Method for taking a large-area X-ray image

Abstract

Method for providing a large-area X-ray image of an object stored by means of an object support and providing further synthetic X-ray projections, comprising: • Recording of multiple X-ray projections using an X-ray device, wherein the majority of X-ray projections are recorded during a linear movement of the object support and / or the X-ray device; • Creation of at least one tomosynthetic volume from the majority of X-ray projections; • Creation of at least one first synthetic forward projection from the at least one tomosynthetic volume, wherein the at least one synthetic forward projection produces a large-area X-ray image; • Display of the large-area X-ray image on a suitable display device; • Selecting and / or marking an area of ​​interest within the large-area X-ray image; • Create at least one second synthetic forward projection encompassing the area of ​​interest; • Representation of at least one second synthetic forward projection on a suitable display device; characterized in that the at least one second synthetic forward projection is created in a projection geometry that corresponds to the current orientation of the X-ray device in space.

Description

[0001] The invention relates to a method for taking a large-area X-ray image.

[0002] X-ray devices, such as mobile or stationary C-arms, can only image a limited area of ​​a patient on a 2D X-ray image. This area is primarily determined by the size of the X-ray detector used, such as a flat panel detector or an image intensifier, e.g., 20×20 cm. 2 or 30×30 cm 2, specified. In most cases, it is sufficient to take a simple X-ray image. However, there are also applications that benefit from images that exceed the size of the detector used. These include, for example, the visualization of spinal columns and their pathological changes, such as scoliosis (curvature along the longitudinal axis), as well as the tracking of contrast agent injections, for example, along an entire leg or during aortic interventions, which are referred to below as bolus chase or bolus.

[0003] The aforementioned use cases are interventional applications, where the comparison before, during, and after a surgical procedure is of particular interest. In the case of orthopedic applications, a static scene can generally be assumed, so that X-ray images of different areas of the patient can be acquired by changing the image section, particularly by shifting the patient position, for example, a patient couch.

[0004] Regardless of how the change in the image section is implemented, these are initially individual or isolated X-ray images that exist without any relationship to one another and can only be viewed by the user individually or as a sequence. Particularly when the position of the patient bed is continuously changing, it would be advantageous to combine the individual X-ray images into a single, larger image.

[0005] An obvious possibility for generating a large-area X-ray image is known from DE10 2019 001 988 B3. Here, the individual images are simply stitched together, for example using an image registration algorithm, in particular a so-called "stitching" process. However, simply merging at least two X-ray images to create a large-area X-ray image can increasingly lead to artifacts occurring in the overall image to be created. The causes for this can be the unavoidable cone beam geometry of the X-ray machine and the use of planar (flat) detectors in this geometry. When using an X-ray point source, such as in conventional X-ray tubes, this ensures that the corresponding areas in the two images to be merged, for example, are recorded from different angles, thus resulting in different representations of the same objects.The different representations of the same objects therefore mean that the X-ray images can only be brought into registration inadequately or inconsistently using a simple image registration procedure.

[0006] Furthermore, the use of planar detectors in conjunction with an X-ray point source that generates divergent X-ray beams means that objects or patient areas become increasingly distorted the farther they are from the center of the detector. In this case, a situation also arises where a region that is located at the edge of the detector in one image and, for example, in the center of the detector in another image cannot be represented consistently in the overall image.

[0007] Furthermore, if there are changes in the scene during the recording, for example due to the injection of contrast agent, image registration can no longer be carried out successfully, since the temporal change in the scene, for example due to the spread of the contrast agent, no longer results in sufficiently consistent image areas in the image sequence.

[0008] The document DE10 2015 204 957 A1 discloses a tomosynthesis system that displays depth information in color code.

[0009] The document DE 10 2018 212 389 B3 discloses a tomosynthesis system which moves along the patient and in which a collimator changes the cone beam angle depending on the position of the system.

[0010] DE102020209714 discloses selecting an area of ​​interest on a touch display or GUI and then moving the object support and X-ray source to precisely that position. In contrast to the present invention, the object support and X-ray source are moved simultaneously and in different directions.

[0011] EP3669942 also discloses selecting a region of interest on a touch display or GUI and then moving the object and / or X-ray source to precisely that position. However, in this case, multiple regions are selected and marked with an adjustable marker.

