JP4699070B2 - Perimeter - Google Patents

Description

  The present invention relates to a perimeter that subjectively measures the visual field of a patient.

The visual field is measured by moving a target with constant brightness from the periphery toward the center, connecting the points where the target can be seen, and changing the size and brightness conditions of the target. There is a dynamic visual field measurement in which a visual field is measured by drawing an isosensitivity curve (hereinafter referred to as an isopter) similar to a contour line of the above (for example, Patent Document 1). Dynamic visual field measurement is suitable for knowing the whole change. In this dynamic visual field measurement, a measurement point is further added to measure an isolated dark spot (a portion where sensitivity is lost or sinks) existing between the isopters.
JP-A-6-54804

  However, the skill of the examiner is required to measure the isolated dark spot, and it is difficult for the inexperienced examiner to measure in a short time. Further, since conventional dark spot measurement is manual target presentation, it takes time for the skilled person to perform the examination, and the examiner and the subject are burdened.

  An object of the present invention is to provide a perimeter that can shorten the examination time and reduce the burden on the subject and the subject, in view of the problems of the conventional technology.

In order to solve the above problems, the present invention has the following configuration.
(1) The brightness of the target presented in the visual field of the eye to be examined can be changed, and the visual target presenting means capable of moving the presentation position of the visual target is provided. Search for isolated dark spots between adjacent isopters in a perimeter that moves from the periphery to the center and causes the subject to visually respond and measures the dynamic visual field to obtain the isopter connecting the same sensitivity points. Is a calculation control means for determining the starting point of the target presentation for identifying the depression of the isopter based on the shape of the isopter, and based on the position of the edge of the depression of the isopter and the curve on the valley side of the depression Computation control means is provided for determining the direction of the start point of the target presentation and determining the start point of the target presentation based on the inner and outer isopter intervals in the determined direction .
(2) In the perimeter of (1) , the target presentation for controlling the target presentation means based on the target presentation start point determined by the arithmetic control means and presenting a target for dark spot search Control means is provided .
(3) When the eye to be examined cannot visually recognize the dark spot search target presented at the start point, the target presentation control means of (2) moves the target under the presenting condition and moves the subject. An isolated dark spot region is determined by obtaining a response .

  According to the present invention, the examination time can be shortened, and the burden on the subject and the subject can be reduced.

  Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a schematic configuration diagram of a perimeter according to the present invention. A hemispherical dome-shaped screen 1 is arranged in front of the subject's eyes, and a stimulus target is presented from the target projection unit 2 on this screen 1. The target projection unit 2 includes a light source 3, a projection lens 4, a movable mirror 5, and an aperture 6 with a variable aperture diameter. The movable mirror 5 is driven by a driving mechanism (not shown) and changes the position of a spot target (stimulus target) projected on the screen 1 by the light source 3 and the projection lens 4. The movable mirror 5 can be composed of two sets of galvanometer mirrors. By changing the aperture diameter of the aperture 6, the size of the spot target projected on the screen 1 can be changed. A fixation target 10 is provided at the center of the screen 1. The target projection unit 2 is connected to the control unit 20. The control unit 20 drives and controls the movable mirror 5 of the target projection unit 2, and is projected onto the screen 1 to change the position of the stimulus target. Further, the light amount of the light source 3 is adjusted, and the luminance of the stimulus target projected on the screen 1 is changed. Connected to the control unit 20 are a monitor 21, an input device 22 such as a keyboard, a response switch 15, and the like.

  FIG. 2 is a diagram in which a test result obtained by dynamic visual field measurement is simulated. The isopter I displays a portion having the same sensitivity to the visual target of the eye to be examined as a contour line, and is also called an isosensitivity curve. The isopter I is plotted on the XY plane with the fixation target 10 with the subject's visual axis aligned with the origin O. In the isopter I plotting method, a target set to a predetermined brightness and spot size is moved from outside the subject's field of view toward the center of the field of view (the origin O). When the subject determines that the target is visible, he presses the switch 15. The isopter I is formed by interpolating the point where the target of the same condition is seen. The target presentation interval is set to 360 degrees at an interval of 15 degrees from the origin O, for example. Interpolation between each plot is performed by polynomial approximation with a curve function such as spline interpolation. The isopter is created from the peripheral visual field toward the central visual field. The luminance and spot size of the visual target depend on the symptom of the subject and the purpose of screening, but generally the luminance and spot size of the visual target are reduced as the created isopter approaches the center of the visual field. This is because the central part of the visual field has higher sensitivity and resolution than the peripheral part.

