METHOD OF ADJUSTING SETTINGS OF A RADIATION IMAGE RECORDING SYSTEM TAKING INTO ACCOUNT PATIENT MOVEMENT

Abstract

Patient movement is tracked and settings of a component of a radiation image recording system are adjusted so that the same relative position of the patient and the component is retained.

Claims

1-11. (canceled)

12. A method of adjusting settings of a component of a radiation image recording system for generating a radiation image of a patient, the method comprising: tracking movement of the patient starting from a time when a tracking triggering condition occurs; and adapting settings of the component taking into account the tracked movement of the patient so that a position of the patient relative to the component is maintained.

13. The method according to claim 12, wherein the component is a collimation area delimiter.

14. The method according to claim 12, wherein the step of tracking the movement of the patient includes using at least one motion sensor.

15. The method according to claim 12, wherein the step of tracking the movement of the patient includes: performing a measurement of a collimation area after the movement of the patient and registering the measurement of the collimation area and an initial measurement before the movement of the patient; and updating settings for the collimation area to generate an updated collimation area so that a position of the updated collimation area relative to the patient is the same as a position of the collimation area relative to the patient before the movement of the patient.

16. The method according to claim 15, wherein the measurement of the collimation area includes a depth measurement.

17. The method according to claim 12, wherein the step of tracking the movement of the patient is initiated upon generating an initiating signal.

18. The method according to claim 12, wherein the step of tracking the movement of the patient is initiated after identifying an adjustment of a collimation area delimiter as a final adjustment.

19. The method according to claim 12, wherein the step of tracking the movement of the patient is initiated when an operator of the radiation image recording system distantiates from the patient.

20. The method according to claim 12, wherein the step of tracking the movement of the patient is initiated when an operator looks away from the patient.

21. The method according to claim 12, wherein the step of tracking the movement of the patient is initiated when a collimation light providing a visual indication of a collimation area is extinguished.

22. The method according to claim 12, wherein the step of tracking the movement of the patient includes positioning markers on or near the patient.

Description

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0093] The invention will now be described in detail for the particular situation in which the position of the radiation source is set and in which a region of interest is determined by means of an x-ray collimator and occasionally changed in accordance with a movement of the patient.

[0094] When an X-ray image of a body part of a patient is to be taken, the patient is positioned with the aid of an operator in a suitable position for x-ray image recording. Depending on the type of examination the patient is positioned on a so-called wall stand in a vertical position or alternatively he is positioned on a supporting table in a horizontal position.

[0095] An intelligent patient analysis is then performed. First the patient is identified. Patient data may be entered in a workstation coupled to the x-ray recording device or they may be retrieved from a radiology information system (RIS).

[0096] Next the patient's weight and length are measured and the patient's body mass index is calculated. From this body mass index the body type of the patient can be derived. In accordance with the patient's body type, the radiation dose adequate for image recording is derived. In addition, the patient's thickness of the specified body part can be derived from the depth measurements from the camera.

[0097] The patient's weight can be measured with a sensor in front of the wall stand or in the support table. The patient's height can be derived from the depth measurements. The height measurements can be done directly or indirectly based on a skeletonization of the depth measurements and after identification of the patient.

[0098] In one embodiment patient data (such as name, photo of patient, length, weight, body mass index) are projected onto the wall of the x-ray recording room and/or on an additional monitor or display device attached to the modality or detachable from the modality so that the operator as well as the patient himself can verify the data. In this way errors can be avoided.

[0099] Next, the settings for the x-ray source are determined and set: if the body part of the patient is known and is successfully tracked with the camera, the position of this body part is mapped from the camera's coordinate system to the coordinate system of the modality and the modality is positioned as best as possible to a position which is optimal for the requested acquisition protocol. In addition the size and position of the collimated area is adjusted based on the size measurements and position of the patient. Additionally, dose acquisition parameters such as kV and mAs can be adapted to fit the patient's physiology as good as possible. Hereby the thickness of the patient's body part, patient's body type and tissue type of the body part to be irradiated can be taken into account.

