Automated qualitative description of anatomical changes in radiotherapy

Abstract

A system and method for monitoring anatomical changes in a subject during radiation therapy, as well as an apparatus for medical imaging and analysis and a computer program product for performing the method are provided. To monitor anatomical changes in a subject during radiation therapy, the following steps are performed: receiving first anatomical image data and subsequent anatomical image data of the subject; analyzing the first anatomical image data and the subsequent anatomical image data, the analysis including registering the subsequent anatomical data to the first anatomical data; identifying a change between the first anatomical image data and the subsequent anatomical image data as a change state; matching the identified change state with a corresponding qualitative description; and providing a monitoring report including a qualitative description of the identified change.

Description

Technical Field

[0001] The present invention relates generally to monitoring of anatomical changes and, in particular, but not exclusively, to automated monitoring of anatomical changes in a subject undergoing radiation therapy. Background Art

[0002] In radiation therapy, target structures (e.g., tumors) in a patient's body are treated by subjecting them to radiation. The radiation can be in the form of external radiation (e.g., photons or particles (e.g., protons), such as in external beam radiation therapy (EBRT). The treatment is delivered in such a way that the radiation delivered to the target structure (TS) is as high as possible, while the radiation delivered to the surrounding healthy tissues and structures (often referred to as organs at risk (OARs)) is as low as possible.

[0003] During treatment, radiation doses are usually delivered to the patient in several radiation therapy sessions, with a recovery period between each phase. This approach is called fractionated radiation therapy, and the sessions are called fractions. The reasoning behind this approach is that TS tumor tissue is expected to recover less well from dose fractionation than healthy tissue, including OARs.

[0004] Fractions are typically delivered once a day, allowing treatment to be carried out in a timely manner. During this time, the patient may experience anatomical changes, such as organ movement and deformation that occur naturally in the body. Each fraction of treatment delivered to the patient can also affect the anatomical structure. In particular, during radiation therapy, it is expected that the target structure (e.g., a tumor) will shrink. In the current context, "during" radiation therapy refers to the entire time span in which all treatment fractions, including the recovery period, are delivered, rather than a specific point in time when the patient is actually subjected to radiation. "During" radiation therapy or "in" radiation therapy is an indication of the context in which anatomical changes occur, but monitoring changes occurs before or after the delivery of the treatment fractions, rather than during the actual delivery of the fractions. It is necessary to monitor and evaluate the anatomical changes that occur in the patient during treatment to ensure that the treatment is proceeding properly and that changes or additional measures can be made if necessary.

[0005] In current clinical practice of radiation therapy, changes in the patient's anatomy that occur during radiation therapy are monitored and recorded in a treatment log. Qualitative descriptions are used to describe, communicate, and record these changes along with quantitative measurements. Typically, these changes are visually assessed and manually recorded. Typically, visual inspection involves manual processing of quantitative measurements and comparison of treatment planning images (e.g., CT scans) with in-room acquisition images (e.g., cone beam CT (CBCT) images or planar X-ray images) or separately recorded MRI images. In-room images are typically taken immediately prior to delivering treatment fractions and are used for status monitoring and patient setup.

[0006] Based on this visual assessment and analysis of quantitative data, clinicians qualitatively interpret and describe the type of anatomical changes. However, visual assessment, manual processing, and interpretation are error-prone and time-consuming. Errors can occur because clinicians may overlook relevant anatomical changes or may misinterpret the spatial relationships between objects in different images. Furthermore, manually entering qualitative descriptions can lead to variations in descriptions of similar conditions, causing inconsistencies. In addition, the interpretation of anatomical changes requires the analysis of large amounts of data and specialized skills. Summary of the invention

[0007] The present invention seeks to provide an automated method for monitoring anatomical changes in a subject undergoing radiation therapy.The present invention further seeks to address the need for consistent qualitative description of changes in anatomical structures in a subject during treatment.

[0008] To this end, systems and methods for monitoring anatomical changes in a subject during radiation therapy as well as apparatus for medical imaging and analysis and computer program products for performing the methods are provided.

[0009] A system for monitoring anatomical changes in a subject during radiation therapy comprises an analysis unit comprising an input configured to receive first anatomical image data and subsequent anatomical image data of the subject, the analysis unit being configured at least to register the subsequent anatomical data to the first anatomical data. The system further comprises: a change state identification unit configured to identify a change between the first anatomical image data and the subsequent anatomical image data as a change state; and a qualitative translator configured to match the identified change state with a corresponding qualitative description. A change state is a number of predefined categories, wherein the changes are grouped according to their characteristics. The change state is identified by assigning a category that best corresponds to a particular change between the first anatomical image data and the subsequent anatomical image data. Preferably, the qualitative description provides an indication of the characteristics of the change state. More preferably, the qualitative description reflects how a physician would describe the characteristics of the change when monitoring a patient during treatment. The system further comprises a reporting unit configured to provide a monitoring report including a qualitative description of the identified change state. The corresponding qualitative description is a natural language description and / or a graphic image, preferably in the form of a pictogram. The advantage of a pictogram is that it is easy to understand not only for the clinician himself, but also for example for the patient or a consulting physician from a different medical field.