  In this way, the subjective dynamic visual field inspection in which the visual target is presented to the subject and the recognition of the visual target is responded by the switch 15 is automatically controlled by the control unit 20. The position of the target projected on the screen 1 from the target projection unit 2 is obtained from the angle of the movable mirror 5. The position of the visual target recognized by the subject is obtained by the angle of the movable mirror 5 at that time when a response from the switch 15 is sent to the control unit 20. A program for sequentially projecting the visual target on the screen 1 is stored in the control unit 20 in advance.

An interpolation function used when drawing an isopter will be described. Here, a cubic spline function related to θ is used. The function on the XY plane is expressed as a parameter by θ as follows.
xi (θ) = aiθ ³ + biθ ² + ciθ + di (section: i to i + 1)
yi (θ) = eiθ ³ + fiθ ² + giθ + hi (section: i to i + 1)
ai to hi are coefficients of each term calculated by the spline function in the sections i to i + 1. These values determine the characteristics of the spline function. This function is connected to form one isopter loop. Therefore, the position of a certain point on the isopter can be identified from the above function.

  In the isopter I created in this way, there is an isolated dark spot S, which is a portion where the sensitivity decreases between adjacent isopters. As a result of obtaining and examining dynamic visual field data of 15 eyes in which a dark spot S exists between the isopters, the dark spot S is sandwiched between a portion (concave part) where the isopter Iin is recessed inward (recessed part) Ho and the outer isopter Iout. It was found that a large number of areas W appeared.

  A method for specifying the depression (concave portion) of the isopter Iin will be described. FIG. 3 is an enlarged view of a certain isopter. 3A is mathematically obtained by searching the point P on the isopter counterclockwise (counterclockwise) from the 0 degree direction. A search is performed by moving the point P from the origin O at a constant angle.

FIG. 3B is an enlarged view of the portion A, and is one of two peaks (convex portions) sandwiching the depression. Now, let the points before and after the point Pi be Pi−1 and Pi + 1, respectively. With reference to the origin O, a vector from the point Pi-1 to the point Pi is a1, and a vector from the point Pi to the point Pi + 1 is a2. The outer product of the vectors a1 and a2 is calculated as follows.
a1 × a2> 0
In the part A corresponding to the mountain, since the vector a2 is on the right side of the vector a1, the sign of the value obtained by taking the outer product is positive.

FIG.3 (c) is an enlarged view of the part B, and shows a hollow. Similar to the method described above, the points before and after the point Pj are now denoted as Pj−1 and Pj + 1, respectively. With reference to the origin O, a vector from the point Pj−1 to the point Pj is b1, and a vector from the point Pj to the point Pj + 1 is b2. The outer product of the vectors b1 and b2 is calculated as follows.
b1 × b2 <0
In the portion B corresponding to the depression, the vector b2 is on the left side of the vector b1, so the sign of the value obtained by taking the outer product is negative.

  Thus, the position of the bottom of the depression (valley) or the top of the mountain can be determined from the sign of the result obtained by calculating the outer product of two adjacent vectors.

  In the above description, the continuous depression search process has been described. However, in implementation, the depression P is searched by discretely taking points P on the isopter formed by a polynomial function such as a cubic spline function. For example, the search is performed at an interval such that the origin O is the center and the vector on the isopter forms an angle of 0.1 degrees. The angle is not limited to 0.1 degree, and may be smaller than the angle for searching for the dark spot. For example, it may be once or twice. Therefore, it is not always possible to determine the bottom of the depression and the top of the mountain across the depression. In that case, the most likely point among the discrete search results, for example, the outer product component is the maximum point (minimum point). Alternatively, a calculation method may be used in which the polarity is obtained by the outer product and the point having the maximum angle after determining whether it is a mountain or a valley is the peak of the mountain or the bottom of the valley (described later).

