path of the near-infrared spectrometer according to the scan path;

[0018] guiding the near-infrared spectrometer to perform an expansion scan based on the planned boundary expansion scan path, and determining a lesion attribute and a lesion type of an expansion scan point from near-infrared spectrum data obtained by the scan;

[0019] repeating the above until the near-infrared spectrometer cannot scan abnormal near-infrared spectrum data, and / or;

[0020] planning a boundary convergence scan path of the near-infrared spectrometer according to the scan path;

[0021] guiding the near-infrared spectrometer to perform a convergence scan based on the planned boundary convergence scan path, and determining a lesion range and a lesion type of a convergence area from near-infrared spectrum data obtained by the scan;

[0022] repeating the above until a new boundary convergence scan path cannot be planned.

[0023] In combination with the third implementation manner of the first aspect, in a fourth implementation manner of the first aspect, planning the boundary expansion scan path and the boundary convergence scan path of the near-infrared spectrometer comprises:

[0024] determining position coordinates of each scan point corresponding to the expansion scan and position coordinates of each scan point corresponding to the convergence scan from position coordinates of each scan point of the near-infrared spectrometer on the scan path;

[0025] The boundary outward scanning path is planned according to position coordinates of all scanning points corresponding to the outward scanning, and the boundary converging scanning path is planned according to position coordinates of all scanning points corresponding to the converging scanning.

[0026] With reference to the first aspect, in a fifth implementation manner of the first aspect, the near-infrared spectrum data obtained by the near-infrared spectrometer is used to determine the lesion range and the lesion type, and the method comprises the following steps:

[0027] The near-infrared spectrum data of each scanning point is subjected to spectrum analysis to determine a characteristic absorption frequency band corresponding to each scanning point;

[0028] The characteristic absorption frequency band corresponding to each scanning point is compared with a preset near-infrared characteristic absorption frequency band of a lesion to determine a lesion attribute and a lesion type corresponding to each scanning point;

[0029] The lesion range is determined according to scanning regions corresponding to scanning points with positive lesion attributes, and the lesion type is determined according to lesion types corresponding to scanning points with positive lesion attributes.

[0030] In a second aspect, a brain injury identification system is provided, and the system comprises:

[0031] An image acquisition module configured to acquire a cranial EIT image;

[0032] A path planning module configured to plan a scanning path of a near-infrared spectrometer through an abnormal resistivity distribution area in the cranial EIT image;

[0033] An injury identification module configured to guide the near-infrared spectrometer to perform scanning according to the planned scanning path, and to determine a lesion range and a lesion type according to near-infrared spectrum data obtained by the near-infrared spectrometer.

[0034] With reference to the second aspect, in a first implementation manner of the second aspect, the image acquisition module comprises:

[0035] A boundary extraction unit configured to acquire a head top view image and determine a cranial boundary through the head top view image;

[0036] An image reconstruction unit configured to acquire cranial EIT data, take the cranial boundary as an imaging area, and reconstruct a cranial EIT image of a real boundary according to the cranial EIT data.

[0037] With reference to the second aspect, in a second implementation manner of the second aspect, the path planning module comprises:

[0038] A coordinate space construction unit configured to construct a two-dimensional coordinate space with a midpoint of a cranial left-right symmetry axis projection curve segment on a cranial surface as an origin;

[0039] an image mapping conversion unit, configured to map the craniocerebral EIT image into a two-dimensional coordinate space, and determine an initial boundary position of an abnormal resistivity distribution area in the craniocerebral EIT image;

[0040] a scanning path planning unit, configured to plan a scanning path of the near-infrared spectrometer according to the initial boundary position of the abnormal resistivity distribution area.

[0041] With reference to the second implementation manner of the second aspect, in a third implementation manner of the second aspect, the path planning module further includes:

[0042] an outward expansion path planning unit, configured to plan a boundary outward expansion scanning path of the near-infrared spectrometer according to the scanning path;

[0043] the injury identification module guides the near-infrared spectrometer to perform outward expansion scanning based on the planned boundary outward expansion scanning path, and determines a lesion attribute and a lesion type of an outward expansion scanning point through near-infrared spectrum data obtained by scanning;

[0044] the process is repeated until the near-infrared spectrometer cannot scan abnormal near-infrared spectrum data, and / or;

[0045] a converging path planning unit, configured to plan a boundary converging scanning path of the near-infrared spectrometer according to the scanning path;

[0046] the injury identification module guides the near-infrared spectrometer to perform converging scanning based on the planned boundary converging scanning path, and determines a lesion range and a lesion type of a converging area through near-infrared spectrum data obtained by scanning;

[0047] the process is repeated until a new boundary converging scanning path cannot be planned.