[0010] In an embodiment of the system, the change state identification unit includes at least one qualitative spatial reasoning algorithm, and the qualitative spatial reasoning algorithm is preferably an RCC-8 calculation algorithm and / or a Cardinal spatial reasoning algorithm.

[0011] Preferably, the analysis unit of the system further comprises an input configured to receive dose distribution data of a treatment plan for the subject; and the change state identification unit is additionally or alternatively configured to identify a change between the dose distribution data and subsequent anatomical structure image data as a change state. By including such an optional input, the system is able to additionally or alternatively monitor anatomical changes in the subject compared to the planned dose distribution.

[0012] Preferably, the qualitative translator is configured to match the identified change state with the corresponding qualitative description by using a lookup table.

[0013] Preferably, the reporting unit comprises a display for visually displaying the monitoring report. This is particularly preferred when the monitoring report is a pictographic report. It is particularly advantageous that the display is configured to visually display the monitoring report in the form of a graphical user interface. The advantage of presenting the report in the form of a user interface is that the report version most suitable for the situation can be selected and displayed. Another advantage is that additional data in the background (for example, acquired patient images, dosimetry information or quantitative data obtained from another analysis) can be used and these additional data can be selected and displayed when appropriate or necessary. When these additional data are not selected, the information is not displayed, thereby simplifying the main view of the report and making it easier to understand. Preferably, in order to optimally use these options, the graphical user interface includes at least one control for selecting the information to be displayed as part of the monitoring report.

[0014] A method for monitoring anatomical changes in a subject during radiation therapy comprises the steps of receiving first anatomical image data and subsequent anatomical image data of the subject; and analyzing the first anatomical image data and the subsequent anatomical image data, wherein the step of analyzing comprises registering the subsequent anatomical data to the first anatomical data. The method further comprises the steps of identifying a change between the first anatomical image data and the subsequent anatomical image data as a change state; and matching the identified change state with a corresponding qualitative description. The method further comprises the steps of providing a monitoring report including a qualitative description of the identified change. The method is preferably computer-implemented.

[0015] In a preferred embodiment of the method, the step of identifying changes includes supplying the first anatomical structure image data and the subsequent anatomical structure image data to at least one qualitative spatial reasoning algorithm, which is preferably an RCC-8 algorithm and / or a Cardinal spatial reasoning algorithm.

[0016] Preferably, the step of receiving first anatomical structure image data and subsequent anatomical structure image data of the object further comprises receiving dose distribution data of a treatment plan for the object, and the step of identifying changes additionally or alternatively comprises identifying changes between the dose distribution data and subsequent anatomical structure image data as a change state.

[0017] In another preferred embodiment, which may be combined with other embodiments and preferences, the method step of analyzing the anatomical structure image data further comprises obtaining quantitative image data, the quantitative image data preferably comprising at least one of the following: the size of the region of interest, the change in size of the region of interest, the distance the region of interest has shifted, the total radiation dose the region of interest has received. When obtaining quantitative image data, the monitoring report preferably further comprises at least one quantitative description of at least one of the identified changes.

[0018] An apparatus for medical imaging and analysis comprises: one or more imaging devices configured to provide images of an object to be treated; a contouring tool configured to provide anatomical image data based on the images provided by the one or more imaging devices; and a system as described above for monitoring anatomical changes during radiation therapy.

[0019] One advantage of the present invention is that anatomical changes in patients undergoing radiation therapy can be monitored more quickly and accurately. Identifying changes as change states allows for consistent classification of changes. Automated analysis of first and subsequent images combined with automated identification of changes can be faster and more accurate than manual comparison of images by a clinician.

[0020] Another advantage of the present invention is that anatomical changes in patients undergoing radiation therapy can be monitored more consistently. Identifying changes as change states allows for consistent classification of changes. Moreover, each change state is matched with a corresponding qualitative description. Thus, changes belonging to the same category will have the same qualitative description. This has the additional advantage that it is easier for a physician or clinician to find similar cases in a database. This can also have the additional advantage that the data set can be used as input to a machine learning algorithm.

[0021] Another advantage is that a monitoring report is presented that allows for immediate understanding of changes without further analysis. Because the changes are reported in a qualitative manner (e.g., by using language similar to that used by a physician when describing a patient's status), the report can be read or shown to a patient or consulting physician for informational purposes without requiring extensive interpretation of the data.

[0022] Other advantages from the described invention will also be apparent to the skilled person. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In the following figures:

[0024] Figure 1 A system for monitoring anatomical changes in a subject during radiation therapy and other components of an apparatus for medical imaging and analysis are schematically and exemplarily illustrated.

[0025] Figure 2a and Figure 2b State detection by a qualitative spatial reasoning algorithm is schematically and exemplarily illustrated.

[0026] Figure 3 An example of a method for monitoring anatomical changes during radiation therapy is schematically illustrated.