In this embodiment, since the search interval of the point P is discrete with a constant angle from the origin O, the vector length is not constant. Therefore, even if the outer product of two adjacent vectors is the maximum value (minimum value), it does not necessarily indicate the top (bottom) of the mountain (valley). In this case, the calculation method of the top of the mountain or the bottom of the valley is determined by taking the inner product of two adjacent vectors and obtaining the angle, and determining the angle based on the magnitude of the angle and the polarity obtained by the outer product. If the angle formed by two adjacent vectors a1 and a2 is θi, the inner product is as follows.
a1 · a2 = | a1 || a2 | cosθi
Accordingly, the angle θi is obtained as follows.
θi = cos ⁻¹ (a1 · a2 / | a1 || a2 |)
The same applies to two adjacent vectors b1 and b2. In this way, the angle between two adjacent vectors can be obtained. At the top of the mountain (bottom of the valley), the angle is maximum in the vicinity, so it is determined whether it is a mountain or a valley according to the polarity obtained by the outer product, and then the angle of the part corresponding to the mountain or valley is obtained. The point that has the maximum angle in the vicinity is defined as the top of the mountain or the bottom of the valley. A valley sandwiched between two peaks (convex portions) becomes a recess (concave portion) of an isopter. The value of the outer product that determines the bottom of the valley or the angle obtained from the inner product is a parameter indicating the degree of depression.

  Note that the above algorithm also detects small depressions. Looking at the dynamic visual field data of 15 eyes, the dark spot S did not exist in a very small hollow part. Therefore, in order to avoid complication, a part that is not regarded as a hollow is an object of dark spot search. It is outside. For example, in the method of calculating the depression by setting a threshold for the determination of the outer product of the vector or the peak of the mountain (bottom of the valley) at an angle, if the outer product is equal to or less than a predetermined value, It shall not be regarded as a pit for searching.

  Also, in this embodiment, by obtaining the outer product and inner product of two adjacent vectors sandwiching the top portion of the isopter mountain and the bottom portion of the valley, the polarity of the mountain and valley is obtained, and the angle of that portion is obtained. However, it is not limited to this. The calculation of the angle is not a method for obtaining the inner product, but the portion where the radius of curvature at that point is the shortest may be the top of the mountain or the bottom of the valley.

  Next, a description will be given of a method for determining the starting point of the target presentation for searching for a dark spot between the isopter Iin depression obtained by the above-described method and the outer isopter Iout. FIG. 4 is a diagram for explaining a method of determining the start point of the dark spot search based on the depression. In FIG. 4, the points (edges of the depressions) on the tops of the two peaks sandwiching the depression are denoted by C and D, respectively. When the obtained dynamic visual field data of 15 eyes were examined in further detail, the location where the dark spot S exists was related to the interval between the points D and the valley side shape of the depression, and the isopter Iin of the depression and the outside thereof. It was found to be related to the interval of isopter Iout. Regarding the relationship between the dark spot S and the shape of the valley side of the depression, there is a vicinity where the curve (curvature) of the isopter becomes small on the valley valley side between the points C and D of the isopter Iin, and the center between the points C and D. , Found that there is much in the direction of the straight line passing through.

  Therefore, first, a method for determining the direction in which the dark spot S is searched in relation to the dark spot S and the curve shape on the valley side of the depression will be described. The line segment CD is divided at equal intervals, for example, once, and a perpendicular line is drawn to the isopter Iin from the line segment CD, and the intersection point is Ci (i = 1, 2,..., N). Let Hi be the straight line passing through Ci and Ci + 1, and let Vi be the normal line perpendicular to Hi at its midpoint. The distances from the points C and D to the normal line Vi are assumed to be diC and diD, respectively. Then, the point where the absolute value (| diC−diD |) of the difference between the distance diC and the distance diD is minimized is obtained. Let that point be o, and Hi and Vi at that time be the x-axis and y-axis, respectively. The coordinate system is replaced from the XY coordinate system to the xy coordinate system by the newly calculated x-axis and y-axis.

  Using this new xy coordinate system, the barycentric coordinates (xG, yG) of isolated dark spots in the dynamic visual field data of 15 eyes were obtained. As a result, xG was −1.54 to 9.9 °, the average ± standard deviation was 1.25 ° ± 2.78 °, and it was found that the center of gravity of the dark spot S was near the y-axis.