[0012] The object of the invention is therefore to provide a method that makes it possible to create an artifact- and distortion-free large-area X-ray image from at least two X-ray images, which can be displayed on suitable display devices, such as screens, displays, and / or computers. These display devices can preferably include a touchscreen, front-end, or GUI function. The object support mentioned below can correspond to a patient couch or any other suitable support device for patients and / or objects.

[0013] The object of the invention is achieved by the features of the independent claim. Advantageous embodiments are specified by the dependent patent claims.

[0014] Preferably, the invention is implemented in software for the purpose of providing a large-area X-ray image. A largely software-based implementation of the method has the advantage that even methods for image acquisition systems previously used for this purpose can be easily retrofitted via a software update to operate in the manner according to the invention. In this respect, the object is also achieved by a corresponding computer program product with a computer program that can be loaded directly into a memory device of an X-ray device, for example, a C-arm X-ray device, with program sections for executing all steps of the method according to the invention when the computer program is executed in the control device.Such a computer program product may, in addition to the computer program, include additional components such as documentation and / or additional components, including hardware components for using the software.

[0015] A computer-readable medium, for example a memory stick, a hard disk, or another portable or permanently installed data storage device, on which the program sections of the computer program that can be read and executed by a computer unit of the control device are stored, can be used for transport to the control device and / or for storage on or in the control device. A connection to a hospital information system connected to a network, to a radiology information system, or to a global network, in which systems the program sections of the computer program that can be read and executed by a computer unit of the control device are stored, can also be used for transport. The computer unit can, for example, have one or more cooperating microprocessors or the like for this purpose.

[0016] To solve the problem, a method is used for recording a large-area X-ray image, in particular of a patient, on an object support by means of an X-ray device which generates X-rays in a fan beam geometry, comprising the following steps: a. Acquiring a first X-ray image using a first acquisition geometry; b. Acquiring at least one second X-ray geometry using a second acquisition geometry, wherein the first X-ray image partially overlaps with the second X-ray image; c. reconstruction of a tomosynthetic volume from the images of at least two X-ray images; d. Creating synthetic forward projections from the tomosynthetic volume, whereby the totality of the synthetic forward projections creates a large-area X-ray image; e. Display of the large-area X-ray image on a display device; f. Marking an area of ​​interest on a suitable display device, g. Creation of a synthetic forward projection of the area of ​​interest as a preview image, based on an evaluation of the relative position of the X-ray machine and the object positioning, so that the preview image is created in an acquisition geometry that corresponds to the current orientation of the X-ray machine in space.

[0017] The method according to the invention is based on a first exposure of an X-ray image which was produced using a first exposure geometry. The exposure geometry is in particular the position (orientation) of the X-ray device relative to the object support, wherein the exposure geometry can be adjusted, for example, by rotating or translating the position of the X-ray device relative to the object support, for example by raising or lowering the X-ray generator and / or the X-ray detector and, for example, by translating the X-ray generator parallel and / or perpendicular in a plane. The method according to the invention is based on the exposure of a second X-ray image, wherein the exposure geometry of the first X-ray image and the exposure geometry of the second X-ray image are different from one another.Preferably, the acquisition geometry of the second X-ray image differs from the acquisition geometry of the first X-ray image by a translational offset along the direction of the object's position. According to the method according to the invention, it is irrelevant whether the change in the acquisition geometry results from a change in the position of the X-ray device or from a change in the position of the object's position (relative position change).

[0018] Subsequently, a tomosynthetic 3D volume is reconstructed from the first and second X-ray images. Unlike 3D computed tomography, the tomosynthetic 3D volume has incomplete angular coverage and thus limited depth resolution. The tomosynthetic volume can be generated using known methods, in particular using filtered or unfiltered backprojection, an iterative or algebraic reconstruction method, or by applying a machine learning method trained for this purpose. Optionally, a calibration of the C-arm, the object support, and / or, in particular, the relative positioning or alignment of both to one another can be used. Alternatively, reconstruction can also be performed without such calibration.

[0019] After reconstructing a tomosynthetic 3D volume, synthetic forward projections are generated from the tomosynthetic volume, preferably in parallel or fan-beam geometry. The totality of all synthetic forward projections combined produces a large-area X-ray image. The resulting large-area X-ray image is then displayed on a display device. It is also possible to perform the forward projection alternatively using cone-beam geometry.

