DE102008046346B4 - Method and device for monitoring a spatial area, in particular the environment of a movable medical device - Google Patents

Abstract

Method for monitoring a spatial area (R), in particular the surroundings of a movable medical device (1), in which:- the spatial area is recorded three-dimensionally with a plurality of actively distance-measuring detection means (7, 8, 9, 10), wherein the active distance-measuring detection means (7, 8, 9, 10) actively transmit and receive a signal, wherein the distance to objects is determined from the change in the signal due to reflection or scattering on objects or from the propagation time of the transmitted and received signal;- a three-dimensional environment model is created from the recorded spatial area, which specifies spatial volumes in the spatial area (R) occupied by one or more objects (1, 2, ..., 6, A, D).

Description

The invention relates to a method and a device for monitoring a spatial area, in particular the environment of a movable medical device.

In a large number of technical applications, there is a need to monitor spatial areas with devices located therein, for example to avoid collisions between the devices and other objects in the room or the operating personnel.

In the field of medical technology, medical devices are increasingly being used that carry out automated or semi-automated movements when carrying out diagnostics or therapy. The particular problem here is that due to the high dynamics in medical treatment rooms, e.g. due to movements of the treatment staff or the patient, there is an increasing number of disruptive influences that can cause collisions between the moving medical device and people or objects.

The publication EP 0 087 198 B1 discloses a method for avoiding collisions between two bodies moving relative to each other. For this purpose, the outer contour of the bodies is described using sub-volumes, whereby the distances between the sub-volumes of the bodies are calculated and compared with a safety distance.

In the print DE 297 24 767 U1 discloses a medical device with a device for detecting the position of at least one object located in a room. The device for detecting the position uses a light transmitter to emit a fan of light onto the object and a camera to detect the light reflected by the object. The signals from the camera are fed to an evaluation device, which generates 3D data based on these signals using a triangular 3D technique.

In the document DE 43 35 301 C1 A medical system is shown that avoids a collision of adjustable components while taking moving obstacles into account. For this purpose, television cameras are available to monitor the space of the system, with electronics combining the coordinates of the contours of the components and the obstacles into a multidimensional vector and forwarding them to a three-dimensional neural network that blocks movements that lead to collisions.

The publication EN 10 2007 003 876 B3 describes a method for collision monitoring of a medical device, in which at least one image recording device is used that moves with a component of the medical device. From the current images recorded with the image recording device, a change value is determined by direct comparison with a reference image of a previously recorded image, which change value reflects the degree of deviation between the compared images.

The publication EN 10 2007 002 401 A1 shows a medical examination or intervention device which comprises at least one user-controlled component which can be moved via a manual operating device and a collision monitoring device designed to determine an impending and/or occurred collision. The operating device has a haptic display means which is controlled by the collision monitoring device in order to indicate an impending and/or occurred collision of a component.

The publication EN 10 2004 042 489 A1 discloses a medical examination or treatment device which has an internal and an external position determination system for measuring the position of a tool which is used to carry out a medical examination or treatment and is guided by a robot. The positions of the tool determined by the two position determination systems are compared in a position monitoring system, whereby a restriction of the mobility of the tool is provided if an adjustable threshold is exceeded with regard to the deviation of the tool positions of the two position determination systems. The internal position determination system measures the position of the tool by determining the setting of adjustment elements of the robot. In contrast, the external position determination system measures the position of the tool independently of the robot, preferably with a camera-supported measuring system.

In the document DE 693 27 436 T2 a device with a collision protection arrangement is described, the arrangement comprising power and/or current measuring means for determining values of a power and/or a current. The power or the current is supplied to control means for moving components of the device. The collision protection arrangement repeatedly compares the values of the supplied power and/or the supplied current with relevant reference values, a relevant component of the device is disconnected when a value exceeds the relevant reference value.

In the document EN 10 2004 037 859 A1 describes a medical device with a safety device that includes at least one sensor for detecting an optical signal emitted by a source. The source and the sensor are arranged in such a way that unwanted objects can be detected in an area to be monitored. When an unwanted object is detected, the device is switched to a safe operating state by means of a control unit.

In the document EN 10 2005 001 133 A1 A system is described for positioning devices that can be assigned to a patient to be treated in his or her environment. The devices are recorded in their intended spatial distribution, preferably by at least one camera, whereby this distribution is checked for sufficient freedom of movement of the devices and a corrected distribution is reported back.

The invention is based on the object of presenting a method and a device which quickly and reliably monitor a hazardous area.

This object is achieved by the method according to patent claim 1 or by the device according to patent claim 18. Further developments of the invention are set out in the dependent claims.

In the method according to the invention, the spatial area to be monitored is recorded three-dimensionally using a plurality of actively distance-measuring detection devices. A three-dimensional model of the environment is then created from the recorded spatial area using computer support, which specifies the volume of space occupied by one or more objects in the spatial area. Actively or directly distance-measuring detection devices are those detection devices which actively send out and receive a signal, whereby the distance to objects is determined from the change in the signal due to reflection or scattering on objects or from the transit time of the transmitted and received signal. The signal can be designed in any way, in particular it can be electromagnetic waves (e.g. light in the visible or non-visible range) or sound waves (in particular ultrasonic waves).