Next, the relationship between the center-of-gravity coordinates yG and the distance (interval: angle) between the isopters in the dynamic visual field data of 15 eyes was examined. FIG. 5 is a diagram showing the results. The horizontal axis represents the distance ymax between isopters on the y axis, and the vertical axis represents the coordinates yG. From the figure, it can be seen that the coordinate yG of the gravity center position is related to the distance ymax between isopters on the y-axis. That is, the value of the coordinate yG is a small value when the distance between the isopters is short, and a large value when the distance between the isopters is wide. The solid line graph in FIG. 5 shows a regression line for 15 yGs. The correlation coefficient between yG and ymax was 0.774, indicating a high correlation. The regression line yG is
yG = 0.5049 × ymax−2.0989
It became. The significance probability was 0.05 or less, and it was confirmed that the regression line equation was meaningful at a significance level of 0.05. From the above regression line, it was found that the center of gravity of the dark spot S is near the center of the distance ymax between the isopters. Therefore, it can be seen that the starting point of the dark spot search is preferably near the center of the distance ymax on the y-axis. Preferably, it is slightly on the isopter Iin side from the center of the distance ymax.

  The above y-axis is determined in relation to the shape of the valley side of the depression, but the line segment CD connecting the points C and D that become the peak (edge) of the depression is the x-axis direction, and the midpoint of the line segment CD It is also possible to use a determination method in which the y-axis is the direction orthogonal to the line segment CD with the origin as the origin. Also in this case, there was a dark spot almost on the y-axis.

  Next, an algorithm for searching for a dark spot from the start point will be described. FIG. 6 is a diagram showing a procedure for starting a dark spot search in the depression. Let G and F be the intersections of the perpendiculars drawn from points C and D to the x-axis (straight line Hi) and Iout, respectively. Now, a dark spot search is performed in an area surrounded by CDFG. The control unit 20 determines the dark spot search point Si (i = 1, 2,..., N) at the position determined by the regression line on the y axis. Here, the point where the dark spot is first searched is S1. The control unit 20 presents a spot light target in S1, and waits for a response from the subject. The brightness and size of the spot light are the same as when the isopter Iout is determined, or brighter than the condition of the isopter Iin, and the spot size is increased. If there is no response within a predetermined time (if it cannot be seen), the dark spot area is found by the following procedure.

  As shown in FIG. 7, the control unit 20 moves the spot light radially from the point S1 toward the periphery. First, the spot light is moved from the center toward a specific direction to find a point where there is a response from the subject and recorded. Again, the spot light is moved in the other direction from the center, and the response of the subject is waited. The point to which the subject responds (the point that can be seen) becomes the boundary (region) of the dark spot. This process is repeated, and the response points of the subject formed so as to surround the point S1 are interpolated with a cubic spline function. The inside of the loop formed by the curve is the dark spot area. For example, eight directions are used when the spot light is moved radially. However, the present invention is not limited to this, and if it is desired to find a dark spot region with high accuracy, a direction smaller than 8 directions may be taken. After the dark spot area is calculated, if there is still room for search in the dark spot search area CDFG, a process for searching for the next dark spot is started (described later).

  First, when the point S1 is presented and a response is received from the subject (seen), the spot light is presented at a point S2 that is slightly shifted from the point S1 to find the dark spot. At this time, the point S2 is presented at a point moved by a predetermined distance on the y-axis from the point S1 toward Iin (field center). If no dark spot is found at point S2, the dark spot search is continued with points S3, S4,. The dark spot search point at this time is directed toward the Iin side, but if a dark spot is not yet found even after reaching Iin, a search is made from the point S1 on the y axis in the Iout direction. If not, spot light moved by a predetermined angle in the x-axis direction from the point S1 is presented. The search rule in the x-axis direction is performed spirally with the point S1 as the center. In this process, when there is no response from the subject (not visible), the process of scanning the dark spot region is started by the above-described procedure.

  As explained, the spot light is presented, the following processing is sequentially performed from the response of the subject, and when the spot light is presented to the dark spot search point Sn, the dark spot search in the region CDFG is completed. Shall be. If there is another depression that has not been searched for a dark spot, the dark spot is searched similarly by a series of algorithms. In that case, the size and brightness of the spot light are changed. The reason for searching in the Iin direction on the y-axis is that there is a high probability that a dark spot exists. The predetermined distance mentioned here indicates a search interval at the time of dark spot search, and an angle of 1 to 4 degrees is preferable.

  In this embodiment, the dark spot search is performed from the points S1 to Iin after performing the dark spot search from the points S1 to Iin on the y-axis. However, the present invention is not limited to this. You may search from the point close | similar on the y-axis to each direction of Iin and Iout from point S1. Moreover, you may search helically from point S1.

