Relative calibration for dosimetric devices

This application was accepted on May 27, 2003, under the name "Radiation Detection Medium Calibration System or Method," and was accepted on September 22, 2000, US Provisional Applications 60 / 234,745, and 2000. A continuation of US Patent No. 6,675,116, which was granted on June 1, 2001, claiming priority to US Provisional Application 60 / 252,705, which was accepted on 22 November, and is currently pending US Patent Application 10 Related to / 445,587. This application also relates to the application of US Pat. No. 6,528,803, which was accepted on January 21, 2000. This application was also accepted on December 10, 2004, under the name "Image Arrangement Optimization," a pending US patent application, and on July 30, 2003, "Image Arrangement System and. In connection with the pending US patent application 10 / 630,015 under the name "Method". All of the above related applications are incorporated herein by reference.

The present invention relates to methods and devices for efficiently performing radiation dosimetry, more specifically radiation therapy related radiation dose calibration.

An important use of radiation therapy, especially intensity-modulated radiation therapy (IMRT), is the destruction of tumor cells. In the case of ionizing radiation, tumor destruction depends on the "absorbed dose", the amount of energy stored in the tissue mass. Radiation physicists usually represent absorbed doses in cGy units or centimeter gray. 1cGy is equal to 0.01J / kg.

Radiation dosimetry usually describes a method of measuring or predicting the dose absorbed by various tissues of a patient undergoing radiation therapy. Accuracy in predicting and measuring absorbed doses is important for effective treatment and prevention of complexity due to excessive or under-radiation exposure. Many methods exist to measure and predict absorbed doses, but many rely on the development of calibration-calibration curves, look-up tables, equations, etc. for the response of the detection medium to absorbed doses. Effective detection media are known to those of skill in the art and include radiation sensitive films and three-dimensional gels that darken or discolor upon radiation exposure (eg, "BANG" and "BANANA" gels). Other effective detection media include electronic portable imaging devices, computer radiation (CR) devices, digital radiation (DR) devices, and amorphous silicon detector sequences that generate signals in response to radiation exposure.

There are various known methods of creating calibration curves. For example, Patent Document 1, granted to the applicant of the present application and fully incorporated herein by reference, provides a detection medium that responds to exposure to ionizing radiation and sets a predetermined region of the detection medium to a different ionizing radiation. Disclose that creating a calibrated dose response pattern by exposure to dose levels. Patent Document 1 further measures the response of the detection medium in a predetermined region to generate a calibration associated with the next response to an ionized radiation dose. Different dose levels are obtained, for example, by selectively shielding a portion of the detection medium from ionizing radiation using a multi-leaf collimator, secondary collimator, or attenuation block. Different dose levels can also be obtained by moving the detection medium during exposure. Patent Document 1 further discloses a software routine immobilized on a computer-readable medium configured to generate a calibration of the detection medium's response to an ionized radiation dose.<patcit num="1"><text>U.S. Pat. No. 6,675,116</text></patcit>

<p> Methods such as those disclosed in U.S. Pat. No. 6,675,116 use linear accelerators or similar devices to expose separate parts of the detection medium to different known radiation doses to create calibration curves and look-up tables. Need that. Typically twelve, often as many as 25, different radiation dose levels are measured to create calibration curves and look-up tables. Usually, as the number of radiation dose levels measured increases, the accuracy of calibration increases. However, as the number of measurements increases, the calibration process becomes more costly and time consuming. Therefore, it would be desirable to have a system and method that provides calibration information by applying the radiation therapy program to a quality assurance device and analyzing the "acquired image" obtained by capturing the radiation intensity distribution.</p><p> There is known a method of modifying the acquired image so that the once calibrated dosimetry acquisition system does not need to be recalibrated. For example, U.S. Pat. No. 6,528,803, granted to the applicant of this application and fully incorporated herein by reference, exposes a portion of the test film to an array of standard illuminants, followed by one of a set of calibration films. Or obtain an optical density step gradient comparable to all corresponding optical density step gradients. However, existing methods, such as those disclosed in US Pat. No. 6,528,803, require additional equipment and time to collect data on optical density step gradients. In some cases, when it is only a "relative" calibration (displayed as a percentage) and not an absolute calibration (dose and traces can be national standards), no special equipment is required and minimal time is required. In short, it would be desirable to have a system and method that provides calibration information for the next acquired image.</p><p> In addition, it is generated by a system that evaluates the characteristics of the experimentally obtained calibration curve and creates a treatment plan to determine where the difference occurs by modeling the inaccuracy as for the true experimental difference. It may also be desirable to have a system that models the dose distribution to be performed and other predictions of the dose distribution.</p>