[0048] With reference to the third implementation manner of the second aspect, in a fourth implementation manner of the second aspect, the outward expansion path planning unit determines position coordinates of each scanning point corresponding to the outward expansion scanning through position coordinates of each scanning point of the near-infrared spectrometer on the scanning path, and plans the boundary outward expansion scanning path according to the position coordinates of all scanning points corresponding to the outward expansion scanning;

[0049] the converging path planning unit determines position coordinates of each scanning point corresponding to the converging scanning through position coordinates of each scanning point of the near-infrared spectrometer, and plans the boundary converging scanning path according to the position coordinates of all scanning points corresponding to the converging scanning.

[0050] With reference to the second aspect, in a fifth implementation manner of the second aspect, the injury identification module includes:

[0051] a spectrum analysis unit configured to perform spectrum analysis on the near-infrared spectrum data of each scanning point to determine a characteristic absorption frequency band corresponding to each scanning point;

[0052] a frequency band comparison unit configured to compare the characteristic absorption frequency band corresponding to each scanning point with a preset lesion near-infrared characteristic absorption frequency band respectively to determine a lesion attribute and a lesion type corresponding to each scanning point;

[0053] a lesion identification unit configured to determine a lesion range according to a scanning region corresponding to a scanning point with a positive lesion attribute and determine a lesion type according to a lesion type corresponding to a scanning point with a positive lesion attribute.

[0054] Beneficial effects: the brain lesion identification method and the identification system can combine the electrical impedance imaging technology and the near-infrared spectrum technology organically, perform electrical impedance imaging on the measured cranium first, then take the abnormal resistivity distribution target presented by the electrical impedance imaging as a guide for the near-infrared probe to detect on the cranium surface, so as to realize automatic and rapid identification of the brain lesion range and properties. BRIEF DESCRIPTION OF DRAWINGS

[0055] In order to more clearly illustrate the specific embodiments of the present application, the drawings required to be used in the specific embodiments will be briefly introduced below. In all the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0056] Figure 1 a flowchart of the brain lesion identification method provided by an embodiment of the present application;

[0057] Figure 2 a flowchart of the EIT image acquisition of the cranium provided by an embodiment of the present application;

[0058] Figure 3 a flowchart of the scanning path planning provided by an embodiment of the present application;

[0059] Figure 4 a flowchart of the identification of the lesion range and the lesion type provided by an embodiment of the present application;

[0060] Figure 5 a flowchart of the outward expansion scanning provided by an embodiment of the present application;

[0061] Figure 6 a flowchart of the convergent scanning provided by an embodiment of the present application;

[0062] Figure 7 a system block diagram of the brain lesion identification system provided by an embodiment of the present application;

[0063] Figure 8 a system block diagram of the image acquisition module provided by an embodiment of the present application;

[0064] Figure 9 The system block diagram of the path planning module provided by an embodiment of the present application;

[0065] Figure 10 The system block diagram of the damage identification module provided by an embodiment of the present application;

[0066] Figure 11 The scanning path diagram planned by the brain damage identification method and identification system provided by an embodiment of the present application;

[0067] Figure 12 The outward expansion scanning point diagram planned by the brain damage identification method and identification system provided by an embodiment of the present application;

[0068] Figure 13 The convergent scanning point diagram planned by the brain damage identification method and identification system provided by an embodiment of the present application;

[0069] Figure 14 The identification result diagram of the brain hemorrhage patient identified by the brain damage identification method and identification system provided by an embodiment of the present application;

[0070] Figure 15 The identification result diagram of the brain ischemia patient identified by the brain damage identification method and identification system provided by an embodiment of the present application;

[0071] Figure 16 The identification result diagram of the brain edema patient identified by the brain damage identification method and identification system provided by an embodiment of the present application. DETAILED DESCRIPTION

[0072] The embodiments of the technical solutions of the present application will be described in detail below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present application, and therefore only serve as examples, and cannot limit the protection scope of the present application.

[0073] Embodiment one

[0074] As shown in the flowchart of the brain damage identification method, the identification method comprises: Figure 1

[0075] Step 1, acquiring a craniocerebral EIT image, and planning a scanning path of a near-infrared spectrometer through an abnormal resistivity distribution area in the craniocerebral EIT image;

[0076] Step 2, guiding the near-infrared spectrometer to scan according to the planned scanning path, and determining a lesion range and a lesion type through near-infrared spectrum data obtained by the near-infrared spectrometer scanning.

[0077] ​Specifically, first, the patient's brain can be electrically impedance imaged to obtain the patient's brain EIT image, the region with abnormal resistivity distribution in the brain EIT image is taken as a reference for near-infrared spectrometer scanning, and the scanning path of the near-infrared spectrometer is planned, and then the near-infrared spectrometer is scanned along the scanning path, and the lesion range and lesion type are determined through the obtained near-infrared spectral data. In this way, electrodes do not need to be arranged on the entire head of the patient, and automatic and rapid identification of the brain injury lesion range and nature is achieved.

[0078] The following will be described in combination with Figure 2 The step of obtaining the brain EIT image in step 1 will be described in detail.

[0079] In this embodiment, the brain EIT image includes:

[0080] Step 1-1-1, obtaining a top view image of the head and determining the brain boundary through the top view image of the head;

[0081] Step 1-1-2, obtaining brain EIT data and taking the brain boundary as the imaging region, and reconstructing the brain EIT image with the real boundary according to the brain EIT data.