[0020] In alternative embodiments, the method according to the invention comprises a zoom functionality, in particular for the large-area X-ray image, for example for enlarging or reducing areas of the large-area X-ray image, wherein an area does not necessarily have to correspond to one image, but an area can also consist of several X-ray images.

[0021] Subsequently, the method according to the invention provides that, based on an evaluation of the relative position of the X-ray device and the object positioning, a synthetic forward projection is displayed as a preview or preview image on a display device. When calculating the preview image, the relative position of the X-ray device to the object positioning is evaluated such that the preview image at least approximately corresponds to an X-ray image acquired in this position, in particular to the acquisition geometry present at the time. According to the invention, it is therefore provided to create this synthetic forward projection preferably in a cone beam geometry which takes into account the distance between the X-ray focus and the X-ray detector. This is advantageous in order to generate the preview image in a manner that comes as close as possible to an actually acquired X-ray image.A preferred embodiment of the method according to the invention therefore provides for the preview image to be automatically updated when the user adjusts the orientation of the X-ray device in space, so that it continues to reflect the current orientation of the X-ray device.

[0022] To reduce the radiation dose for patients, a preferred embodiment of the method according to the invention further provides for generating a preview image that does not correspond to the current relative position of the X-ray machine and the object support, but rather to one that is particularly advantageous for the selected medical workflow. This allows users to select an area of ​​interest according to their requirements and based on the large-area X-ray image, and subsequently, an automatic, preferably motorized adjustment of the object support and / or the X-ray machine is performed such that the desired acquisition geometry is obtained after the adjustment.

[0023] A further preferred embodiment provides that the calculation of a movement of the X-ray device and / or the object support from the received area of ​​interest is carried out under the proviso that, after execution of the movement, a projection geometry relative to the object results, which subsequently allows the creation of an X-ray projection that at least approximately corresponds to the second synthetic forward projection. The calculated movement is then evaluated to determine whether it is feasible given the kinematic boundary conditions of the X-ray device and / or the object support. Finally, the movement is only carried out if it is feasible. Otherwise, a visual and / or acoustic indication is provided if the movement cannot be carried out.

[0024] In alternative embodiments, however, the above steps are only performed if a safety criterion is met, in particular if, based on the determined relative position, it is at least approximately ensured that all frequencies required for this forward projection are contained in the Fourier representation of the tomosynthetic volume, thus allowing the forward projection to be created largely artifact-free. If this is not the case, a message can be output on the display device indicating that a selected acquisition geometry is unsuitable for creating a preview image.

[0025] In alternative embodiments, the method according to the invention comprises functions for measuring lengths and / or angles on the synthetically generated forward projections, in particular on the large-area X-ray image.

[0026] In alternative embodiments, the method according to the invention can display a sequence of the recorded X-ray images on a display device, for example in the form of a film sequence, preferably combined with a highlighting, in particular a visualization of a part of the sequence which corresponds to the current position of the X-ray device for object storage.

[0027] In alternative embodiments of the method according to the invention, it is provided that based on the current speed of the position change, for example the movement of the patient bed or the movement of the X-ray device or the relative movement between the X-ray device and the patient bed, it can lead to the adaptation of the temporal pulse rate and pulse duration of the X-ray device, preferably in order to determine a sampling rate suitable for tomosynthesis and to prevent or attenuate motion blur in the acquired X-ray images.

[0028] According to the invention, the method is applied in the event of a scene change (scene change) during the acquisition of the X-ray images for the large-area X-ray image. In particular, these scene changes can be caused by the injection of contrast agent(s). Visible artifacts can arise in the acquired X-ray images because inconsistencies arise between the individual X-ray images due to temporal changes, in particular temporally variable contrast agent dynamics. These arise because the scene changes, caused, for example, by the inflow and outflow of contrast agent, are reproduced at different times in the X-ray images.Before the large-area X-ray image is taken, the respective scene changes on the acquired X-ray images are identified, i.e., those image areas that exhibit dynamic activity, for example, in the case of injected contrast agents, the blood vessels, where the blood vessels are at least partially filled with contrast agent. According to the invention, the relative movement of the X-ray machine and the object position is not considered a scene change, as this is treated accordingly during tomosynthesis. Relevant scene changes can be advantageously identified using image processing methods. For this purpose, conventional methods or previously trained machine learning methods can be used, each adapted to the application, for example, to the flow of a contrast agent.Preferably, a set with at least one derived projection, but preferably several, is created from an initial X-ray projection. The derived projections correspond to specific classes of image content, in particular the image content without scene changes and one or more classes of scene changes, for example, blood vessels filled with contrast medium. In these alternative embodiments of the method according to the invention, the identified scene changes are then taken into account as follows: Based on the sets created from the at least one derived projection, at least one tomosynthetic reconstruction is created, specifically based on the derived projections belonging to a specific class. In one possible embodiment of the method according to the invention, a tomosynthetic reconstruction is created exclusively taking into account the class corresponding to those image regions that have not undergone any scene changes, for example, the surrounding anatomy of a blood vessel, in particular bones and / or soft tissue. This tomosynthetic reconstruction forms the basis for creating a large-area X-ray image using forward projection, which can be interpreted as the background.