<p> According to the embodiment, the system that calibrates the dose response of the image acquisition device is a means of creating a dose map showing the dose provided in the treatment plan, acquisition including an indication of the dose intensity recorded from the application of the treatment plan. A means for generating an image and a means for generating a self-calibration curve that associates the donated dose with the donated dose intensity are provided.</p><p> Further, according to the embodiment, the system for calibrating the dose response of the image acquisition device includes a first self-calibration curve, a second self-calibration curve, an initial calibration, and (1) a first self-calibration curve. Determine at least one difference or fit with the second self-calibration curve, (2) modify the acquired image based on the at least one difference or fit, and (3) apply the initial calibration to the acquired image. With a means of performing calculations, including generating relative calibrations based on.</p><p> Further, according to the embodiment, the method of calibrating the dose response of the image acquisition device compares the first self-calibration curve with the second self-calibration curve to determine the relationship between these curves and makes at least one difference. Based on that the acquired image is modified and the initial calibration is applied to the acquired image, which calibrates the dose response of the image acquisition device.</p>

For the first treatment plan, a calibration is created that associates the planned dose with the response of the detection medium to the absorbed dose. Also, an IMRT self-calibration curve (ISCC or ISCC curve) that correlates the dose intensity delivered in the first treatment plan with the pixel strength of the acquired image is created for the first treatment plan. The ISCC curve is then created for the second treatment plan. Comparing the ISCC curve for the first treatment plan with the ISCC curve for the second treatment plan relates to the second treatment plan so that calibration can be used to provide calibration information for the second treatment plan. Allows adjustment of the acquired image. Therefore, the systems and methods disclosed herein provide a simple, fast, efficient, and cost-effective calibration of the imaging device used to capture test images before the treatment plan is applied to the patient. To do.

The use of the various components of System 100 shown in FIG. 1 is described in detail below. Usually, the treatment plan is applicable to a radiation detector and is therefore recordable on a given medium or device that produces the acquired image 112, which is shown and described more fully with reference to FIG. .. The acquired image 112 showing the intensity of the dose delivered as part of the procedure is compared to the dose map 106, which is more fully described below with reference to FIG.

System overview FIG. 1 provides an overview of System 100 used in at least one embodiment to create an IMRT self-calibration curve (ISCC). The treatment planning system 102 is any of the various treatment planning systems known to those of skill in the art, the Pinnacle 3 system manufactured by Phillips Medical Systems in Andover, Massachusetts, manufactured by Brainlab AG in Heimstetten, Germany. BrainSCAN, PLATO SunRise by Nucletron in Veenendaal, Netherlands, and Eclipse manufactured by Varian Medical Systems in Palo Alto, California, but not limited to.

The treatment planning system 102 is used to create one or more radiation treatment plans 104. The treatment planning system 102 is also used to create a dose map 106, which is sometimes referred to as a planning image. Such use of treatment planning system 102 will be well known to those of skill in the art. In addition, one of ordinary skill in the art will recognize that dose map 106 provides a predicted distribution of planned radiation doses in quality assurance phantoms or patients. An example of dose map 106 is shown in Figure 6A.

The radiation detector 108 is a device capable of detecting and receiving radiation as known to those skilled in the art. In some embodiments, the radiation detector 108 is a quality assurance phantom, also known as a test phantom, as known to those of skill in the art. The purpose of the test phantom is to simulate a medium that should receive a radiation dose, such as human tissue.