[0082] Specifically, first, a clear static top view image of the patient's head can be taken by a camera, the camera can send the captured top view image of the head to the host computer, the host computer can use existing image segmentation algorithms to segment the top view image of the head, and then identify the brain boundary through existing boundary recognition programs.

[0083] Then, the existing EIT system can collect EIT data by contacting the brain at a specific position through the scalp electrode, and send the collected EIT data to the host computer, and the host computer can calculate the resistivity distribution of the obtained EIT data to determine the EIT data with abnormal resistivity distribution. The host computer can also calculate the relative coordinates of all boundary points in the brain boundary, construct a finite element model for imaging within the brain boundary, determine the sensitivity coefficient matrix and regularize the matrix, and use the sensitivity coefficient method image reconstruction algorithm to perform brain EIT imaging with the real boundary on the collected EIT data with abnormal resistivity distribution, to obtain an abnormal resistivity distribution image in the imaging domain of the real brain boundary, i.e., the brain EIT image.

[0084] In the embodiment, the specific position of the brain can be 16 points equally spaced on the intersection line of the plane where the points are located at 1 cm above the upper edge of the left and right ears, 1 cm above the eyebrow center, and 1 cm above the occipital tuberosity, and the scalp. The EIT data collected is the boundary voltage data collected by the EIT system at the brain under multiple excitations. Among them, the EIT data of the abnormal resistivity distribution refers to the EIT data with large resistivity change after differential imaging; the abnormal resistivity distribution is composed of pixels with resistivity change absolute value in the range of 50%-100% of the maximum resistivity change absolute value of the whole brain EIT image.

[0085] The following will be described in conjunction with the accompanying Figure 3 The process of planning the scanning path of the near-infrared spectrometer in step 1 will be described in detail.

[0086] In the embodiment, the scanning path of the near-infrared spectrometer includes:

[0087] Step 1-2-1, constructing a two-dimensional coordinate space with the midpoint of the curve segment projected on the surface of the brain along the left-right symmetry axis of the brain as the origin;

[0088] Step 1-2-2, mapping the brain EIT image into the two-dimensional coordinate space to determine the initial boundary position of the abnormal resistivity distribution area in the brain EIT image;

[0089] Step 1-2-3, planning the scanning path of the near-infrared spectrometer according to the initial boundary position of the abnormal resistivity distribution area, specifically.

[0090] In the embodiment, the near-infrared spectrometer can use a 512-pixel InGaAs line array detector, which measures a wavelength range of 900-2500 nm, a wavelength resolution of 3 nm, and a signal-to-noise ratio of 15000:1, and can obtain continuous spectra in the near-infrared band of the area below the center point of the near-infrared probe in real time. The center point of the near-infrared probe is the midpoint of the line connecting the center point of the light source and the center point of the detector.

[0091] In order to facilitate the control of the near-infrared spectrometer to scan along the scanning path, the probe of the near-infrared spectrometer can be installed in the end probe box of the mechanical arm. The part of the probe that contacts the scalp can extend out of the probe box, and the probe body can make a linear motion in the probe box towards the part of the probe that contacts the scalp. The linear motion is that when the probe contacts the scalp, the part of the probe body is forced to retract into the box, and the amplitude of the motion is limited by the spring between the probe body and the bottom end of the box. The amplitude of the motion is supervised by the pressure sensor arranged between the spring and the bottom end of the box. The host computer can collect the pressure value on the pressure sensor in real time, and adjust the movement distance of the end of the mechanical arm to the scalp to adjust the pressure value, so that the pressure value is in the range of 180-200 N.

[0092] The host computer can send the generated cranium EIT image to the mechanical arm to plan a scanning path for the near-infrared spectrometer through the mechanical arm. When planning the scanning path, the mechanical arm can first take the midpoint of the line segment of the left-right symmetry axis in the imaging domain in the cranium EIT image as the origin, the right as the X-axis, and the upward as the Y-axis to construct a two-dimensional coordinate space. Then, the cranium EIT image is mapped into the two-dimensional coordinate space to determine the position coordinates of each pixel point in the cranium EIT image in the two-dimensional coordinate space. Finally, the scanning path of the probe is planned according to the initial boundary position of the abnormal resistivity distribution area in the two-dimensional coordinate space. Specifically, the coordinates (x i , y i ) of the discrete points on the initial boundary are read as scanning points in an equidistant manner, i is the discrete point number, and the scanning path of the near-infrared spectrometer is constructed through the discrete points. The constructed scanning path is shown in Figure 11 .

[0093] The initial boundary refers to the outer boundary of the region composed of the pixel points with the resistivity change absolute value in the range of 50%-100% of the maximum resistivity change absolute value of the entire image.