[0027] Figure 4 Another example of a method for monitoring anatomical changes in radiation therapy is schematically illustrated.

[0028] Figure 5 An example of a monitoring report of anatomical changes in a subject in the form of a pictographic report is schematically illustrated.

[0029] Figure 6 Another example of a monitoring report of anatomical changes in a subject in the form of a graphical user interface is schematically illustrated. DETAILED DESCRIPTION

[0030] Figure 1 A system 110 for monitoring anatomical changes in a subject during radiation therapy is illustrated. In this example, the system is illustrated as part of an apparatus 100 for medical imaging and analysis.

[0031] In the apparatus 100 for medical imaging and analysis, a planning image 101 of an object to be treated is acquired by an imaging device 102. The image can be a computed tomography image (CT), a magnetic resonance image (MR), a positron emission tomography (PET) image, another medical image or a combined image, for example a combined PET / CT or PET / MR image. Figure 1 In the figure, as an example, a PET / CT imaging device 102 is illustrated. A region of interest (ROI) is delineated in the first image using a contouring tool 103 to provide first anatomical structure image data 104. The ROI in the object will include at least one target structure (TS), typically a tumor, but may also include one or more organs at risk (OARs). The contouring tool typically interacts with a user (the user can be, for example, a trained medical imaging technician or radiologist) to define the contours of the ROI. This process can be completely manual, or partially or completely automated.

[0032] During the treatment process, at least one subsequent image 105 of the object is acquired. The image may be acquired using the same imaging device 102 as the planning image, but can also be acquired using one or more alternative imaging devices 106. Such alternative imaging devices can be CT, MR or other medical imaging devices. In a preferred embodiment, subsequent images are acquired using a second medical imaging device 106, which is a cone beam CT (CBCT) or planar X-ray imaging device configured to image the object in the treatment room. The advantage of this setup is that the images can be used to monitor anatomical changes in the object and for object setup for delivery of radiation therapy treatment fractions.

[0033] In subsequent images 105, the ROI is also delineated using the contouring tool 103 to provide subsequent anatomical structure image data 107. Contouring can be started from scratch, such as manually, automatically or semi-automatically, interactively with the user in the planning image. Contouring of subsequent images can also be done by automatically propagating contours from the planning image 101 to the subsequent image 105 using a deformable image registration process. When using this process, the user preferably has the option of approving and / or manually correcting the automatically propagated contours.

[0034] Next, the anatomical structure image data 104 and 107 are analyzed in order to monitor anatomical structure changes in the subject. This is done using a system 110 for monitoring anatomical structure changes in radiation therapy. The system comprises an analysis unit 111, a change state identification unit 112, a qualitative translator 113 and a reporting unit 114. The analysis unit 111 comprises an input configured to receive the first anatomical structure image data 104 and the subsequent anatomical structure image data 107 of the subject.

[0035] The analysis unit 111 is at least configured to register the subsequent anatomical structure image data 107 to the first anatomical structure image data 104. As explained above, the first planning image 101 may be acquired using an imaging device 102 that is different from the device 106 used to acquire the subsequent image 105. Moreover, for each image acquisition, the position of the object in the imaging device will be slightly different. As a result, the reference system for the anatomical structure image data can change for each image and the scaling. However, in order to make an accurate and reliable comparison, the first anatomical structure image data 104 and the subsequent anatomical structure image data 107 should have the same reference system. In order to be able to detect changes caused by changes in the anatomical structure of the object rather than changes in the image acquisition environment, the images should have the same scaling and the same coordinate system. To this end, the subsequent image and its anatomical structure image data need to be matched with the first anatomical structure image and its anatomical structure image data in terms of scaling and reference system. This matching process is called "registration". Registering the subsequent anatomical structure image data to the first anatomical structure image data can be accomplished by identifying and matching anatomical structure landmarks. The anatomical landmarks are structures in the anatomy of the subject that are known to change little or not at all over the time frame of the radiation treatment, e.g., bone structures. The anatomical landmarks can be part of the anatomical image data in outline form, e.g., an outline of a bone structure as part of an image. The subsequent anatomical image data is then scaled, translated, and rotated until the size and position of the anatomical landmarks correspond as closely as possible to the size and position of the anatomical landmarks in the first anatomical image data. Preferably, this registration is automated.

[0036] The analysis unit 111 may also be configured to quantify the anatomical changes between the first anatomical structure image data and the second anatomical structure image data. Such quantification may be absolute, for example, the displacement size of the OAR in millimeters. Such quantification may also be relative, for example, the percentage by which the TS has shrunk. In particular, the analysis unit 111 may also be configured to perform the following Figure 4 Optional additional analytical method steps are described.

[0037] The change state identification unit 112 is configured to identify the change between the first anatomical structure image data 104 and the subsequent anatomical structure image data 107 as a change state. The change state is a number of predefined categories, in which they are grouped according to the characteristics of the change. The change state is identified by assigning a category that best corresponds to the characteristics of the change between the first anatomical structure image data and the subsequent anatomical structure image data. The change state can be based on categories with characteristics that the user has predefined, but can also be based on a spatial model or a size model. The spatial model defines the change state according to the possible relationship between two areas. The number and nature of the change states of the model will depend on the characteristics of interest. One or more models for identifying the change state can be selected accordingly.