The image acquisition device 110 can be a device or medium for recording the detected radiation, as known to those skilled in the art, such as radiographic film, computer radiographers, electronic portable imaging devices, charge coupling elements ( Contains, but is not limited to, CCD) cameras, or BANG gels. The image acquisition device 110 generates one or more acquired images 112. As described below, the acquired image 112 and the dose map 106 are used to create the ISCC curve 114. An example of the acquired image 112 is shown in FIG. 6B. Further, as described below, most embodiments create at least two ISCC curves 114 for at least two treatment plans 104.

Those skilled in the art perform the processes described herein with reference to System 100 by using one or more computers capable of functioning as described herein with respect to System 100, as known to those skilled in the art. It will be recognized that it is possible and may include any device or combination of devices that receives, outputs, processes, transforms, incorporates, and / or stores information. Therefore, the process described herein can be performed by executing a computer-executable command incorporated in a computer-readable medium. For example, the computer used in System 100 can be a general purpose computer capable of running a wide variety of different software applications. Further, such a computer may be a special device limited to a specific function. In some embodiments, the computer is a network of computers. Generally, the system 100 can incorporate a wide variety of different information technology configurations. Computers are not limited to any type, number, form, or configuration of processors, memory, computer-readable media, peripherals, computing units, and / or operating systems.

In addition, some of the elements of System 100 may exist as displays in the computer. For example, a treatment plan 104, a dose map 106, an acquired image 112, and / or an ISCC curve 114 can exist as one or more in-computer displays. Thus, the computer can include, or can connect to, an interface and an access device that provides the user (eg, a radiologist) with access to the system 100. As such, users can access the processes and elements of System 100 using any access device or interface known to those of skill in the art.

Initial calibration process FIG. 2 describes a processing flow for the initial calibration process. Step 200 represents the process of acquiring the first acquired image 112a representing the radiation distribution from the first radiation treatment plan 104a. Shown in step 200 to process it is described in detail with reference to Figure 3A. Step 202 represents a calibration that associates the radiation intensity distribution of the acquired image 112a with the radiation dose provided by the application of treatment plan 104a, i.e. the process of creating a calibration curve or equation. As mentioned above, various means, methods, and devices for performing the calibration of step 202 will be known to those of skill in the art. Step 204 represents the process of generating the ISCC curve 114. The process shown in step 204 will be described in detail with reference to FIG. 3B.

Image acquisition process FIG. 3A is a flow diagram illustrating the processing of the acquired image that should be used to create the ISCC curve.

In step 300, a radiation treatment plan 104 is created. The creation of radiation treatment plans is well known and can be achieved using various known treatment planning systems 102. As is well known, the radiation treatment plan may include the intensity, duration, and location of the radiation dose applied to the tumor site during ongoing radiation therapy.

In step 302, the treatment planning system 102 is used to create a dose map 106 corresponding to the treatment plan 104, which is optionally referred to as a planning image.

In step 304, the treatment plan 104 is applied to the radiation detector 108. The radiation distribution of the detected radiation is recorded in the image acquisition device 110. The image acquisition device 110 is used to generate an acquisition image 112 representing the radiation distribution created from the application of treatment plan 104. Those skilled in the art will recognize that the acquired image 112 can be generated in a variety of ways. For example, 112 acquired images shown in FIG. 6B represent a scanned digital image of a quality assurance film in the IMRT treated area. In some embodiments, the acquired image 112 can be filtered using a 5x5 median filter or some other filtering technique that may be known to those of skill in the art. Filtering can be used to reduce noise and / or to adjust pixels or elements between the treatment plan and the image acquisition device.

ISCC generation process The ISCC generation process will be described with reference to FIG. 3B. The use of the ISCC generation process is described herein for certain embodiments, but it should be understood that the process is applicable to other embodiments that will be apparent to those skilled in the art upon reading this disclosure. .. In step 306, the acquired image 112 is registered in the dose map 106. Image registration refers to the process of arranging images so that they occupy the same image space and can be compared and / or combined. Various methods and devices for registering images are known to those of skill in the art, some of which were accepted in the pending US applications 10 / 630,015 and December 10, 2004, "Optimization of Image Arrangement". It is explained in a US application named.