  Next, the search interval for dark spot search will be described. FIG. 8 is a diagram showing the interval between dark spot search points in a certain search range. As shown in the figure, dark spot search candidates are secured in a grid pattern at predetermined intervals. A dark spot is searched from this candidate using the algorithm described above. When search points are sequentially determined by these algorithms, dark spots and dark spot areas can be found in a short time.

  FIG. 9 is a diagram showing a dark spot search point after the dark spot area is found. As shown in the figure, candidate points for the next dark spot search are plotted around the dark spot area LS determined by the dark spot search process. In the room portion obtained by subtracting the dark spot area LS from the search area CDFG, a point separated from the boundary of the dark spot area LS by a predetermined width is plotted as a candidate point. The probability that another dark spot area exists empirically around the already determined dark spot area LS is considered to be low. Therefore, re-searching around the dark spot region LS having a predetermined width is avoided. The predetermined width is a value determined under the same condition as the dark spot search interval described above. This eliminates the need for searches that are not very valuable. Furthermore, if the dark spot area is determined at an early stage of the examination for searching for a dark spot, the visual field examination is completed in a short time, and the burden on the subject is reduced.

  FIG. 10 is a diagram showing dark spot search points when there are two dark spot areas. Since two dark spots are defined, it is understood that the vicinity of the dark spot areas LS1 and LS2 need not be searched again.

  It should be noted that the dark spot search area is defined as G, F at the intersections of perpendicular lines extending from points C, D and the x-axis (straight line Hi) and Iout, respectively, and surrounded by CDFG formed there. This is not a limitation. A dark spot search may be performed in a region surrounded by a total of four intersections of straight lines drawn from the origin O of the XY plane and the origin o of the xy plane to the points C and D and Iout. In the embodiments described above, all point interpolation is performed by a cubic spline function. However, the present invention is not limited to this. The spline function may use a quadratic function or may be a quadratic or higher order function. A Lagrangian function or the like may also be used. Furthermore, each interpolation may be performed under different conditions. For example, the isopter may be calculated by a quartic spline function, and the dark spot area may be calculated by a cubic spline function.

  A series of dark spot search processing of the present embodiment described above is shown in a flowchart. FIG. 11 is a flowchart of dark spot search. There are four major steps. An isopter mapping step S1-i for creating a patient's isopter, a depression calculation step S2-i for calculating a depression in the isopter, and a dark spot search start point determination step S3 for calculating a starting point for searching for a dark spot in the obtained depression. -I, a dark spot area detection step S4-i for detecting a dark spot in the search area from the starting point of the dark spot search, where i = 1, 2,..., N.

  First, step S1-i for creating an isopter will be described. When the field of view and dark spot measurement is started, the apparatus enters a step of creating an isopter. In S1-1, conditions for the spot size and brightness of the presentation target when creating an isopter are set. In S1-2, the target is moved 360 degrees over a predetermined angle, for example, 15 degrees from the peripheral part of the visual field to the central part of the visual field, and the points to which the subject responds are plotted. In S1-3, the points plotted in S1-2 are interpolated with a spline function or the like to create an isopter. In S1-4, it is determined whether or not there is another isopter to be created. If there is a process, the same steps are performed from S1-1. If not, all necessary isopters have been completed. Proceed to

  In step S2-i, an isopter depression portion is calculated, and a portion in which the existence probability of a dark spot is empirically low is excluded. In S2-1, the isopter is searched for a dent counterclockwise from the 0 degree direction. In S2-2, a vector at each point on the isopter is calculated, and an outer product and an angle of two adjacent vectors are collectively calculated for one round of the isopter. In S2-3, the outer product for one round of the obtained isopter having a negative value and an absolute value equal to or less than a threshold value is excluded from candidates for the subsequent depression calculation. Here, the threshold value is a value that divides a portion that is not regarded as the above-described depression having a dark spot. In S2-4, processing is performed on the depressions that are not excluded in S2-3. A search is made for a point where the outer product is positive and the angle at that time is maximum in the left-right direction from the negative value to the positive direction. In S2-5, each point obtained in the previous step is determined as a mountain sandwiching the depression. In S2-6, it is determined whether there is any other isopter that is being searched for and whose outer product is negative and whose absolute value is greater than or equal to the threshold value. If so, the process returns to step S2-3, and if not, the process proceeds to the next step. In S2-7, it is determined whether there is an isopter for which no depression has been calculated, and if there is a return to step S2-1, a search for a depression is performed with a different isopter, and if not, all the isopter depressions have been calculated. So go to the next step.