[0029] In alternative embodiments of the method according to the invention, a tomosynthetic reconstruction can be created based on further derived projections, which only includes a specific class of scene changes, for example, blood vessels with contrast agent. For the method steps described above, it is advantageous to have prior knowledge as to whether scene changes could be present and whether these are to be computationally utilized. This can be achieved, for example, using the selected organ program, so that the method according to the invention will only search for scene changes in the corresponding case. For example, when taking spinal images, no scene changes would be searched for, whereas when acquiring a bolus, for example when contrast agent is injected into the vascular region of a patient, a scene change would be searched for.

[0030] The derived projections can be calculated using various methods, particularly image processing techniques. In the case of regions containing contrast agents, these can be detected, for example, via a change in brightness. Alternatively, dark image regions that are sharply defined and oriented in the direction of movement can be identified as blood vessels containing contrast agents, for example, using a structure tensor. In general, machine learning methods, such as deep learning, particularly convolutional neural networks, can also be trained to identify, for example, areas with a scene change across the image sequence. Coordinate transformation between consecutive images can be performed, in particular, via the evaluation of position, angle, and / or velocity encoders, for example to identify changes in brightness.

[0031] In these embodiments, the synthetic forward projection can then be generated, in particular, in parallel-beam or fan-beam geometry for all tomosynthetic reconstructions. The large-area 2D X-ray image can thus be displayed with and without scene changes, for example, contrast-enhanced blood vessels. In particular, it is possible to switch between the available displays, preferably to display or hide the contrast-enhanced vascular tree or the patient background.

[0032] The method according to the invention is based on a first recording of an X-ray image, which was produced using a first recording geometry. The recording geometry is in particular the position (orientation) of the X-ray device relative to the object support, wherein the recording geometry can be adjusted, for example, by rotating or translating the position of the X-ray device relative to the object support, for example by raising or lowering the X-ray generator and / or the X-ray detector and, for example, by translating the X-ray generator parallel and / or perpendicular in a plane. The method according to the invention is based on the recording of a second X-ray image, wherein the recording geometry of the first X-ray image and the recording geometry of the second X-ray image are different from one another.Preferably, the acquisition geometry of the second X-ray image differs from the acquisition geometry of the first X-ray image by a translational offset along the direction of the object's position. According to the method according to the invention, it is irrelevant whether the change in the acquisition geometry results from a change in the position of the X-ray device or from a change in the position of the object's position (relative position change).

[0033] Subsequently, a tomosynthetic 3D volume is reconstructed from the first and second X-ray images. Due to the process / technical limitations, the tomosynthetic 3D volume has a limited depth resolution. The tomosynthetic volume can be generated using known methods, in particular using filtered backprojection, an iterative or algebraic reconstruction method, or by applying a machine learning method trained for this purpose. Optionally, a calibration of the C-arm, the patient bed, and in particular the relative positioning and alignment of both to each other can be used. Alternatively, reconstruction can also be performed without such calibration.

[0034] After reconstructing a tomosynthetic 3D volume, synthetic forward projections are created from the tomosynthetic volume, preferably in parallel or fan-beam geometry. The totality of all synthetic forward projections combined results in a large-area X-ray image. The resulting large-area X-ray image is then displayed on a display device. In alternative embodiments, the forward projection can be performed alternatively in parallel, fan, or cone-beam geometry.