In step 308, a common region of interest (ROI) is acquired for the dose map 106 and the acquired image 112. The common ROI can never be greater than the smaller of the dose map 106 and the acquired image 112 and is selected to exclude any special non-dose related markings of the acquired image 112. For example, the ROI should not contain any description or reference markings and should not include any area away from the edges of the dose map 106 or acquired image 112.

As part of step 308, various automated techniques known to those of skill in the art exclude anomalous small areas of dose map 106 and acquired image 112, such as areas where one image contains needlestick marks. Can be adopted to do. In addition, various known thresholding techniques can be employed to exclude regions that are considered to have low correlation for the selected dose range. These regions include low-dose regions, high-gradient regions, regions near the physical medium boundary, and the like.

In step 310, the values in the dose map 106 are standardized for the maximum pixel values in the dose map 106 and converted to percentage values if "relative" measurements are desired instead of "absolute" measurements. Will be done.

In the first embodiment, in step 312, the dose map 106 is divided into smaller geometric areas. An example of dose map 106 divided into smaller geometric regions (and after statistical functions have been applied to the regions, as described below) is shown in Figure 7A. In one embodiment, the geometric regions are rectangles (cubes for 3D images), each containing 1% of the region of the dose map 106. It should be noted that the geometric regions of the image may or may not be adjacent and may or may not physically overlap depending on the particular image adopted.

In the second embodiment, in step 312, the dose levels in the dose map 106 are divided into dose ranges. These dose ranges may or may not be adjacent and may or may not overlap. In one embodiment, each dose range spans 1% of the total dose range of the planned image. That is, each 1% increment extends from 0 to maximum dose in dose map 106 on a scale of 0 to 100. Statistical measures such as mean and median are calculated for each range of pixels in the dose curve. That is, the process finds the planned image, i.e. all pixels in each dose range of the dose map 106, and takes the average of those pixels (or some other measure of median or median tendency). The index placed on the pixels of the registered image in each range is maintained.

Further in a third embodiment, in step 312, the dose map 106 is divided into sub-regions having a percentage of the region of the dose map 106, such as the geometric region described above. The dose map 106 is then transformed into what is referred to as the "column format". The use of columnar format is adopted by choice for the purpose of simplifying the process, but the steps following step 312 and step 312 can be performed without displaying the columnar dose map 106 and the acquired image 112. Should be understood. An example of a columnar dose map 106 is shown in Figure 9A. Each subregion of the dose map 106 shown in FIG. 6A is represented in each column of the image shown in FIG. 9A.

In the first embodiment described with reference to step 312, in step 314, the pixels are defined as described above for step 312 and are placed on the acquired image 112 corresponding to each geometric region of the dose map 106. .. The dose map 106 and the acquired image 112 can be selectively excised before the corresponding pixels are arranged so that each of the dose map 106 and the acquired image 112 has a region obtained by adding the total number of each geometric region. Is. The acquired image 112 divided into geometric areas is shown in FIG. 7B. Several statistical scales or characteristics are calculated for each set of pixels in such an arrangement. In some embodiments, an average pixel intensity is calculated for each set of arranged pixels, as depicted, for example, in FIGS. 7A and 7B. Other embodiments can calculate some other statistical properties as known to the median or those skilled in the art. For example, median values preserve the edges of an image, while averaging tends to smooth the edges. The choice of statistical properties may depend on several factors that can include the rate at which the dose changes within the region.

In a second embodiment described with reference to step 312, in step 314, the pixels are placed on the acquired image 112 corresponding to the dose level identified in the dose map 106 as described above for step 312. A statistical scale for each dose range of the acquired image 112 (eg, mean, median, etc.) is then calculated for the dose map 106 in step 312 as described above.