The method according to the invention has the advantage that three-dimensional spatial points can be determined reliably and quickly by using several actively distance-measuring detection means, which can be arranged in arrays, for example, so that the occupancy of the space by objects does not have to be determined using computationally complex processing steps from uncertain measurement data. In this way, the creation of an accurate three-dimensional model of the environment is guaranteed in a short computing time.

In the method according to the invention, one or more actively distance-measuring 3D cameras and/or one or more pivoting laser scanners and/or one or more ultrasonic sensors are preferably used as active distance-measuring detection means. For example, distance-measuring cameras from MESA, e.g. the SR 3000 model, or two-dimensional laser scanners, e.g. the URG 04-LX model from Hokuyo or the LMS 200 model from Sick, can be used. The three-dimensional environmental detection is achieved in two-dimensional laser scanners by a pivot drive of the laser scanner, whereby the scanning plane of the laser scanner is moved in space and the space is thus measured three-dimensionally.

In a preferred embodiment, at least part of the spatial area, in particular the entire spatial area, is recorded by actively distance-measuring 3D cameras, and in addition a partial area of the spatial area is recorded by a pivoting laser scanner. In this way, partial areas in which there is a high need for precise environmental modeling can be recorded again separately by a laser scanner, thereby increasing the accuracy of the environmental model in these partial areas. Such partial areas are in particular areas in which increased activity is suspected due to the movement of objects or people.

The actively distance-measuring detection means used in the method according to the invention can be arranged anywhere in the room. In particular, they can be fixed in place in the room. However, it is also possible to arrange at least some of the detection means on one or more movable devices in the room. This makes it possible to detect areas of the spatial environment in particular which cannot be seen by fixed detection means in certain positions of the movable device because they are shaded by the movable device.

In a particularly preferred embodiment, the dynamics of the spatial area are also recorded, which is achieved by continuous monitoring over time. This means that the The spatial area is recorded as an image stream of temporally successive images, whereby the term image stream is to be interpreted broadly and can include any type of three-dimensional measurement data, in particular video images or measuring points from a laser scanner.

In a further embodiment of the method according to the invention, each actively distance-measuring detection device detects the spatial volumes occupied by one or more objects in a respective subsection of the spatial area with a detection probability, whereby spatial volumes in the three-dimensional environment model are specified as occupied, taking into account the detection probabilities of all detection devices. This achieves a suitable calculation of the different measurement data of the detection devices, taking into account their detection probabilities. A spatial volume can be identified as occupied, for example, if the detection probability for this spatial volume exceeds a predetermined value for at least one detection device or for several detection devices.

The individual spatial volumes in the three-dimensional environment model can be designed as desired, but they are preferably specified as cuboid cells. In order to save storage space, unused spatial volumes in the environment model can be combined into one spatial volume.

In a particularly preferred embodiment of the invention, the occupied spatial volumes in the three-dimensional environment model are compared with three-dimensional object types from a database in order to assign objects to the spatial volumes based on the comparison result. In this way, a computer-aided scene description of the environment can be obtained, which can be used, for example, to monitor collisions.

In a particularly preferred embodiment of the method according to the invention, a movable device is monitored in the spatial area, the spatial volumes occupied by the device being identified in the three-dimensional environment model. This identification can be carried out, for example, based on the comparison of the spatial volumes with object types from a database described above, the database containing the device as an object type. Based on such an identification of a movable device, a collision risk between the movable device and other objects in the environment model can be determined in a preferred embodiment of the invention. The determined collision risk can assume values in a continuous value range or in a discrete value range, the values making it possible to state how high the risk of collision is. For example, an increase or decrease in the value can represent an increase in the collision risk. It is possible that the collision risk can also only distinguish between the states "existing risk of collision" and "non-existent risk of collision".

In a particularly preferred embodiment, the determined collision risk is designed in such a way that it is greater the smaller the distance between the device and the other object(s) in the room. The distance can, for example, be the smallest distance between the object and the device resulting from the environment model.

In a particularly preferred embodiment of the invention, the future direction of movement of the device and/or the other object(s) in the room is also determined when determining the risk of collision. A risk of collision is determined if, assuming that the device and/or the other object(s) continue to move, a collision occurs within a certain period of time. In particular, the speed of movement of the objects or the device is also taken into account. The future direction of movement can be determined in various ways. In particular, it can be determined from the current and/or past direction of movement of the device and/or the object(s). However, it is also possible to determine the future direction of movement from information about the movement to be carried out by the device and/or the other object(s). In particular, information from the control system of the device relating to the movement to be carried out can be read out.