[0035] The invention is explained in more detail with reference to the following descriptions of the figures. Fig. 1 shows a device according to the invention in the form of a mobile C-arm. Fig. Figure 2 schematically shows the movement of the patient bed. Fig. 3 reveals the synthetic forward projection in parallel and cone beam geometry. Fig. 4 schematically represents the object to be x-rayed including an area of ​​interest. Fig. 5 shows the preview image generated for the area of ​​interest.

[0036] In Fig. 1 schematically shows a device according to the invention in the form of a mobile C-arm 11, which is intended for the implementation of the method according to the invention.

[0037] The C-arm 11 carries an X-ray generator 13 at one end and an X-ray image detector 12, for example, a flat panel detector or an image intensifier, at the other end opposite the X-ray generator 13. The C-arm 11 is motor-controlled and adjustable in several axes. The axes are equipped with sensors to detect the extent of the adjustment.

[0038] Furthermore, the device includes an image processing unit 121, a storage unit 122, a control unit 123, and a network interface 124. By means of the network interface 124, data, for example image data sets and results of the method according to the invention, can be distributed or made available in a network.

[0039] The image processing unit 121 comprises a control unit 122, on which the two- or three-dimensional image data sets used for the inventive method can be stored or loaded. These image data sets can either be loaded from a server or acquired using the C-arm 11 before or during an intervention. Furthermore, the storage unit comprises instructions used for executing the inventive method by means of a computing unit.

[0040] Furthermore, the device can include a GUI with an image output unit (16, 17) and an input unit 19, with which corresponding settings can be made in corresponding organ programs for the image processing unit 121.

[0041] In Fig. 2 schematically shows a device according to the invention in the form of a mobile C-arm 11, together with an object support 3 that is adjustable / displaceable / travelable in several directions of movement 4. The object support 3 can be moved back and forth by the C-arm 11, with its axes having sensors for detecting the extent of the adjustment.

[0042] In Fig.3, the synthetic forward projection can be seen as a large-area X-ray image. After the tomosynthetic volume 2 has been created, one or more synthetic forward projections are created from it, for example, by sending parallel rays 1 through the tomosynthetic volume 2 using software or a computer program to provide the large-area X-ray image or panoramic image 7. On this panoramic image 7 displayed on the screen 16, the region of interest 9 is then selected, and the preview image 10 is calculated for this region, preferably in cone-beam geometry with the aperture angle of the central projection 6. This preview image can then be displayed on a suitable display device 16, with the preview image 10 being created in a projection geometry that corresponds to the current orientation of the X-ray device in space. According to the invention, this display device 16 can be a touchscreen or GUI-capable screen.Now, by means of an input command, the object support and / or the C-arm 11 can be moved directly to this area of ​​interest 9 and an actual X-ray image can be acquired there, which is close to the previously created preview image and which can be displayed according to the invention.

[0043] Figure 4 shows the large-area X-ray image 7 and the selected area of ​​interest 9.

[0044] Figure 5 shows the preview image 10 of the object 8 to be scanned, generated and displayed according to the invention from the area of ​​interest 9. List of reference symbols 1 Parallel projection or projection in parallel ray geometry 2 Tomosynthetic volume 3 Object storage 4 Direction of movement 11 C-arms 12 X-ray image detector 13 X-ray generator 16, 17 Image output unit 19 Input unit 121 Image processing unit 122 storage unit 123 computing unit 124 Network interface 5 Central projection or projection in cone beam geometry 6 Aperture angle of the central projection 7 Large-area X-ray image or panoramic image 8 Object 9 Area of ​​interest 10 Thumbnail