In the third embodiment described above with reference to step 312, in step 314 the acquired image 112 is divided into sub-regions and then displayed in a columnar format as shown in FIG. 9B. For each such sub-region, a correlation is taken between the reference image, ie the pixels of the dose map 106, and the pixels of the corresponding geometric region of the acquired image 112. Such correlations are known to those of skill in the art. For example, FIG. 10 shows a graph representing the correlation coefficient, as known to those of skill in the art, for each corresponding set of columns (numbered 1 to 100) of the images shown in FIGS. 9A and 9B. The corresponding sub-regions represented by the corresponding columns are ranked in the order of the correlation scale. FIGS. 11A and 11B show columnar images showing the dose map 106 and the acquired image 112, respectively, in which the columns are sorted according to the value of the correlation coefficient. Starting from the region with the highest correlation, the corresponding pixels are used to create the calibration curve in a manner similar to that described in the previous paragraph.

In some embodiments, a correlation threshold is generated such that only columns whose correlation exceeds a predetermined threshold are considered when creating the ISCC curve. For example, with reference to FIGS. 11A and 11B, columns 1-33 of the column image have a correlation less than or equal to -0.97, and a value of -0.97 is selected as the correlation threshold. Therefore, in this example, only columns 1-33 should be considered when generating the ISCC curve.

In step 316, the original ISCC curve is created. An example of an ISCC curve created in one performed embodiment is shown in Figure 4. The ISCC curve of this embodiment is the dose of the treatment plan 104 for each of the ranges of the dose map 106 defined in step 312 with respect to the value representing the pixel intensity of each of the corresponding pixels of the acquired image 112. An intensity value is illustrated, which is associated with any statistical scale selected in step 314. Another example of the original ISCC curve is provided in Figure 8A. The original ISCC curve shown in FIG. 8A was created by dividing the dose map 106 shown in FIGS. 6A and 7A and the acquired image shown in FIGS. 6B and 7B into smaller geometric regions. FIG. 12A shows the original ISCC curve based on a plot of pixel values for the region of dose map 106 and acquired image 112 that exceeds the correlation threshold described above for step 314.

In step 318, the ISCC curve created in step 316 is post-processed to ensure that the pixel values decrease monotonically as the dose value increases. In addition, other post-treatments known to those of skill in the art are applicable. For example, a technique for smoothing or fitting ISCC curves is applicable in step 318. FIG. 8B shows the ISCC curve generated by the post-processing of the original ISCC curve shown in FIG. 8A. FIG. 12B shows the ISCC curve generated by the post-processing of the original ISCC curve shown in FIG. 12A.

Next calibration process FIG. 5 illustrates the following calibration process, i.e., relative calibration performed for treatment plan 104b other than treatment plan 104a for which the calibration curve was created in step 202 of the initial calibration process described above with reference to FIG. Treatment plan 104b is referred to as a second or next treatment plan.

In step 502, the image acquisition process is performed for the next treatment plan 104b. Step 502 includes performing the steps described with reference to FIG. 3A for treatment plan 104b. Therefore, step 502 produces the dose map 106b and the acquired image 112b.

In step 504, the ISCC generation process described above with reference to FIG. 3B is performed for the dose map 106b and the acquired image 112b. Therefore, step 504 produces the following ISCC curve 114b.

At step 506, a comparison is performed between the first ISCC curve 114a and the next ISCC curve 114b to identify the difference between the two curves or to fit them. In step 508, the next acquired image 112b is modified based on the difference or fit identified between the first ISCC curve 114a and the next ISCC curve 114b. The purpose of this change is to transform the acquired image 112b into a calibable state using the calibration curve created in step 202 described above with reference to FIG. The conversion of captured image 112b can be performed using a variety of methods known to those of skill in the art, including those described, for example, in US Pat. No. 6,528,803. The relationship between two curves or two sets of points can easily be a difference, but can be more complex and can take the form of a look-up table or a curve fit as known to those of skill in the art. it can.

At step 510, the calibration curve created in step 202 described above with reference to FIG. 2 is applied to the acquired image 112b.

