Star sheet verification method for analyzing isocenter of radiotherapy accelerator

Abstract

The present application relates to a kind of star-shaped film verification methods for analyzing the center of radiotherapy accelerator, the irradiation film of radiotherapy accelerator is photographed by mobile phone or other equipment;The photo taken is introduced into the star-shaped film analysis of computing analysis unit, in computing analysis unit, in turn, after image calibration, field center line search, the minimum incircle calculation of all field, laser line indication center search and the maximum distance calculation of the center point to all field, the minimum incircle of all isocenter line and the maximum distance of the indication center of laser line to all field center line are output.The proved is the comparable alternative for the commercial software for star-shaped film analysis, while increasing the flexibility of use, reducing the dependence on film processor, scanner, special software.Compared with the current commercial product, it can be used in more hospitals, and is conducive to popularization.It is conducive to improving the accelerator maintenance level of relevant units, and ensuring the treatment effect of patients.

Description

Technical Field

[0001] The present invention relates to an information extraction technology, in particular to a star slice verification method for analyzing the isocenter of an accelerator for radiotherapy. Background Art

[0002] During the use of the radiotherapy accelerator, the field center of the entire accelerator will shift due to the elastic deformation and mechanical misalignment of the machine. According to the recommendations of the American Association of Physics in Medicine (AAPM) Task Force 142 (TG142), the coincidence of the field and the mechanical isocenter must be verified annually to ensure the accuracy of treatment with a maximum deviation of 2mm. A common method for testing isocenters is star film analysis, which generates a film that irradiates at different angles to form strip fields. By finding the smallest circle that touches each field centerline at least once, the maximum change in the field isocenter due to machine movement can be found. The flowchart of the entire star film verification system is shown below. Figure 1 As shown, the hardware and software equipment required include film, film processor, scanner, dedicated analysis software, and corresponding film calibration curve. Multiple devices are used alternately in the process, resulting in accuracy being affected by the process and equipment precision.

[0003] The current standard validation process involves numerous devices and steps, resulting in high costs for both equipment and consumables. This means many hospitals lack access to all the necessary equipment and software. Furthermore, validation results are delayed, preventing timely feedback from testing to facilitate machine calibration and maintenance. Summary of the Invention

[0004] To address the challenges of current star-shaped film verification systems, a new star-shaped film verification method for analyzing radiotherapy accelerator isocenters has been proposed. Based on computer vision, this method bypasses traditional workflows by directly photographing the developed film using a mobile phone or other device. The photographs are then analyzed and processed to produce the desired results. Multiple comparisons have shown good consistency between the results of this method and those from existing commercial software. This convenient method will improve accelerator maintenance at relevant institutions and ensure patient treatment outcomes.

[0005] The technical solution of the present invention is: a star film verification method for analyzing the isocenter of an accelerator for radiotherapy, using a mobile phone or other device to take a photo of the developed film after irradiation of the radiotherapy accelerator; importing the taken photo into a calculation and analysis unit for star film analysis, and in the calculation and analysis unit, after image calibration, field centerline search, calculation of the minimum inscribed circle of all fields, search for the laser line indication center and calculation of the maximum distance from the center point to all fields, outputs the minimum inscribed circle of all isocenter lines and the maximum distance from the laser line indication center to all field center lines.

[0006] Furthermore, the image calibration performs an affine transformation on the photographed photo and associates the image with the size of the real-world film to correct the image distortion.

[0007] Furthermore, the field center line search is as follows: the corrected image is converted into a grayscale image, and inverted, black is converted to white, and white is converted to black, and the field part in the image is enhanced by using limited contrast adaptive histogram equalization technology; with the center of the laser line as the origin, a circle is drawn with a set length as the radius, and each pixel point encountered on the circle is searched counterclockwise from 0°, and an angle-grayscale value curve is formed with the angle as the horizontal coordinate and the grayscale value as the vertical coordinate; the curve is mean filtered to eliminate the influence of noise on peak search, and then all local maxima on the curve are searched. If the number of maxima is odd, the radius is increased and the search is repeated until an even number of local maxima is found; with each maximum value as the center, the half-height width corresponding to the peak is calculated, the angle corresponding to the midpoint of the half-height width is calculated, and the pixel coordinates corresponding to the angle are searched. The number N of field center lines is half the number of pixel points, and all coordinates are paired in pairs according to (i, N+i) to form N straight lines, which are the searched field center lines.

[0008] Furthermore, the minimum inscribed circle of all the fields is calculated by using an exhaustive search method to calculate the distances from all pixel points to the center lines of all the fields within a square of a set side length centered on the laser point, where the point corresponding to the minimum distance is the center of the minimum inscribed circle, and the minimum distance is the radius of the minimum inscribed circle.