  In step S3-i, processing for determining a starting point for dark spot search in the depression calculated in step S2-i is performed. In S3-1, an isopter with a depression calculated in step S2-i is selected from the peripheral side. In S3-2, a tangent and a normal to the tangent are calculated at each point from one end to the other end on a depression sandwiched between two peaks. In S3-3, a perpendicular is drawn from each vertex of two peaks sandwiching the depression to each normal obtained in S3-2, and a ratio of the two perpendiculars is calculated. In S3-4, the minimum point is obtained by the ratio of the perpendiculars calculated at the respective points, and the tangent and normal at that point are set as new coordinate x-axis and y-axis. In S3-5, an isopter surrounded by a total of four points, a point where the extension of the normal line drawn from the y-axis of the new coordinate to each vertex of the mountain intersects with the outer isopter and each vertex of the mountain Set the space to search for dark spots. In S3-6, a predetermined point on the y-axis where the section is divided between the isopters is set as the starting point of the dark spot search.

  In step S4-i, a dark spot is searched with reference to the dark spot search start point, and a process of calculating a dark spot area is performed. In S4-1, a target suitable for an isopter for searching for a dark spot is set, and the target is presented at the dark spot search start point obtained in S3-6. In S4-2, the patient's response is awaited. If there is, the process proceeds to step S4-6, and if not, the process proceeds to step S4-3. In S4-3, the target is scanned radially from the target presentation position, and the place where the subject has responded is set as the boundary point of the dark spot. In S4-4, the boundary point of the dark spot is interpolated, and the range is set as the dark spot area. In S4-5, the next candidate point for the dark spot search is calculated. In S4-6, if the calculated candidate point is next, the process proceeds to step S4-7, and if not, the process proceeds to step S4-8. In S4-7, the target is presented from the next candidate point in the vicinity of the dark spot search start point according to the above-described rules, and the process returns to the step of S4-2 waiting for the subject's response. In S4-8, it is determined whether there is another isopter that is not searching for a dark spot. If it is, the process returns to step S3-1. If not, all dark spots in the visual field have been searched. finish.

It is a schematic block diagram of the perimeter of this invention. It is the figure which drew the test result obtained by the dynamic visual field measurement in a simulated manner. It is the figure which expanded a certain isopter. It is a figure explaining the method of determining the starting point of a dark spot search based on a hollow. It is a figure which shows the relationship between the gravity center coordinate yG in the dynamic visual field data of 15 eyes, and the distance between isopters. It is the figure which showed the procedure which starts a dark spot search in a hollow. It is a figure which shows the procedure which determines a dark spot area | region. It is the figure which showed the space | interval of a dark spot search point in a certain search range. It is a figure which shows the dark spot search point after a dark spot area | region is discovered. It is a figure which shows the dark spot search point when there are two dark spot areas. It is a flowchart figure which shows the dark spot search process of a present Example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Screen 2 Target projection unit 3 Light source 4 Projection lens 5 Movable mirror 6 Aperture 10 Fixation target 15 Switch 20 Control unit I, Iin, Iout Isopter LS, LS1, LS2 Dark spot area

Claims (3)

The brightness of the target to be displayed in the visual field of the eye to be examined can be changed, and the visual target presenting means capable of moving the target presentation position is provided so that the target with a certain brightness is centered from the periphery of the visual field of the eye to be examined. In order to search for isolated dark spots between adjacent isopters in a perimeter that obtains an isopter that connects the points with the same sensitivity by measuring the dynamic visual field by moving the subject toward the visual response. Computational control means for determining the starting point of the target presentation, which identifies the depression of the isopter based on the shape of the isopter, and presents the target based on the position of the edge of the depression of the isopter and the curve on the valley side of the depression And a calculation control means for determining a start point of the target presentation based on the inner and outer isopter intervals in the determined direction . 2. The perimeter of claim 1, wherein the target presentation control means controls the target presentation means based on a start point of the target presentation determined by the arithmetic control means, and presents a target for dark spot search. perimeter, characterized in that it comprises. When the eye to be examined cannot visually recognize the target for dark spot search presented at the start point, the target presentation control means of claim 2 moves the target under the presenting condition to obtain the response of the subject. A perimeter that determines the area of an isolated dark spot .