Evaluation of experimental calibration The ISCC curves created as described above can be advantageously used to evaluate the effectiveness of experimentally obtained calibration curves using methods known to those of skill in the art. Therefore, in some embodiments, the ISCC curve is comparable to an experimentally obtained or calculated calibration curve. The correlation or correspondence between the ISCC and the experimentally obtained curve is represented by the dose map 106 and is a measure by which the dose distribution as shown in the acquired image 112 can be well modeled. Those skilled in the art will recognize that it is possible to set an acceptable threshold for experimentally obtained calibration curves so that curves with excessive error are not used. In addition, in order to determine whether the deviation between the dose map 106 and the acquired image 112 is due to calibration error, TPS modeling error or radiation application error, the user can determine whether the deviation between the dose map 106 and the acquired image 112 is due to the ISCC curve and the experimentally obtained calibration curve. Correspondence is available.

Evaluation and selection of standardized values The ISCC curves created as described above can also be advantageously used to evaluate and select standardized values for images such as acquired images 112 and dose maps 106 to be compared for quality assurance purposes. is there. Systems and methods for relative dosimetry, such as those newly disclosed herein, typically require standardization of pixel values in plans and acquired images that are arranged over similar ranges. Choosing these standardized values can often be difficult. For example, if the experimental calibration and the ISCC curve are different in shape, optimizing the standardization at one dose level can reduce agreement at other dose levels. Changing the standardized value in the planned image, ie dose map 106, is replaced by the ISCC curve generated between the standardized planned image and the acquired image 112. Therefore, the match between the ISCC curve and the experimentally obtained curve is optimized over the entire curve or by changing the standardized values for selected ranges or points of the curve. In this way, the optimized standardized values can be generated for different criteria.

FIG. 13 shows an experimentally obtained calibration curve 1310, i.e. a graph 1300 containing a calibration curve created according to a method known to those of skill in the art or the ISCC generation process newly disclosed above. In addition, FIG. 13 shows the ISCC curve 1320 created by dividing the dose map 106 and the acquired image 112 into the dose ranges as described above with respect to FIG. There are numerous ways to compare and standardize curves 1310, 1320. For example, one way to compare curves 1310, 1320 is to evaluate the difference between these curves. For the purposes of this embodiment, curves 1310, 1320 are evaluated for a selected range of pixel values, usually for a common set of pixel values that have a common pixel value range for the two curves 1310, 1320. .. In this case, each pixel value in the selected range is linearly interpolated, but those skilled in the art will recognize that there are various methods in which this interpolation can be performed.

Focusing on the dose difference between the two curves 1310 and 1320 illustrated in the graph shown in FIG. 14, when using the experimental calibration curve 1310 to calibrate the acquired image 112 from pixel value to dose level, Differences are discernible, and the acquired image 112 can be compared to dose map 106 for quality assurance purposes, overresponse in the low dose region (0-10 cGy) of calibration curve 1310, and calibration curve 1310. Under-responses in the range of 10-30 cGy will be recognizable. Focusing on the plot shown in FIG. 14 in combination with the plot shown in FIG. 13, conclude that additional experimental calibration is required for points in the 0-10 cGy range to match the ISCC curve 1320 well. Will be able to. After performing additional experimental calibration, the analysis described with respect to FIGS. 13 and 14 can be repeated to see if the difference between ISCC curve 1320 and experimental calibration curve 1310 has improved.

Instead of focusing on a particular dose region, curves 1310, 1320 are comparable in their entirety, as described for FIG. For example, the correlation between the two curves can be evaluated, or the root mean square (RMS) of the difference can be calculated. Tolerance thresholds can be set for these parameters to accept or reject the experimental curve 1310 for use in calibrating dosimetry devices. Instead, the portion of curve 1310 can be evaluated independently, especially if certain regions of the curve (eg, high dose regions) are of interest. When curve fitting is used, this is referred to as spline fitting. For example, curve 1310 can be divided into 10 equivalence parts, and correlations such as those shown in FIG. 15A or RMS-like statistics such as those shown in FIG. 15B are calculated for each.