[0100] Then the position of the x-ray source including the collimator is to be set or fine-tuned so that x-rays emitted by the source of radiation irradiate the region of interest.

[0101] In this specific embodiment, the position of the source of radiation relative to the patient as well as the setting of the collimator blades is controlled by means of hand gestures (possibly non-contact: no contact with the patient, nor the recording device) of the operator and tracking of the change of the location of these hands. It is also possible to position the source of radiation and collimated area with standard input as currently is implemented.

[0102] In order to avoid mistakes when tracking the hand movements of the operator, the operator is first to be identified so that only his hand movements and not these of another person that is present in the room (e.g. the patient) are tracked and used for setting of the location of the x-ray source and the collimator.

[0103] For that purpose in this specific embodiment a picture of the operator is taken by means of at least one of the cameras that is provided in the x-ray room. One camera which has a field of view containing the operator and patient is sufficient but also multiple cameras can be used. If the positioning of the cameras is known with respect to each other or with respect to the modality, the information of the multiple cameras can be merged to create a more detailed image or representation of the room.

[0104] Systems such as Microsoft Kinect and Intel RealSense identify and track persons within a video sequence. So once a person has been identified with face recognition, it is therefore possible to track this person with the associated identification label generated by the person tracking software. One could position a depth sensing camera or regular camera facing the entry of the modality room or positioned at a place where the patient and operator are guaranteed to pass. If a frame of such a camera is good enough for face recognition, the person identification from the face recognition is linked with person identification from the person tracking software.

[0105] The operator can be identified by face recognition and person tracking links. Alternatives are possible, for example on the basis of the location where the operator is standing a difference can be made between the operator and the patient (the patient being the person that lies on the supporting table or that stands on the wall stand, the operator being the person in the room that is not on the supporting table or on the wall stand). If even more persons are present in the room, the operator can be identified and tracked as the first person assisting the person on the supporting table or on the wall stand.

[0106] In a specific embodiment the generation of radiation is prevented when 2 or more persons are detected in a given area.

[0107] Once a person is identified as being the operator, movements of a specified body part made by this person are taken into account for controlling the operation of components of the x-ray recording device. The movement of a body part will be measured and the amount of change of movement or an amount which is proportional to the measured amount will be used to control the positioning of the x-ray source as well as to adjust the collimator settings.

[0108] In order not to take into account body part movement, in this case hand movements, which were not intended to be used for controlling one of the above mentioned components, the movement tracking is only initiated once a tracking start indication is generated and detected.

[0109] Likewise the tracking is stopped once a stop tracking indication is given and detected.

[0110] This indication may have different forms. However one of the described embodiments the tracking start indication is a gesture in which each of the hands poses the thumb and index in an angle of approximately 90 degrees while closing the other fingers and the tracking stop indication is releasing the pose of this gesture.

[0111] In order to delineate a region of interest, the operator forms a rectangle with the fingers of both hands above the region of interest for x-ray imaging on the patient.

[0112] In another described embodiment, the tracking start indication is a gesture where both hands are positioned parallel as flat hands in either a vertical or horizontal plane and the tracking stop indication is the closure of one or both of the hands.

[0113] In order to adjust the width of the collimated area, the operator poses his hands parallel vertically. The distance between the start of this gestures defines the current width of the collimator. If the distance between the hands increases, the width of the collimator also increases proportionally. For example, the width is increased proportionally with the ratio between the width of the collimator and the width between the hands at the start indication moment. Another implementation would be to increase the width of the collimator identical to the increase of the distance between both parallel hands.

[0114] In order to adjust the height of the collimated area, a similar method is implemented for hands positioned parallel horizontally.

[0115] A depth camera provided in the x-ray room records the image of the hands and measures the area. This information is applied to the controller of the x-ray source and collimator and the collimator blades are adjusted so that they delineate an opening for x-rays emitted by the x-ray source to pass through which is proportional to the recorded area. The proportional factor can be the ratio between the area of the collimated area at the start tracking area and the area of the indicated area with the hands. Another possibility is that if the width or height of the indicated area increases or decreases with one cm, the corresponding width or height increases or decreases with one cm or a factor thereof.