[0038] For example, the displacement of the ROI may be a characteristic of interest. In this case, the change can be classified or modeled based on the size or direction of the displacement. A well-known direction model is the compass direction, which defines four states of change: north, south, east, and west. The size of the displacement can be classified using a range defined by the user (for example, three states of change are defined: small, medium, and large). In an alternative example, the overlap of the ROI with its original position is of interest. A simple overlap model defines four states of change: complete overlap, partial overlap, touching, and no overlap. In another alternative example, the change in the size of the ROI may be a characteristic of interest. For this example, a simple model defines three states of change: growth, shrinkage, and no change.

[0039] The qualitative translator 113 is configured to match the change state that has been identified by the change state identification unit 112 with a corresponding qualitative description. Preferably, the qualitative description provides an indication of the characteristics of the change state. For example: "The OAR has shifted to the left" or "The TS has shrunk". More preferably, the qualitative description reflects how a physician would describe the characteristics of the change when monitoring the patient during treatment. The qualitative translator 113 is preferably configured to match the identified change state with the corresponding qualitative description by using a lookup table. The lookup table provides a simple way to consistently use the same qualitative description for each change state when it is detected.

[0040] The reporting unit 114 is configured to provide a monitoring report including a qualitative description of the identified change status. Figure 1 In the example of FIG. 1 , the reporting unit includes a display 115 for visually displaying the monitoring report. In a preferred embodiment, the monitoring report is displayed in a graphical user interface (e.g., Figure 6 The graphical user interface (as shown) is displayed in the form of a graphical user interface.

[0041] exist Figure 1In the exemplary embodiment shown, the analysis unit 111 of the system 110 for monitoring anatomical changes additionally includes an input for receiving dose distribution data 108 of a treatment plan for an object. This can be the same input as the input configured to receive anatomical image data, but can also be a different input. The dose distribution data preferably includes defined dose areas (e.g. in the form of isodose contours). The system is able to use these defined dose areas to monitor anatomical changes in the object relative to the planned dose (instead of the first anatomical image data 104). To this end, the change state identification unit 112 is preferably also configured to identify a change between the dose distribution data 108 and subsequent anatomical image data 107 as a change state. By including this option, the system is able to additionally or alternatively monitor anatomical changes in the object compared to the planned dose distribution.

[0042] In an embodiment of the monitoring system 110, the change state identification unit 112 includes at least one qualitative spatial reasoning algorithm, which is preferably an RCC-8 algorithm and / or a Cardinal spatial reasoning algorithm. Qualitative spatial reasoning is a field of artificial intelligence that deals with the problem of generating qualitative descriptions that summarize similar quantitative measurements.

[0043] Figure 2a and Figure 2b Change state detection by a qualitative spatial reasoning algorithm is schematically and exemplarily illustrated.

[0044] Figure 2a The diagram shows the detection of change states in the form of a cardinal orientation relationship between two objects. This diagram is for 2D cardinal space reasoning, but this form of modeling can also be used in 3D. 3D cardinal qualitative space reasoning is particularly suitable for detecting and qualitatively describing the displacement of ROIs.

[0045] like Figure 2a As shown on the left side of the figure, the ROI 201 forming the basis for comparison is adapted to the bounding box, and a direction grid is defined around the bounding box. The direction grid forms a model of possible displacement states. The nine change states of this model (here 2D) are from top to bottom and from left to right: SL to the upper left, L to the left, IL to the lower left, S to the upper left, O to no displacement, I to the lower right, SR to the upper right, R to the right, and IR to the lower right.

[0046] Figure 2aThe right side of 200 shows the detection of the change state of the ROI. The ROI in the first anatomical structure image data 202 is shown with a dashed outline and a grid, and the same ROI is shown with a solid outline and a bounding box in the subsequent anatomical structure image data 203. The identified change state is "SL", and its corresponding qualitative description is in the natural language form of "the ROI has shifted to the upper left".

[0047] Figure 2b The diagram shows the change states identified by the RCC-8 algorithm. The RCC-8 algorithm is a regional connection algorithm. The RCC-8 algorithm models the relationship between two regions as eight possible spatial relationships that form a change state. The RCC-8 algorithm can be used to describe the change of a single ROI, but is particularly suitable for describing the location and overlap of the TS and OAR with the high-dose area of ​​the dose distribution of the treatment plan. The RCC-8 algorithm can also help describe the change when the TS and OAR overlap or are close.

[0048] The changing state of the RCC-8 calculation model is Figure 2b From top to bottom and from left to right in the figure: disconnected DC, partially overlapped PO, externally connected EC, equal EQ, tangent proper portion TPP, inverse tangent proper portion TPPi, non-tangent proper portion NTPP, and non-tangent proper portion NTPPi.