[0094] After the mechanical arm plans the scanning path of the near-infrared spectrometer, the probe of the near-infrared spectrometer can be manually pulled to the position where the center point of the probe coincides with the midpoint of the cranium surface projection curve segment of the left-right symmetry axis, so that the center point of the probe coincides with the virtual coordinate origin. The host computer can capture a position image of the position of the probe through the camera, position the spatial coordinates of the center point of the probe through the position image, and send the spatial coordinates to the mechanical arm. The mechanical arm can map the spatial coordinates of the center point of the probe into the two-dimensional coordinate space according to the position of the midpoint of the cranium surface projection curve segment in the two-dimensional coordinate space.

[0095] The mechanical arm can start to drive the center point of the probe to sequentially traverse the scanning points according to the scanning path. When reaching each scanning point, the mechanical arm sends the coordinates (x i , y i ) of the scanning point to the host computer, and the host computer immediately collects the near-infrared spectrum data NIRS i under the scanning point through the near-infrared spectrometer.

[0096] The process of determining the lesion range and lesion type from the near-infrared spectrum data scanned by the near-infrared spectrometer in step 2 will be described below in conjunction with the accompanying Figure 4 , including:

[0097] Step 4-1, performing spectral analysis on the near-infrared spectrum data of each scanning point to determine the characteristic absorption frequency band corresponding to each scanning point;

[0098] Step 4-2, compare the characteristic absorption frequency band corresponding to each scanning point with the preset lesion near-infrared characteristic absorption frequency band respectively to determine the lesion attribute and lesion type corresponding to each scanning point;

[0099] Step 4-3, determine the lesion range according to the scanning area corresponding to the scanning point with positive lesion attribute, and determine the lesion type according to the lesion type corresponding to the scanning point with positive lesion attribute.

[0100] Specifically, the host computer can first calculate the wavelength range corresponding to the 90% of the maximum absorption peak of the near-infrared spectrum data NIRS i obtained from the region below the scanning point to determine the characteristic absorption frequency band corresponding to the scanning point. Then, compare the calculated characteristic absorption frequency band with the pre-stored characteristic absorption frequency bands of common lesions such as cerebral hemorrhage, cerebral ischemia and cerebral edema.

[0101] If there is an overlap, the host computer marks the lesion attribute value of the scanning point (x i , y i ) as positive, and if there is no overlap, the lesion attribute of the scanning point (x i , y i ) is marked as negative. If the characteristic absorption frequency band of the scanning point (x i , y i ) has the largest range of overlap with the characteristic absorption frequency band of one of the cerebral hemorrhage, cerebral ischemia, cerebral edema and brain contusion lesions, the lesion type of the scanning point (x i , y i ) is marked as the corresponding lesion type.

[0102] The host computer can mark all scanning points with different geometric shapes on the real boundary cranial EIT image according to their lesion attributes and lesion types, such as solid triangles, solid squares, solid circles and solid prisms to represent different lesion types, so as to display the brain injury lesion range and lesion type, thereby realizing the automatic and rapid identification of brain injury lesion range and nature.

[0103] Embodiment Two

[0104] Embodiment Two is substantially the same as Embodiment One, the main difference is that it further includes:

[0105] Step 3-1, planning a boundary expansion scanning path for the near-infrared spectrometer according to the scanning path;

[0106] Step 3-2, guiding the near-infrared spectrometer to perform boundary expansion scanning based on the planned boundary expansion scanning path, and determining the lesion attribute and lesion type of the boundary expansion scanning point through the obtained near-infrared spectrum data;

[0107] Step 3-3: Repeat this process until the near-infrared spectrometer can no longer detect abnormal near-infrared spectral data.

[0108] Specifically, such as Figure 5 As shown, when planning the scanning path, the robotic arm can also plan an extended scanning path based on the scanning path planned based on the initial boundary. The extended scanning path is to plan all surface points (x, y, y) on the initial boundary. i y i Move one unit distance (1cm) in the opposite direction to the centroid of the initial boundary, forming a new outward scanning point (x). j y j ), through these extended scanning points (x j y j This forms a new boundary expansion scan path.

[0109] After the robotic arm drives the probe to measure the near-infrared spectral data of all scanning points along the scanning path, it then drives the probe to scan the outer scanning points along the outer scanning path. The main control computer receives the near-infrared spectral data of each outer scanning point sent by the near-infrared spectrometer, and uses the same process as described above to determine the lesion attributes and lesion types corresponding to the scanning points, to determine the lesion attributes and lesion types of each outer scanning point based on the corresponding near-infrared spectral data.

[0110] If there are outward scanning points with positive lesion attributes, the main control computer can control the robotic arm to scan all scanning points (x) on the initial boundary. i y i The scanner moves two units in the opposite direction to the centroid of the initial boundary-enclosed shape, and continues to plan a new boundary expansion scanning path. This process is repeated until the main control computer determines that no more expansion scanning points with positive lesion attributes appear, at which point the expansion scanning stops. The main control computer can then determine the extent and type of the patient's brain injury based on all expansion scanning points and the lesion attributes and types at each point. This allows for the detection of lesions outside the abnormal resistivity distribution area on the cranial EIT image, thereby improving the accuracy of the identification results.