[0116] From detection of the start tracking signal to detection of the stop tracing signal movements of the hands are recorded and measured by the depth camera and spatial changes of the hand positions (and consequentially of the area delineated by the hands) are applied to the controller of the x-ray source and the collimator. The collimator opening is adjusted in accordance with the detected and measured spatial change of the hand position.

[0117] Once the stop tracking signal is generated and detected, no tracking of the spatial change of the hand positions is performed anymore and no corresponding further changes are applied to the x-ray source and collimator.

[0118] Visual control by the operator can be obtained by displaying the hand movements on the display device of the operator's work station.

[0119] In order to have an additional check of the location of the region of interest on which x-rays will be projected, visible light is projected from the collimator position onto the patient, said visible light delimiting the region of interest.

[0120] As an alternative, the collimated area can be computed by taking into account the position of the 3D camera, the position of the X-ray source and the measured depth data. The estimated collimation area computed based on the known geometry, can be presented as an overlay on a (color) image from another camera or a (color) image from the visual camera in the same 3D camera system.

[0121] Once the position of the x-ray source and the collimation area are set to the satisfaction of the operator, a radiation image of the patient can be taken.

[0122] According to the invention the movement of the patient after the x-ray source and collimation area are set correctly is tracked. For example, depth measurement data of the collimation area can be taken after the final adjustment of the operator. This depth data can be registered with newly obtained depth measurements. If the registration differs from the initial position, the system can update the X-ray source and collimation area such that the original object of interest is imaged in the same manner. If this is not possible, a warning to the operator can be generated.

[0123] To detect movement of the patient, motion sensors can be used. If motion is detected, the system can track the object that is present in the collimation area.

[0124] Motion sensors are not needed if tracking of the object is done robustly. If the patient stands still, the tracking will detect that there was no movement and will not adjust the settings of the components.

[0125] For such a method, it is essential to determine a condition that triggers the tracking of the object in the collimation area.

[0126] In a first embodiment the movement of the patient is tracked after the x-ray source and collimation area are set correctly by for example taking depth measurement data of the collimation area after the final adjustment of the operator. As soon as the last adjustment has been made, tracking starts. If after this time, a new adjustment is made, tracking is reset completely and tracking starts after the final adjustment.

[0127] In a second embodiment the movement of the patient is tracked when the operator has moved away from the patient, when there is no more contact between operator and patient or when the operator looks away from the patient. This embodiment can be implemented by using a depth camera and person tracking software. As mentioned above, the person closest to the wall stand or lying on the table is the patient. If a second person is detected, tracking is started when this person is more than a given distance away from the patient. If no second person is detected, the tracking should be started already or otherwise is started immediately. From the skeletonization software from eg. Microsoft Kinect, one could also determine the location of the operator's hands. If the operator assists the patient for correct position, his hands are touching or almost touching the patient, hereby guiding or indicating the patient where to position some body parts. If the hands of the operator are a given distance away from the patient, the positioning of the patient is finished. The tracking of the patient can start from this moment. The same principle can be used if the skeletonization data indicates that the operator's face is looking away from the patient.

[0128] In a third embodiment the movement of the patient is tracked when the operator signals the system to track the patient. The signal can be given with a button, foot switch or on a wearable device. The signal can be a voice command, a gesture or given on a user interface in a software program. In this scenario, the operator positions the patient and notifies the system to start tracking with any input device. Examples of input devices are buttons, foot switches, wearable devices, tablets, smartphones, computers, microphones with voice commands, camera's with gestures, gaming controllers, pointers, etc. It is obvious that this list of input devices is not extensive.

[0129] In a fourth embodiment the movement of the patient is tracked when the collimator light is switched off. In a normal scenario, the operator switches on the collimator light to have visual feedback where the exact collimation area will be. To avoid overheating of the X-ray tube, this collimator light is switched off after a given time period. If the operator is satisfied with the position of the patient, he will not turn this light back on. If he is not satisfied, he will switch the collimator light on for further positioning. In such a scenario, the system will start tracking the body parts in the collimation area if the light is switched off, either manually or automatically. If the light is switched on, tracking will stop and will be re-initiated for the new area when switched off again.