[0049] exist Figure 2b , change states are illustrated for ROI A and ROI B. These can be the same ROI, wherein the first anatomical structure image data is for A and the subsequent anatomical structure image data is for B. As an example of this, when the change state "A NTPPi B" is identified, this would have a corresponding qualitative description in the form of natural language "the ROI has shrunk". Or, for example, when the change state "A DC B" is identified, this would have a corresponding natural language "the ROI has completely moved out of its original position". Alternatively, ROI A can be dose planning data and ROI B can be a ROI of a subsequent image. As an example of this, when the change state "A NTTPi B" is identified, this would have a natural language description "the ROI falls within the high-dose treatment area". As a further alternative, A and B can also be two different ROIs within the same anatomical structure image data. For example, A can be an OAR and B can be a TS. As an example of this, when the change state "A PO B" is detected, this would have a corresponding natural language "the OAR overlaps with the TS".

[0050] For the case where the anatomy of the object has not changed and one or more ROIs in the first anatomical structure image data and the subsequent anatomical structure image data are substantially the same, this is also identified as a change state. A qualitative example description of a match for this state is "ROI no change". Figure 2a In , this state is detected as "O", while in Figure 2b In the EQ state, this state is detected as "EQ".

[0051] Figure 3 The steps of a method for monitoring anatomical changes in a subject during radiation therapy 300 are schematically illustrated. The method for monitoring anatomical changes in a subject is preceded by the steps of obtaining first anatomical image data 301 and subsequent anatomical data 302. Figure 1 The anatomical structure image data is obtained in the same manner as described above with respect to the apparatus 100 for medical imaging analysis shown.

[0052] A method 300 for monitoring anatomical changes in a subject during radiation therapy according to the present invention comprises the steps of receiving first anatomical image data and subsequent anatomical image data of the subject 310, and analyzing the first anatomical image data and subsequent anatomical image data 320, wherein the step of analyzing comprises registering the subsequent anatomical data to the first anatomical data. The method further comprises the steps of identifying changes between the first anatomical image data and the subsequent anatomical image data as change states 330, and matching the identified change states with corresponding qualitative descriptions 340. The method further comprises the steps of providing a monitoring report 350 including a qualitative description of the identified changes.

[0053] Figure 4 The steps of another example of a method for monitoring anatomical changes in a subject during radiation therapy 400 are schematically illustrated. Figure 4 The method 400 includes Figure 3 Method steps 410, 420, 440 and 450 are similar to steps 310, 320, 340 and 350 of method 300 shown in FIG. Figure 3 Compared with the method 300, Figure 4 The method steps include a number of additional options. These steps represent advantageous additional options that can be used individually, in sub-combinations or in full combination, such as Figure 4 Advantageously, Figure 1 The illustrated system 110 for monitoring changes in anatomy may also be configured to perform any or all of these additional steps.

[0054] The method for monitoring anatomical changes in a subject is preceded by the step of obtaining first anatomical image data 301 and subsequent anatomical data 302. The method is also preceded by determining a treatment plan for the subject comprising dose distribution data 401 .

[0055] The method 400 for monitoring anatomical changes in a subject during radiation therapy according to the present invention comprises a step of receiving data 410. The step of receiving data 410 comprises receiving first and subsequent anatomical image data of the subject and also comprises receiving dose distribution data of a treatment plan of the subject.

[0056] The method then comprises a step of analyzing the data that has been received 420. This step of analyzing the data comprises registering the subsequent anatomical structure data to the first anatomical structure data 421. Figure 4 In the exemplary embodiment shown, the step of analyzing the anatomical structure image data also includes obtaining quantitative image data. Obtaining quantitative data can include any one or more of the following: determining the size or volume of the ROI 422, determining the change in the size or volume of the ROI 423, determining the distance the ROI has shifted 424 and / or determining the radiation dose for the ROI 425. The ROI can be one or more OARs and / or one or more TS. When obtaining quantitative data for the change of the ROI, this can be expressed as an absolute value of the change, and can also be expressed as a relative value, such as a percentage.

[0057] The method further comprises a step of identifying a change in the received data as a change state 430. A change between the first anatomical structure image data and the subsequent anatomical structure image data is identified 431, and additionally or alternatively a change between the dose distribution data and the subsequent anatomical structure image data is identified 432. In both steps, the first anatomical structure image data or the dose distribution data and the subsequent anatomical structure image data are preferably supplied to one or more spatial reasoning algorithms.