[0111] Example 3

[0112] Example 3 is largely the same as Example 2, with the main difference being:

[0113] Step 4-1: Plan the boundary convergence scanning path of the near-infrared spectrometer according to the scanning path;

[0114] Step 4-2: Guide the near-infrared spectrometer to perform convergence scanning based on the planned boundary convergence scanning path, and determine the lesion range and lesion type in the convergence area through the near-infrared spectral data obtained from the scan.

[0115] Step 4-3, repeat the above steps until a new boundary convergent scanning path cannot be planned.

[0116] As shown in Figure 6 , after the outward expansion scanning is completed, the mechanical arm can continue the boundary convergent scanning according to the planned boundary convergent scanning path. Specifically, first, the mechanical arm can move all the scanning points (x i , y i ) on the initial boundary to the direction of the centroid of the figure surrounded by the initial boundary by a unit distance (1 cm), thereby forming new convergent scanning points (x k , y k ), and these convergent scanning points (x k , y k ) constitute a new boundary convergent scanning path.

[0117] Then, the mechanical arm moves the probe to the first scanning point on the initial boundary, and traverses all the convergent scanning points (x k , y k ) according to the planned boundary convergent scanning path to perform scanning. The host computer receives the near-infrared spectrum data of each convergent scanning point sent by the near-infrared spectrometer, and determines the lesion attribute and lesion type of each convergent scanning point according to the corresponding near-infrared spectrum data by using the same process as that of determining the scanning point. Repeat the steps until the mechanical arm cannot generate a new convergent scanning path, the entire scanning process is completed, and the host computer determines the lesion range and lesion type of the brain injury according to the lesion attributes and lesion types corresponding to all the scanning points, convergent scanning points and outward expansion scanning points. The identification method of this embodiment is used to identify the injuries of three patients with cerebral hemorrhage, cerebral ischemia and cerebral edema respectively, and the identification results are shown in Figure 14 , Figure 15 , Figure 16 .

[0118] Example Four

[0119] The system block diagram of the brain injury identification system is shown in Figure 7 , which comprises:

[0120] An image acquisition module configured to acquire a cranial EIT image;

[0121] A path planning module configured to plan a scanning path of a near-infrared spectrometer through an abnormal resistivity distribution area in the cranial EIT image;

[0122] An injury identification module configured to guide the near-infrared spectrometer to scan according to the planned scanning path, and determine a lesion range and a lesion type through near-infrared spectrum data obtained by the near-infrared spectrometer scanning.

[0123] Specifically, the image acquisition module of the host computer can perform electrical impedance tomography on the patient's brain to obtain the patient's brain EIT image. The host computer can send the brain EIT image to the path planning module of the mechanical arm, and the path planning module can take the region with abnormal electrical resistivity distribution in the brain EIT image as a reference for guiding the near-infrared spectrometer to scan, and plan a scanning path for the near-infrared spectrometer. Thus, the near-infrared spectrometer can be guided to scan the patient's head along the scanning path, and the near-infrared spectrometer can send the scanned near-infrared spectral data to the damage identification module of the host computer, and the damage identification module can determine the lesion range and lesion type based on the obtained near-infrared spectral data. In this way, electrodes do not need to be arranged on the entire head of the patient, and automatic and rapid identification of the brain damage lesion range and nature is achieved.

[0124] The following will be described in detail with reference to the accompanying drawings Figure 8 The image acquisition module will be described in detail.

[0125] In this embodiment, the image acquisition module comprises:

[0126] A boundary extraction unit configured to obtain a top view image of the head and determine the brain boundary through the top view image of the head;

[0127] An image reconstruction unit configured to obtain brain EIT data and reconstruct a brain EIT image with a real boundary based on the brain EIT data, taking the brain boundary as an imaging region.

[0128] Specifically, a still top view image of the patient's head is clearly captured by a camera, and the camera can send the captured top view image of the head to the boundary extraction unit of the host computer. The boundary extraction unit can use existing image segmentation algorithms to separate the top view image of the head, and then use existing boundary programs to identify the brain boundary.

[0129] EIT data is collected by the existing EIT system using scalp electrodes to contact specific positions of the brain, and the collected EIT data is sent to the image reconstruction unit of the host computer. The image reconstruction unit can calculate the electrical resistivity distribution of the obtained EIT data to determine the EIT data with abnormal electrical resistivity distribution. The image reconstruction unit can also calculate the relative coordinates of all boundary points in the brain boundary based on the brain boundary extracted by the boundary extraction unit, construct a finite element model for imaging within the brain boundary, determine the sensitivity coefficient matrix and regularize the matrix, and use the sensitivity coefficient method image reconstruction algorithm to perform real boundary brain EIT imaging on the collected EIT data with abnormal electrical resistivity distribution to obtain an abnormal electrical resistivity distribution image within the imaging domain of the real brain boundary, i.e., a brain EIT image.