Moreover, as mentioned above, the standardized values are often applied to adjust the calibration curve to reduce the system error between the images being compared, i.e. the dose map 106 and the acquired image 112. For example, the ISCC curve 1320 and the experimental curve 1310 can be standardized for their maximum values. The experimental dose curve 1310 can then be adjusted by a range of factors, and the RMS of the experimental curve 1310 adjusted against the ISCC curve 1320 can be plotted. Then, by evaluating the minimum point of this plot, the optimum standardization factor can be determined. It is noted that this technique is feasible for a portion of curve 1310 and allows high dose optimization, eg, as shown in FIG. The minimum point of the curve shown in FIG. 16 appears to be approximately 1.01, suggesting that 1.01 is the optimal standardization factor for the region shown in FIG. Those skilled in the art will recognize that this factor can be further improved by fitting the curve to a polynomial.

The above description is intended to be graphic and not restrictive. Numerous embodiments and uses other than the examples provided will be apparent to those skilled in the art by reading the above description. The scope of the invention should be determined without reference to the above description, but instead with reference to the accompanying claims, the scope of such claims will be determined along with the full scope of the described equivalent. Should be. Future improvements will occur in dosimetry images and it is planned and intended that the present invention will be incorporated into such future embodiments.

<figref num="1">FIG. 6 is a block diagram that provides an overview of the system that creates the IMRT self-calibration curve (ISCC) used in at least one embodiment.</figref><figref num="2">It is a process flow diagram explaining the process flow of the initial calibration process which concerns on embodiment.</figref><figref num="3A">FIG. 5 is a processing flow diagram illustrating the processing of an acquired image to be used to create an ISCC curve according to an embodiment.</figref><figref num="3B">It is a process flow diagram explaining the ISCC generation process which concerns on embodiment.</figref><figref num="4">An example of the ISCC curve is shown.</figref><figref num="5">It is a process flow diagram explaining the next calibration process.</figref><figref num="6A">An example of a dose map is shown.</figref><figref num="6B">An example of the acquired image is shown.</figref><figref num="7A">An example of a dose map after being divided into smaller geometric regions and statistical functions applied to those regions is shown.</figref><figref num="7B">An example of the acquired image after being divided into small geometric areas and the statistical function applied to the area is shown.</figref><figref num="8A">An example of the original ISCC curve is shown.</figref><figref num="8B">An example of a post-processed ISCC curve is shown.</figref><figref num="9A">An example of a columnar dose map is shown.</figref><figref num="9B">An example of the acquired image in column format is shown.</figref><figref num="10">An exemplary graph showing the correlation coefficient for each set of corresponding columns of images shown in FIGS. 9A and 9B is shown.</figref><figref num="11A">An exemplary image of the column format representing the dose map is shown, in which the columns are divided according to the value of the correlation coefficient.</figref><figref num="11B">An example image in a column format representing an acquired image in which columns are divided according to the value of the correlation coefficient is shown.</figref><figref num="12A">An exemplary original ISCC curve is shown based on illustrating pixel values above the correlation threshold for the dose map and the region of the acquired image.</figref><figref num="12B">An example of the ISCC curve of FIG. 12A after post-treatment is shown.</figref><figref num="13">An exemplary graph including the experimentally obtained calibration curve and the ISCC curve is shown.</figref><figref num="14">An exemplary graph showing the dose difference between the experimentally obtained calibration curve and the ISCC curve is shown.</figref><figref num="15A">An exemplary plot of correlation statistics for the 10 equal parts of the calibration curve is shown.</figref><figref num="15B">An exemplary plot of least squares statistics for the 10 equal parts of the calibration curve is shown.</figref><figref num="16">An exemplary standardized curve for some of the calibration curves is shown.</figref>

Code description

102 Treatment planning system 104 Treatment plan 106 Dose map 110 Image acquisition device 112 Acquired image 114 ISCC curve 114a First-order ISCC curve 114b quadratic ISCC curve