[0009] Furthermore, the laser line indication center is searched and the maximum distance from the center point to all fields is calculated: the laser line indication point is calibrated, the laser line indication center is calculated, and the distance from the center point to the center line of all fields is calculated. At the same time, the distance from the laser line indication center to the center line of all fields is calculated, and the maximum distance is the required distance.

[0010] The present invention has the following beneficial effects: The method for validating stellate films for analyzing radiotherapy accelerator isocenters has been demonstrated to be a comparable alternative to commercial software for stellate film analysis, while also increasing flexibility and reducing reliance on film processors, scanners, and specialized software. Compared to current commercial products, it can be used in more hospitals, facilitating widespread adoption. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 Flowchart of the verification system for the entire star slice;

[0012] Figure 2 Flow chart of the operation of the method of the present invention;

[0013] Figure 3Flowchart for realizing the method of the present invention;

[0014] Figure 4 This is an interface diagram of the calculation and analysis unit of the present invention;

[0015] Figure 5 Shooting and generating images for the implementation of the method of the present invention;

[0016] Figure 6 for Figure 5 The rectified image;

[0017] Figure 7 This is a schematic diagram of the position of the laser line after the method of the present invention is implemented;

[0018] Figure 8 Angle-gray value curve diagram for implementing the method of the present invention;

[0019] Figure 9 A schematic diagram of an isocenter line corresponding to each field of view is identified for the implementation of the method of the present invention;

[0020] Figure 10 Schematic diagram of calculating the minimum inscribed circle according to the method of the present invention. DETAILED DESCRIPTION

[0021] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0022] The use process of the present invention is as follows Figure 2 As shown, after irradiation by the radiotherapy accelerator, the radiation field will be developed on the non-processing film. Use a mobile phone or other device to take a photo of the developed film, and then import the photo into the calculation and analysis unit of the present invention to perform star film analysis and obtain the required results. The flow chart of the calculation and analysis unit is shown in FIG. Figure 3 As shown in the figure, when importing a captured photo, due to the angle and distance of the shot, an affine transformation is performed on the photo. This transformation then correlates the image with the size of real-world film to correct image distortion. The corrected image is then processed to extract the centerline of each field of view. Finally, the minimum inscribed circle of these centerlines and the maximum distance from the laser line to these centerlines are calculated, and the test results are output. Aside from the initial process-free film, the entire process does not require any other equipment, consumables, or software products.

[0023] The present invention is based on basic mathematical principles in calculating the minimum inscribed circle of the centerline and the distance from the laser line to the centerline of each field. Image processing primarily involves distortion correction, grayscale conversion, image enhancement, filtering, and peak detection. Based on practical observations, distortion primarily occurs in displacement and tilt, while the image maintains its essentially flat nature. Based on this, an affine transformation is used to correct distortion in the captured image, achieving good results. The corrected image is converted to a grayscale image and inverted, converting black to white and vice versa. Contrast-limited adaptive histogram equalization (CLAHE) is then used to enhance the field portion of the image. A circle is drawn with the center of the laser line as the origin and a radius of a given length. Each pixel encountered on the circle is searched counterclockwise from 0°, forming an angle-grayscale curve with the angle as the horizontal coordinate and the grayscale value as the vertical coordinate. This curve is mean filtered to eliminate the effects of noise on peak detection. The curve is then searched for all local maxima. If the number of maxima is odd, the radius is increased and the search is repeated until an even number of local maxima is found. Taking each maximum value as the center, calculate the half-height width corresponding to the peak, calculate the angle corresponding to the midpoint of the half-height width, and find the pixel coordinates corresponding to the angle. The number of field centerlines N is half the number of pixel points. All coordinates are paired according to (i, N+i) to form N straight lines. Then, using an exhaustive search method, calculate the distance from all pixel points to all field centerlines within a 2cm square centered on the laser point. The point corresponding to the minimum distance is the center of the minimum inscribed circle, and the minimum distance is the radius of the minimum inscribed circle. At the same time, calculate the distance from the laser point to all field centerlines, and the maximum distance is the required distance.

[0024] The calculation and analysis unit includes four modules: image calibration, field centerline search, calculation of the minimum inscribed circle of all fields, search for the center of laser line indication, and calculation of the maximum distance from the center point to all fields.

[0025] Regarding the first image calibration module, it mainly includes three parts: image import, image affine transformation, and association between image pixels and real-world size.

[0026] The second field centerline search module mainly includes six parts: converting the image from RGB to grayscale image, enhancing the image contrast using CLAHE technology, searching for the grayscale values ​​of pixels on a circle at a certain distance from the center of the laser line indicator, mean filtering of the grayscale value-angle curve, searching for all local maxima of the grayscale value-angle curve, and generating the field centerline.