[0058] Figure 4The monitoring method of an exemplary embodiment of includes an optional step 426 of determining risk factors for the patient. Preferably, quantitative information about the dose that the ROI has received is calculated and stored for further reference after each treatment fraction is delivered. This information may include a dose volume histogram, a dose distribution in the ROI, or an average dose or total dose received by the entire area of ​​the ROI. The advantage of calculating and storing this quantitative dose information is that the dose that the ROI has received so far can be combined with the expected dose to be delivered in the next fraction based on subsequent image data. In clinical practice, the ROI predefines tolerances for one or more statistical descriptors of radiation received that are considered safe. These statistical descriptors can be a dose volume histogram, a minimum dose for the entire ROI area, a maximum dose for the entire ROI area, a maximum dose for the hottest sub-area of ​​the ROI, an average dose, or an average maximum dose or an average minimum dose. The appropriate statistical descriptor is determined by the nature of the ROI. For example, for TS, it can be specified that at least a minimum dose is delivered in the entire TS area. For another example, for OAR, it can be specified that the average dose received should not exceed a certain threshold. If the value of one or more statistical descriptors calculated for the ROI is within the predefined tolerance, the risk factor will be determined to be low. If the value of one or more statistical descriptors is below the tolerance, but is approaching that value, the risk will be determined to be medium. If the value of the statistical descriptors for the ROI will exceed the tolerance level for the next fraction, the risk will be determined to be high.

[0059] At step 440 of the method, the quantitative data obtained from the analysis and the identified change states and identified risk factors are collected and combined. The identified changes are matched with the corresponding descriptions. This step includes matching the identified change states with the corresponding qualitative descriptions 441. Where applicable, the qualitative description can be supplemented with quantitative data 442, as well as the identified risk factors 443. Preferably, the qualitative natural language description of the identified changes is supplemented with quantitative data, for example, "the tumor TS has shrunk by 10%", or "the organ OAR has moved 5mm to the left into the high-dose area". The risk factors determined at 426 can also be indicated individually (for example, indicating them as "high", "medium" or "low"), but can also be preferably added to the pictographic qualitative description in the form of color coding, for example, red is high, yellow is medium, and green is low. The quantitative data can also be used to supplement the pictographic information, for example, to illustrate the relative size of the elements of the pictogram.

[0060] At step 444, the collected and combined data is further used to identify whether the current patient status requires review and follow-up by a physician. For example, if the TS has shifted significantly outside the high-dose region of the treatment plan, the treatment plan may need to be revised. In another example, the OAR may have shrunk to the point where additional measures are required to stabilize the patient and / or reduce adverse side effects of the treatment.

[0061] The method also includes the step of providing a monitoring report including a qualitative description of the identified changes 450. In this embodiment, quantitative information is also included in the report and in any follow-up identified at 444 to alert a physician for review.

[0062] Figure 5 An example of a monitoring report of anatomical changes in an object in the form of a pictographic report 500 is schematically illustrated. The report is preferably presented on a display. However, the format also allows hard copy printing, so that it can be easily distributed as information to patients undergoing treatment. In this example, the monitoring report 500 is presented in the form of a table. The rows in the table are used for the regions of interest being monitored. In this example, the regions of interest are two OARs (OAR1 and OAR2 located in the treatment area) and a TS in the form of a tumor GTV. The region of interest can be, for example, the left and right parotid glands as OARs and a TS in the form of a tumor located near the left parotid gland. The columns represent the status of the patient before the treatment fraction is delivered. Column F1 is used for fraction 1, F2 is used for fraction 2, and so on. The illustrated report has been provided to fraction 5-F5. In this example, anatomical changes have been analyzed and reported according to the dose distribution of the treatment plan. The dose distribution data consists of the outline of the planned target volume PTV. This is the information indicated in the title row of the report ROI-PTV.

[0063] The pictogram report shows the change status determined using the RCC-8 algorithm. The PTV is represented as a circle with a dotted outline in each pictogram, and each ROI is represented by a circle with a solid outline. Quantitative data is added in the form of relative size and distance. The pictogram is further enhanced by a risk indicator in the form of a shadow. Low-risk ROIs have no shadows, medium-risk ROIs have crossed shadows, and high-risk ROIs have solid shadows. Alternatively, green, orange, and red can be used as a "traffic light" indication of risk. This color scheme is well known and easy to recognize.

[0064] Figure 5The pictographic report can be easily understood not only by the clinicians themselves, but also by, for example, patients or consulting physicians from different medical fields. For example, for OAR2, it can be seen that "OAR2 is outside the high-dose treatment area and has a low risk." For example, for OAR1, from F2 to F5 it can be seen that "OAR1 is shrinking and moving toward that high-dose treatment area, which is a high-risk situation." As an example for TS, it can be seen that before delivering fraction 2, "the tumor has shrunk, but has also moved toward the edge of the high-dose treatment area, which indicates an increased risk."

[0065] exist Figure 5 In the example of , fractions F3-F4 are delivered according to the original treatment plan. However, in a possible alternative strategy, based on the report provided for the setup prior to the delivery of fraction 3, the physician can also decide not to proceed with the delivery, but to acquire new planning images and revise the treatment plan for the patient. Having an automated report available that can be provided quickly and is easy to read means that setup images acquired from the patient in the treatment room immediately prior to delivery can be used for these assessments. Additional imaging for this purpose can be omitted.