[0130] In the embodiment, the specific positions of the brain can be 16 points equally spaced on the intersection line of the plane where the points are located and the scalp, the plane being at 1 cm above the upper edge of the left and right ears, 1 cm above the center of the eyebrows, and 1 cm above the occipital protuberance. The collected EIT data are the boundary voltage data collected by the EIT system at the brain under multiple excitations. The EIT data of the abnormal resistivity distribution are the EIT data with a large change in resistivity after differential imaging. The abnormal resistivity distribution is composed of pixels with an absolute value of resistivity change in the range of 50%-100% of the maximum absolute value of resistivity change in the whole EIT image of the brain.

[0131] The following will be described in detail with reference to the accompanying drawings. Figure 9 The path planning module is described in detail.

[0132] In the embodiment, the path planning module of the mechanical arm comprises:

[0133] A coordinate space construction unit configured to construct a two-dimensional coordinate space with the midpoint of the curve segment projected on the surface of the brain on the left-right symmetry axis of the brain as the origin;

[0134] An image mapping and conversion unit configured to map the EIT image of the brain into the two-dimensional coordinate space and determine the initial boundary position of the abnormal resistivity distribution area in the EIT image of the brain;

[0135] A scanning path planning unit configured to plan a scanning path of the near-infrared spectrometer according to the initial boundary position of the abnormal resistivity distribution area.

[0136] Specifically, in the embodiment, the near-infrared spectrometer can adopt a 512-pixel InGaAs linear array detector, which measures a wavelength range of 900-2500 nm, a wavelength resolution of 3 nm, and a signal-to-noise ratio of 15000:1, and can obtain continuous spectra of the region below the center point of the near-infrared probe in the near-infrared wavelength range in real time. The center point of the near-infrared probe is the midpoint of the line connecting the center point of the light source and the center point of the detector.

[0137] In order to facilitate the control of the near-infrared spectrometer to scan along the scanning path, the probe of the near-infrared spectrometer can be installed in the end probe box of the mechanical arm. The part of the probe that contacts the scalp can extend out of the probe box, and the probe body can make a linear motion in the probe box towards the part of the probe that contacts the scalp. The linear motion is that the part of the probe body is forced to retract into the box when the probe contacts the scalp, the amplitude of the motion is limited by the spring arranged between the probe body and the bottom end of the box, and the amplitude of the motion is supervised by the pressure sensor arranged between the spring and the bottom end of the box. The host computer can collect the pressure value on the pressure sensor in real time, and adjust the movement distance of the end of the mechanical arm to the scalp to adjust the pressure value, so that the pressure value is in the range of 180-200 N.

[0138] The host computer can send the generated cranium EIT image to a path coordinate space construction unit of the mechanical arm, and plan a scanning path of the near-infrared scanner through the coordinate space construction unit. When planning the scanning path, the coordinate space construction unit can first take the midpoint of a line segment of a left-right symmetry axis in the imaging domain in the cranium EIT image as an origin, take the right direction as an X axis, and take the upward direction as a Y axis to construct a two-dimensional coordinate space.

[0139] The image mapping conversion unit of the mechanical arm can map the cranium EIT image into the two-dimensional coordinate space to determine the position coordinates of each pixel point in the cranium EIT image in the two-dimensional coordinate space. The scanning path planning unit of the mechanical arm can plan the scanning path of the probe according to the initial boundary position of the abnormal resistivity distribution area in the two-dimensional coordinate space. Specifically, the coordinates (x i , y i ) of the discrete points on the initial boundary are read as scanning points in an equal-interval manner, i is the discrete point number, and the scanning path of the near-infrared spectrometer is constructed through the discrete points.

[0140] The initial boundary refers to the outer boundary of a region composed of pixel points whose resistivity changes in the EIT image are in the range of 50%-100% of the maximum resistivity change absolute value of the entire image.

[0141] After the mechanical arm plans the scanning path of the near-infrared scanner, the probe of the near-infrared spectrometer can be manually pulled to a position where the probe center point coincides with the midpoint of the cranium left-right symmetry axis projection curve segment on the cranium surface, so that the probe center point coincides with the virtual coordinate origin. The host computer can capture a position image of the position of the probe through a camera, position the spatial coordinates of the probe center point through the position image, and send the spatial coordinates of the probe center point to the mechanical arm. The mechanical arm can map the spatial coordinates of the probe center point into the two-dimensional coordinate space according to the position of the midpoint of the cranium surface projection curve segment in the two-dimensional coordinate space.

[0142] The mechanical arm can then drive the probe center point to sequentially traverse the scanning points according to the scanning path. When reaching each scanning point, the mechanical arm sends the coordinates (x i , y i ) of the scanning point to the host computer, and the host computer immediately collects near-infrared spectrum data NIRS i under the scanning point through the near-infrared spectrometer.

[0143] The following will be described in detail with reference to the accompanying drawings. Figure 10 The damage identification module of the host computer will be described in detail.