Claims

[1] A method for providing a large-area X-ray image of an object supported by an object support and providing further synthetic X-ray projections, comprising: • Recording a plurality of X-ray projections by means of an X-ray device, wherein the plurality of X-ray projections are recorded during a linear movement of the object support and / or the X-ray device; • Creating at least one tomosynthetic volume from the plurality of X-ray projections; • Creating at least one first synthetic forward projection from the at least one tomosynthetic volume, wherein the at least one synthetic forward projection creates a large-area x-ray image; • Display of the large-area X-ray image on a suitable display device; • Selecting and / or marking an area of ​​interest within the large-scale X-ray image; • Creating at least a second synthetic forward projection comprising the region of interest; • displaying the at least one second synthetic forward projection on a suitable display device; wherein the at least one second synthetic forward projection is created in a projection geometry that corresponds to the current orientation of the X-ray device in space, characterized by that a scene change caused by an object change is taken into account, where a scene change is a change in a scene during the acquisition of the X-ray images for the large-area X-ray image. [2] Method according to claim 1, characterized by that the linear movement of the object support and / or the X-ray device is superimposed by another movement. [3] Method according to one of the preceding claims, characterized bythat the movement of the object support and / or the X-ray device is detected by determining position and / or orientation information and the reconstruction of the at least one tomosynthetic volume is carried out taking this into account. [4] Method according to one of the preceding claims, characterized by that the creation of at least one tomosynthetic volume is carried out by means of a prior calibration of the acquisition geometry. [5] Method according to one of the preceding claims, characterized by that the creation of the at least one second synthetic forward projection takes place in a cone beam geometry. [6] A method according to any one of the preceding claims, comprising: • Calculating a movement of the X-ray device and / or the object position from the received area of ​​interest under the proviso that, after execution of the movement, a projection geometry relative to the object results, which subsequently allows the production of an X-ray projection which at least approximately corresponds to the second synthetic forward projection; • Evaluation of whether the calculated movement is feasible due to the kinematic constraints of the X-ray device and / or the object storage; and • Perform the movement if it is feasible, or not perform the movement and / or issue a notice if it is not feasible. [7] Method according to one of the preceding claims, characterized by that a zoom function is provided for the large-area X-ray image and / or the second synthetic X-ray projection. [8] Method according to one of the preceding claims, characterized by that functions for measuring lengths and / or angles are provided on the display of the large-area X-ray image and / or the second synthetic forward projection. [9] Method according to one of the preceding claims, characterized by that the object area represented by the second synthetic forward projection is at least approximately marked in the representation of the large-area X-ray image. [10] Method according to one of the preceding claims, characterized by that the creation of a tomosynthetic volume is carried out from derived projections which include at least one certain class of scene change and / or exclude at least one certain class of scene change. [11] Method according to one of claims 1 to 10, characterized by that the scene changes are caused by the temporal dynamics of a contrast agent injection. [12] Method according to claim 11, characterized by that at least two tomosynthetic volumes are created, one of these volumes corresponding to the state of the object without contrast agent and another containing only the contrast agent. [13] Method according to one of the preceding claims, characterized by that the majority of X-ray projections and / or projections derived therefrom are displayed within a temporal sequence. [14] Method according to one of the preceding claims, characterized by that the position of the currently displayed X-ray projection and / or derived projection is marked in the display of the large-area X-ray image. [15] Method according to one of the preceding claims, characterized by that the temporal rate for recording the plurality of X-ray projections and / or the pulse duration of the X-ray device is adapted to the movement, in particular to its speed and / or angular velocity. [16] X-ray device, in particular a C-arm for providing a large-area X-ray image of an object stored by means of an object storage and providing further synthetic X-ray projections, comprising the following method steps: • Recording a plurality of X-ray projections by means of an X-ray device, wherein the plurality of X-ray projections are recorded during a linear movement of the object support and / or the X-ray device; • Creating at least one tomosynthetic volume from the plurality of X-ray projections; • Creating at least one first synthetic forward projection from the at least one tomosynthetic volume, wherein the at least one synthetic forward projection creates a large-area x-ray image; • Display of the large-area X-ray image on a suitable display device; • Selecting and / or marking an area of ​​interest within the large-scale X-ray image; • Creating at least a second synthetic forward projection comprising the region of interest; • displaying the at least one second synthetic forward projection on a suitable display device; wherein the at least one second synthetic forward projection is created in a projection geometry that corresponds to the current orientation of the X-ray device in space characterized by that a scene change caused by an object change is taken into account, where a scene change is a change in a scene during the acquisition of the X-ray images for the large-area X-ray image. [17] Computer program product with a computer program which can be loaded directly into a memory unit of a control unit of an X-ray device, in particular a C-arm X-ray device, with program sections to carry out all steps of the method according to one of claims 1 to 15 when the computer program is executed in the control unit of the X-ray device. [18] Computer-readable medium on which program sections are stored which can be read and executed by a computer unit in order to carry out all the steps of the method according to one of claims 1 to 15 when the program sections are executed by the computer unit.

Images