[0066] Figure 6 Another example of a monitoring report of anatomical changes in an object in the form of a graphical user interface (GUI) 600 is schematically illustrated. The report is preferably presented on a display (e.g., a desktop computer display, a presentation screen, or a wireless display device). An advantage of presenting the report in the form of a user interface is that the report version most suitable for the situation can be selected and displayed. Another advantage is that additional data (e.g., images of the patient collected, dose information, or quantitative data obtained from additional analysis) can be obtained in the background and accessed and displayed as appropriate or necessary.

[0067] exist Figure 6 In the example of FIG. 6 , the main view of GUI 610 is similar to Figure 5 The pictographic report of the report 500 shown and having the same advantages. And here, the row of the table is for the monitored region of interest. In this example, the region of interest is again two OARs (OAR1 and OAR2 located in the treatment area) and a TS in the form of a tumor GTV. The column represents the state of the patient before the delivery treatment fractionation. Column F1 is used for fractionation 1, F2 is used for fractionation 2, and so on. The illustrated report has been provided to fractionation 5-F5. The pictographic report here again shows the change state determined using RCC-8 calculations and supplemented with quantitative data. The first anatomical structure image data is represented as a circle with a dotted outline in each pictogram, and the subsequent image is represented by a circle with a solid outline. Quantitative data is added in the form of relative size and distance, and the pictogram is further enhanced by the risk indicator in the form of a shadow.

[0068] The anatomical changes in this example GUI have been analyzed and reported with respect to both the initial image data used for treatment planning and the dose distribution in the treatment plan. The GUI has a view selector in the form of a drop-down menu 611 that allows the user to select a view in subsequent comparisons of the anatomical image data with the dose distribution data (e.g. Figure 5 ) or a comparison of subsequent anatomical structure image data with the first anatomical structure image data (as shown in the ROI-PTV in FIG. Figure 6 Choose between reports with the ROI-CT indicated in the figure. Figure 6 The front view of shows an option in which subsequent anatomical structure image data of the response fractionation is compared with the first anatomical structure image data obtained from the planning CT image.

[0069] In addition to the example of the drop-down menu 611, Figure 6 The GUI 600 also has more controls for selecting information to be displayed as part of the monitoring report. Each pictogram in the data table of the main view 610 allows access to additional data available for the change state, thereby providing independent controls for each change state. For example, data can be accessed by clicking the mouse on the pictogram or navigating in the table using the arrow keys. Figure 6 In the GUI of FIG. 6 , a new additional view 620 is opened when a state is selected. In this example, additional information is shown for the changed state of OAR1 compared to the planning CT prior to delivery of fraction 4 for radiation therapy.

[0070] The additional view 620 shows a natural language description 621 of the changed state supplemented with additional quantitative information, which for this example is "GTV increased by 15% and shifted right by 5". This provides a simple and consistent understanding of what is visually displayed by the pictogram. The physician or clinician is able to explain the pictographic report to a patient or colleague using the natural language description. The additional view 620 also provides further access to more detailed information. In this example, access is provided by three controls in the form of three buttons 622, 623 and 624, but more or fewer buttons and alternative access options such as drop-down menus are also possible.

[0071] Button 622 is a "show image" button, which is configured to open an additional view showing a subsequent image of the object. Preferably, the image also shows anatomical structure image data in the form of OAR contours and TS contours. A particularly advantageous option for this view is to show the subsequent image and its subsequent anatomical structure image data side by side with the first planning image with the first anatomical structure image data. This will allow the user to perform visual inspection and comparison if desired.

[0072] Button 623 is a "dose report" button that is configured to open an additional view showing a dose report. The dose report can be shown in the form of an image, table, or dose volume histogram. The dose report can also be interactive to allow the user to explore and view multiple forms of dose information.

[0073] Button 624 is a "Change Report" button that is configured to open an additional view 630 showing a change report. The change report provides additional quantitative data about subsequent changes in the anatomical structure image data. Figure 6 In the example shown, the additional view shows the report title 631 and the change report in the form of a data table 632 .

[0074] Any method steps disclosed herein can be recorded in the form of a computer program including instructions, which, when executed on a processor, cause the processor to perform the method steps. Instructions can be stored on a computer program product. The computer program product can be provided by dedicated hardware and hardware capable of running software associated with appropriate software. When provided by a processor, these functions can be provided by a single dedicated processor, a single shared processor, or multiple independent processors (some of which can be shared). In addition, embodiments of the present invention can take the form of a computer program product accessible from a computer-available or computer-readable storage medium, which provides program code for use by or in combination with a computer or any instruction running system. For the purposes of this specification, a computer-available or computer-readable storage medium can be any device that can include, store, communicate, propagate, or transmit a program for use by or in combination with an instruction running system, device, or equipment. The medium can be an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system or device or equipment or a propagation medium. Examples of computer-readable media include semiconductor or solid-state memory, magnetic tape, removable computer disk, random access memory "RAM", read-only memory "ROM", hard disk, and optical disk. Current examples of optical disks include compact disk - read only memory "CD-ROM", compact disk - read / write "CD-R / W", Blu-Ray™ and DVD. An example of a communication medium is the Internet or other wired or wireless telecommunication systems.