[0130] It is obvious that a combination of the above mentioned embodiments is possible. It is also possible that tracking always is done unless any or the inverse of any of the previous conditions is met.

[0131] In another embodiment a notification is given if the system is in a tracking modus. This notification can be a visual indication, an audible notification or a tangible notification. The visual indication can for example be projected on a wall, a led indicator in the wall stand or table, a highlighted icon in a user interface of the workstation. The audible notification can be a beep or a given audible signal which is distinguishable from other audible notifications. A tangible notification can be a vibration on a device like a phone or a haptic alert on for example an Apple Watch.

[0132] It is a goal of the invention to track the object in the collimation area in such a way that after movement of the patient, a similar image of the same object is generated as before the movement. To achieve this goal, data are acquired about the object and newly obtained data about this object are registered and spatial differences are compared and recording system settings (modality settings) are adapted to compensate for these spatial differences.

[0133] The data that are acquired in this process can be taken from various sources. One or more cameras may be used to obtain these data. For example, one or more depth cameras can be used and the depth data are used for registration. Any derivative of these depth data, eg. a skeletonization of a detected person, can be used for registration. If multiple cameras are used, the depth data can be merged in a single point cloud where all further analysis is performed on these merged data.

[0134] It is also possible to use camera's which capture visible light, ultrasound, infrared, X-ray for this purpose or a combination of these techniques.

[0135] The object which is tracked can be all data which is present in the collimation area. It's also possible to use data from a larger or smaller area than the collimation area or to perform some analysis on the data in the collimation area. A form of analysis is for example the body parts of a skeletonization process which overlap with the collimation area. Another form can be a partitioning of the complete captured data where all partitions are tracked which overlap with collimation area.

[0136] Any derivative of the captured data may be used to perform the tracking or registration. A filter can be used that detect invariant features. But also simple generic image processing filters like gradient filters are possible.

[0137] The captured data may also be transformed to another representation. For example, depth data can be transformed to a mesh representation of the surface. In a similar way, depth data is transformed to a model like the skeleton. Simplification of the data will allow faster processing.

[0138] In another embodiment, markers can be placed on or near the patient which are detected with the cameras and for which the displacements are computed. These markers can be visible markers, lead markers which can be imaged with X-rays, magnetic markers for which the locations can be tracked or wearables.

[0139] Several registration techniques can be used to register two data sets. Depending on the result of the registration, the settings of the system can be updated. It is also possible to update the settings of the system, capture new data and verify that the difference between the data set captured at the start of the tracking and the data set captured after the update of the system is minimal.

[0140] A combination of both methods is also possible. Hereby the first registration predicts the update and while the system is adjusting its settings, the differences between the data sets are checked.

[0141] An example of a registration technique is given in Zinsser, Timo and Schmidt, Jochen and Niemann, Heinrich A refined ICP algorithm for robust 3-D correspondence estimation Image Processing, 2003. ICIP 2003. Proceedings. 2003 International Conference on Image Processing, IEEE.

[0142] An example of a measure between two datasets is the Haussdorf distance (https://en.wikipedia.org/wiki/Hausdorff_distance). Other measures may also apply.

[0143] In addition, it is also possible that the modality checks if all acquisition parameters are set correctly. For example, based on the depth measurements and the location of the X-ray source, the system can compute if all active AEC chambers are covered by the patient. If this is not the case, the uncovered AEC chambers can be de-activated or a warning to the operator can be generated, e.g. by display.

[0144] In one embodiment a current position of items in an x-ray room is recorded and movement of parts in the radiology room, e.g. the x-ray source is controlled taking into account the recorded position so as to avoid collision with said items in the radiology room.

[0145] In another embodiment movement of a patient relative to a patient supporting device is tracked and settings, e.g. of an x-ray collimator are adapted taking into account said movement so that the collimation area retains the same relative location to the patient.

[0146] In still another embodiment the generation of radiation is prevented when 2 or more persons are detected in a given area.