[0027] Regarding the third module for calculating the minimum inscribed circle of all fields, it mainly refers to the exhaustive method, which uses three granularities of 0.4mm, 0.2mm and 0.1mm as steps to search the distances from all pixel points to the center lines of all fields in squares with side lengths of 2cm, 1cm and 0.5cm respectively. The one with the largest distance represents the distance of the point, and the point with the smallest distance among all pixel points is the center of the minimum inscribed circle, and the minimum distance is the radius of the minimum inscribed circle.

[0028] The module for finding the center of the laser line indicator and calculating the maximum distance from the center point to all the fields mainly includes three parts: calibration of four laser line indicator points, calculation of the center of the laser line indicator, and calculation of the distance from the center point to the center line of all the fields.

[0029] The interface of the calculation and analysis unit is as follows Figure 4 As shown in the figure, the functional modules are divided into four parts: Image Correction, Laser Coordinate, Segment to Line, and Star Shot. The final result is displayed within the range shown in the red box. At the same time, Export to Excel is used to output the results for subsequent search and use.

[0030] First, based on the standard protocol, a radiotherapy accelerator is used to irradiate the non-processed film, and a mobile phone or other camera device is used to take a picture of the developed star film to generate an image file. Figure 5 shown.

[0031] Then click Open-open file, import the captured image file into the calculation and analysis unit, click ImageCorrection, perform affine transformation on the image, correct the distortion caused by shooting, and associate the photo size with the real film size. The corrected image is as follows Figure 6 shown.

[0032] Click Laser Coordinate and mark the four points indicating the position of the laser line (the position where the laser line is projected on the film). The photo is imported into the calculation and analysis unit of the present invention. The position is identified and calculated in the calculation and analysis unit, and the indicated center of the laser line is calculated. The result is as follows: Figure 7 shown.

[0033] Click Segment to Line, and the calculation and analysis unit will automatically read the calibrated image (such as Figure 8 The obtained curve graph is shown in Figure 1. The horizontal axis is the degree angle, the vertical axis is the gray scale, and the peak selection is the peak selection. The isocenter line corresponding to each field is searched. Figure 9 shown.

[0034] Click Star Shot, and the calculation and analysis unit will automatically calculate the minimum inscribed circle of all isocenter lines based on the previously searched isocenter lines. The results are as follows: Figure 10 At the same time, the calculation and analysis unit will calculate the maximum distance from the laser line center to the center line of all the fields based on the laser line center obtained previously. The two results will be Figure 4 It is displayed in the position shown in the box in the interface diagram.

[0035] The results of the analysis of seven process-free films were compared with those of a commercial APP, as shown in Table 1. The difference in the minimum inscribed circle radius was (0.09 ± 0.09) mm, and the difference in the maximum distance from the laser line to all fields was (-0.02 ± 0.17) mm.

[0036] Table 1

[0037]

[0038] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the patent for this invention shall be determined by the appended claims.

Claims

1. A star slice verification method for analyzing the isocenter of an accelerator for radiotherapy, characterized in that: Use a mobile phone to take a photo of the developed film after irradiation by the radiotherapy accelerator; import the taken photo into the calculation and analysis unit for star film analysis. After the calculation and analysis unit goes through image calibration, field centerline search, calculation of the minimum inscribed circle of all fields, search for the laser line indication center, and calculation of the maximum distance from the center point to all fields, the minimum inscribed circle of all isocenter lines and the maximum distance from the laser line indication center to the center line of all fields are output; The image calibration performs an affine transformation on the photographed photo and associates the image with the size of the real-world film to correct image distortion; The field center line search: convert the corrected image into a grayscale image, and invert it, convert black to white, and white to black, and use the contrast-limited adaptive histogram equalization technology to enhance the field part in the image; use the center of the laser line as the origin, set the length as the radius to draw a circle, search for each pixel point encountered on the circle counterclockwise from 0°, and form an angle-grayscale value curve with the angle as the horizontal coordinate and the grayscale value as the vertical coordinate; perform mean filtering on the curve to eliminate the influence of noise on peak search, and then search for all local maxima on the curve. If the number of maxima is odd, increase the radius and search again until an even number of local maxima are found; with each maximum value as the center, calculate the half-height width corresponding to the peak, calculate the angle corresponding to the midpoint of the half-height width, and find the pixel coordinates corresponding to the angle, the number of field center lines N is half the number of pixels, and all coordinates are calculated according to ( i, N+i ) Paired in pairs, forming N The straight line is the center line of the search field; Calculation of the minimum inscribed circle of all fields: using an exhaustive search method, within a square of a set side length centered on the laser point, calculate the distances from all pixel points to the center lines of all fields, where the point corresponding to the minimum distance is the center of the minimum inscribed circle to be sought, and the minimum distance is the radius of the minimum inscribed circle; The laser line indication center search and the maximum distance calculation from the center point to all fields: laser line indication point calibration, laser line indication equal center calculation, distance calculation from the equal center to all field center lines. The distance from the laser line indication center to all field center lines is calculated at the same time, and the maximum distance is the required distance.

Images