[0075] Variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, by studying the drawings, the disclosure, and the appended claims.It should be noted that the various embodiments can be combined to achieve further advantageous results.

[0076] In the claims, the word "comprising" does not exclude other elements or steps, and the word "a" or "an" does not exclude a plurality.

[0077] A single unit or device may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

[0078] Any reference signs in the claims should not be construed as limiting the scope.

Claims

1. A system (110) for monitoring anatomical changes in a subject during radiation therapy, the system comprising: an analysis unit (111) comprising an input configured to receive first anatomical structure image data (104) and subsequent anatomical structure image data (107) of the object, the analysis unit being configured at least to register the subsequent anatomical structure data (107) to the first anatomical structure data (104); a change state identification unit (112) configured to identify a change between the first anatomical structure image data (104) and the subsequent anatomical structure image data (107) as a change state, wherein the change state is a predefined category corresponding to the change; a qualitative translator (113) configured to match the identified change state with a corresponding qualitative description, wherein the qualitative description provides an indication of a characteristic of the change state; A reporting unit (114) is configured to provide a monitoring report (500, 600) including a qualitative description of the identified change state.

2. The system (110) of claim 1, wherein: The change state recognition unit (112) includes at least one qualitative spatial reasoning algorithm.

3. The system (110) of claim 2, wherein: The at least one qualitative spatial reasoning algorithm is an RCC-8 calculation algorithm and / or a Cardinal spatial reasoning algorithm.

4. The system (110) according to any one of claims 1 to 3, wherein: The analysis unit (111) further comprises an input unit configured to receive dose distribution data (108) of a treatment plan for the object; and the change state identification unit (112) is additionally or alternatively configured to identify a change between the dose distribution data (108) and subsequent anatomical structure image data (107) as a change state.

5. The system (110) according to any one of claims 1 to 3, wherein: The qualitative translator (113) is configured to match the identified change state with the corresponding qualitative description by using a lookup table.

6. The system (110) according to any one of claims 1 to 3, wherein: The corresponding qualitative description is a natural language description and / or a graphical image.

7. The system (110) of claim 6, wherein: The graphic image is in the form of a pictogram.

8. The system (110) according to any one of claims 1 to 3, wherein: The reporting unit (114) includes a display (115) configured to visually display the monitoring report (500, 600).

9. The system (110) of claim 8, wherein: The display (115) is configured to visually display the monitoring report in the form of a graphical user interface (600).

10. The system (110) of claim 9, wherein: The graphical user interface (600) includes at least one control (611, 622, 623, 624) for selecting information to be displayed as part of the monitoring report.

11. A method (300, 400) for monitoring anatomical changes in a subject during radiation therapy, the method comprising the steps of: receiving (310, 410) first anatomical structure image data and subsequent anatomical structure image data of the object; analyzing the first anatomical structure image data and the subsequent anatomical structure image data (320, 420), wherein the analyzing includes registering the subsequent anatomical structure data to the first anatomical structure data (421); identifying (330, 430) a change between the first anatomical structure image data and the subsequent anatomical structure image data as a change state (431), wherein the change state is a predefined category corresponding to the change; matching the identified change state with a corresponding qualitative description (340, 441), wherein the qualitative description provides an indication of a characteristic of the change state; and A monitoring report (350, 450) is provided that includes a qualitative description of the identified changes.

12. The method according to claim 11, wherein: The step of identifying changes (330) includes supplying the first anatomical structure image data and the subsequent anatomical structure image data to at least one qualitative spatial reasoning algorithm.

13. The method according to claim 12, wherein: The at least one qualitative spatial reasoning algorithm is an RCC-8 calculation algorithm and / or a Cardinal spatial reasoning algorithm.

14. The method according to any one of claims 11 to 13, wherein: The step of receiving (310, 410) first anatomical structure image data and subsequent anatomical structure image data of the object also includes receiving dose distribution data of a treatment plan for the object; and wherein the step of identifying a change (330) additionally or alternatively includes identifying a change between the dose distribution data and the subsequent anatomical structure image data as a change state (432).

15. The method according to any one of claims 11 to 13, wherein: The step (320, 420) of analyzing the anatomical structure image data also includes obtaining quantitative image data.

16. The method according to claim 15, wherein: The quantitative image data includes at least one of the following: the size of the region of interest (422), the change in size of the region of interest (423), the distance the region of interest has shifted (424), and the total radiation dose that the region of interest has received (425).

17. The method according to claim 15, wherein: The monitoring report also includes at least one quantitative description of at least one of the identified changes.

18. An apparatus (100) for medical imaging and analysis, comprising: one or more imaging devices (102, 106) configured to provide images (101, 105) of an object to be treated; a contouring tool (103) configured to provide anatomical structure image data (104, 107) based on said images provided by said one or more imaging devices (101, 105); A system (110) for monitoring anatomical changes in a subject during radiation therapy according to claim 1.

19. A computer program product comprising instructions for causing a processor to perform the method according to any one of claims 11 to 17 when the computer program is executed.

Images