[0144] In this embodiment, the damage identification module includes:

[0145] The spectrum analysis unit is configured to perform spectrum analysis on the near-infrared spectrum data of each scanning point to determine the characteristic absorption frequency band corresponding to each scanning point.

[0146] The frequency band comparison unit is configured to compare the characteristic absorption frequency band corresponding to each scanning point with the preset lesion near-infrared characteristic absorption frequency band respectively to determine the lesion attribute and lesion type corresponding to each scanning point.

[0147] The damage identification unit is configured to determine the lesion range according to the scanning area corresponding to the scanning point with positive lesion attribute, and determine the lesion type according to the lesion type corresponding to the scanning point with positive lesion attribute.

[0148] The damage identification module of the host computer includes the spectrum analysis unit, the frequency band comparison unit and the damage identification unit. The spectrum analysis unit can perform spectrum analysis on the acquired near-infrared spectrum data of the scanning point, calculate the wavelength range corresponding to the 90% of the maximum absorption peak of the acquired near-infrared spectrum data NIRS i of the region below the scanning point, and determine the characteristic absorption frequency band corresponding to the scanning point.

[0149] The frequency band comparison unit can compare the characteristic absorption frequency band calculated by the spectrum analysis unit with the pre-stored characteristic absorption frequency bands of common lesions such as cerebral hemorrhage, cerebral ischemia and cerebral edema. If there is a coincidence, the host computer marks the lesion attribute value of the scanning point (x i , y i ) as positive, and if there is no coincidence, marks the lesion attribute of the scanning point (x i , y i ) as negative. If the characteristic absorption frequency band of the scanning point (x i , y i ) coincides with the characteristic absorption frequency band of one of the cerebral hemorrhage, cerebral ischemia, cerebral edema and cerebral contusion lesions in the largest range, the lesion type of the scanning point (x i , y i ) is marked as the corresponding lesion type.

[0150] The damage identification unit can mark all scanning points with different geometric shapes on the real boundary brain EIT image according to their lesion attributes and lesion types, such as solid triangles, solid squares, solid circles and solid prisms to represent different lesion types, so as to display the brain injury lesion range and lesion type, and realize the automatic and rapid identification of the brain injury lesion range and nature.

[0151] Embodiment five

[0152] Embodiment five is substantially the same as embodiment four, with the main difference being that the path planning module further comprises an outer expansion path planning unit configured to plan an outer expansion scanning path for the near-infrared spectrometer according to the scanning path;

[0153] The lesion identification module guides the near-infrared spectrometer to perform outer expansion scanning based on the planned outer expansion scanning path, and determines the lesion attribute and lesion type of the outer expansion scanning points based on the near-infrared spectral data obtained by scanning.

[0154] This is repeated until the near-infrared spectrometer cannot scan abnormal near-infrared spectral data.

[0155] Specifically, the path planning module further comprises an outer expansion path planning unit, which can plan an outer expansion scanning path based on the scanning path planned based on the initial boundary. The outer expansion scanning path is to move all the points (x i , y i ) on the initial boundary by one unit distance (1 cm) in the opposite direction of the direction of the centroid of the figure enclosed by the initial boundary, to form new outer expansion scanning points (x j , y j ), and to form a new outer expansion scanning path by these outer expansion scanning points (x j , y j ).

[0156] After the mechanical arm drives the probe to measure the near-infrared spectral data of all the scanning points along the scanning path, it can drive the probe to scan the outer expansion scanning points along the outer expansion scanning path. The lesion identification module can receive the near-infrared spectral data of each outer expansion scanning point sent by the near-infrared spectrometer, and determine the lesion attribute and lesion type of each outer expansion scanning point according to the corresponding near-infrared spectral data using the same process as described above to determine the lesion attribute and lesion type corresponding to the scanning points.

[0157] If there is an outer expansion scanning point with a positive lesion attribute, the host computer can control the mechanical arm to move all the scanning points (x i , y i ) on the initial boundary by two unit distances in the opposite direction of the direction of the centroid of the figure enclosed by the initial boundary, and continue to plan a new outer expansion scanning path for scanning. This is repeated until the lesion identification module determines that there is no longer an outer expansion scanning point with a positive lesion attribute, and then the outer expansion scanning is stopped. In this way, it can be determined whether there is a lesion outside the abnormal resistivity distribution area on the craniocerebral EIT image, thereby improving the accuracy of the identification result.

[0158] Embodiment six

[0159] Embodiment six is substantially the same as embodiment five, with the main difference being that the path planning module further comprises a converging path planning unit. The converging path planning unit is configured to plan a boundary converging scanning path according to the scanning path;

[0160] The lesion identification module guides the near-infrared spectrometer to perform converging scanning based on the planned boundary converging scanning path, and determines the lesion range and lesion type of the converging area based on the near-infrared spectral data obtained by scanning;

[0161] This is repeated until a new boundary converging scanning path cannot be planned.

[0162] Specifically, after completing the outward expansion scanning, the mechanical arm can continue to perform boundary converging scanning according to the boundary converging scanning path planned by the converging path planning unit. The converging path planning unit can move all scanning points (x i , y i ) on the initial boundary to the direction of the centroid of the figure formed by the initial boundary by a unit distance (1 cm), thereby forming new converging scanning points (x k , y k ), and these converging scanning points (x k , y k ) constitute a new boundary converging scanning path.

[0163] When performing converging scanning, the mechanical arm can first move the probe to the first scanning point on the initial boundary, and then traverse all the converging scanning points (x k , y k ) according to the boundary converging scanning path planned by the converging path planning unit to perform scanning. The main control computer receives the near-infrared spectral data of each converging scanning point sent by the near-infrared spectrometer, and the lesion identification module determines the lesion attribute and lesion type of each converging scanning point according to the corresponding near-infrared spectral data using the same process as that for determining the lesion attribute and lesion type corresponding to the scanning points. This step is repeated until the converging path planning unit cannot generate a new converging scanning path, and the entire scanning process ends. The identification system of this embodiment was used to identify lesions in three types of patients, namely, cerebral hemorrhage, cerebral ischemia, and cerebral edema, and the identification results are shown in Figure 14 、 Figure 15 、 Figure 16 .

[0164] The above examples are only used to illustrate the technical solutions of the present application, but not limit the present application; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that the technical solutions recorded in the foregoing examples can be modified, or some or all of the technical features can be replaced equivalently; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and they should be covered in the scope of the claims and the specification of the present application.

Claims

1. A brain injury identification system, characterized by, The brain injury identification system comprises: an image acquisition module configured to acquire a brain EIT image; a path planning module configured to plan a scanning path of a near-infrared spectrometer through an abnormal resistivity distribution area in the brain EIT image; an injury identification module configured to guide the near-infrared spectrometer to scan according to the planned scanning path, and determine a lesion range and a lesion type according to near-infrared spectrum data obtained by the scanning of the near-infrared spectrometer. The path planning module comprises: a coordinate space construction unit configured to construct a two-dimensional coordinate space with a midpoint of a projection curve segment of a brain left-right symmetry axis on a brain surface as an origin; an image mapping and conversion unit configured to map the brain EIT image into the two-dimensional coordinate space, and determine an initial boundary position of the abnormal resistivity distribution area in the brain EIT image. The scanning path planning unit is configured to read the coordinates of the discrete points on the initial boundary in an equidistant manner As the scanning points, The discrete points are numbered, and the scanning path of the near-infrared spectrometer is constructed through the discrete points.

2. The brain injury identification system of claim 1, wherein The image acquisition module comprises: a boundary extraction unit configured to acquire a head top view image, and determine a brain boundary through the head top view image; an image reconstruction unit configured to acquire brain EIT data, and reconstruct a brain EIT image of a real boundary according to the brain EIT data, with the brain boundary as an imaging area.

3. The brain injury identification system of claim 2, wherein The path planning module further comprises: an outward expansion path planning unit configured to plan a boundary outward expansion scanning path of the near-infrared spectrometer according to the scanning path; the injury identification module guides the near-infrared spectrometer to outward expansion scan based on the planned boundary outward expansion scanning path, and determines a lesion attribute and a lesion type of an outward expansion scanning point according to near-infrared spectrum data obtained by the outward expansion scan; this is repeated until the near-infrared spectrometer cannot scan abnormal near-infrared spectrum data, and / or; a convergent path planning unit configured to plan a boundary convergent scanning path of the near-infrared spectrometer according to the scanning path; the injury identification module guides the near-infrared spectrometer to convergent scan based on the planned boundary convergent scanning path, and determines a lesion range and a lesion type of a convergent area according to near-infrared spectrum data obtained by the convergent scan; this is repeated until a new boundary convergent scanning path cannot be planned.

4. The brain injury identification system according to claim 3, wherein: the outward expansion path planning unit determines position coordinates of each scanning point corresponding to the outward expansion scan according to position coordinates of each scanning point of the near-infrared spectrometer on the scanning path, and plans the boundary outward expansion scanning path according to the position coordinates of all scanning points corresponding to the outward expansion scan; the convergent path planning unit determines position coordinates of each scanning point corresponding to the convergent scan according to position coordinates of each scanning point of the near-infrared spectrometer, and plans the boundary convergent scanning path according to the position coordinates of all scanning points corresponding to the convergent scan.

5. The brain injury identification system of claim 1, wherein The injury identification module comprises: a spectrum analysis unit configured to perform spectrum analysis on near-infrared spectrum data of each scanning point, and determine a characteristic absorption frequency band corresponding to each scanning point; a frequency band comparison unit configured to compare the characteristic absorption frequency band corresponding to each scanning point with a preset lesion near-infrared characteristic absorption frequency band respectively, and determine a lesion attribute and a lesion type corresponding to each scanning point. The damage identification unit is configured to determine a lesion range according to a scanning region corresponding to a scanning point with all positive lesion attributes, and determine a lesion type according to a lesion type corresponding to the scanning point with the positive lesion attribute.

Images