If the method according to the invention detects a risk of collision or if the risk of collision exceeds a predetermined level, suitable measures can be taken to avoid the collision. For example, an alarm can be issued to inform the operating personnel, whereupon the operating personnel can stop the device. The movement of the device can also be stopped automatically. Furthermore, the device can, if necessary, carry out an evasive movement to avoid the collision.

The preferred application of the method according to the invention is the monitoring of a medical treatment room with at least at least a movable medical device arranged therein, such as an X-ray device. The X-ray device can in particular be a C-arm, which is described in more detail below.

In addition to the method described above, the invention also relates to a device for monitoring a spatial area, in particular the surroundings of a movable medical device. The device comprises a plurality of actively distance-measuring detection means for three-dimensional detection of the spatial area. In addition, an evaluation unit is provided which creates a three-dimensional environment model from the detected spatial area, which specifies spatial volumes occupied by one or more objects in the spatial area. The device is designed in particular in such a way that each of the variants of the method according to the invention described above can be carried out with the device.

Embodiments of the invention are described below with reference to the attached 1 This figure shows a medical intervention room that is monitored based on an embodiment of the method according to the invention.

The method according to the invention is described below using the example of a medical intervention room R, which is shown in plan view in 1 is shown. In the intervention room, an X-ray measurement is carried out with the aid of an X-ray device in the form of a so-called C-arm 1, whereby the movement of the C-arm is to be monitored. The C-arm 1 comprises a radiation source with a collimator 1a at one end of the arm and a corresponding detector 1b at the other end. The C-arm can be moved or rotated in any direction in space using a positioning device (not shown). A device known from the state of the art can be used as the C-arm, for example the AXIOM Artis dFA model from Siemens AG. However, any other X-ray device can also be used, for example biplane systems with two C-arms, such as the AXIOM Artis dBC model from Siemens AG. Cartesian-guided systems can also be monitored using the method according to the invention, such as the AXIOM Aristos FX Plus model from Siemens AG.

In the treatment room R, a patient table 2 is provided on which the patient to be x-rayed is placed. Depending on the body part to be x-rayed, the C-arm is then moved into the appropriate x-ray position. In the treatment room R, monitors 3 are also schematically indicated, on which the x-ray images taken by the C-arm are displayed. Furthermore, another medical device 4 in the form of an EEG (EEG = electroencephalography) is arranged at the foot of the patient table. During the x-ray images, the treatment staff is protected against x-rays by an x-ray protection shield 5, whereby the x-ray protection shield can be variably positioned using a corresponding positioning device 6. The positioning device is attached to the ceiling of the room R. The current position of a doctor in the treatment room R is shown by an ellipse D and the current position of an assistant by an ellipse A.

The treatment room is monitored by an array of a total of four actively distance-measuring 3D cameras 7, 8, 9 and 10, with the viewing angle of the individual cameras being designated α. Cameras 7 and 8 are positioned on one long side of room R under the ceiling of the room and cameras 9 and 10 are positioned on the opposite long side of the room, also under the ceiling. The cameras capture a part of the room in three dimensions, with the captured part being an area in which increased activity occurs due to the movement of the C-arm or the patient or the operating personnel. 3D cameras from MESA, for example, are used as cameras. Instead of such cameras, any other actively distance-measuring sensors can be used with which the room area can be captured in three dimensions.

The individual cameras are connected to a corresponding evaluation unit 11, which generates a three-dimensional model of the environment from the images captured by the cameras, in which the volumes of space occupied by objects and in particular the C-arm are specified. By capturing an area of space with cameras or sensors that directly measure distances, a reliable and fast determination of this three-dimensional model of the environment is ensured. If necessary, further active distance-measuring sensors can be present in the room to monitor further areas; in particular, sensors can also be attached to moving objects in the room, for example at the two front ends of the C-arm, in order to monitor areas directly in front of the C-arm. If, for example, pivoting laser scanners are used to capture a part of the room in three dimensions, these laser scanners preferably cover those regions of the room in which there is the greatest need for precise three-dimensional modeling due to a suspected high level of activity.

The generated three-dimensional environment model is further processed by the evaluation unit 11 in a suitable manner, for example to detect collisions. The environment model can also be visualized on a corresponding screen for an operator monitoring the room. Based on the generated three-dimensional environment model, the recorded room can be modeled in different ways. The room is preferably shown in a volume grid with cells of variable resolution. The room can be modeled, for example, using appropriate models (in particular geometric models) taking into account the people and objects expected in the room. The models represent objects or arrangements of objects that are present or suspected to be present in the room. These objects are stored in a database and by comparing the database with the environment model, occupied room volumes can be identified as corresponding objects. For example, the modeling can be based on dynamic models, i.e. models that can change their position and orientation in space and their shape over time. Information can be taken into account here that indicates the extent to which the validity of a model decreases over time. In particular, the models approximately describe the surface of the objects, for example in the form of triangles. The models can also be focused on certain regions, for example in a region where there is increased activity, such as the work area of a medical device, or in an area where there is a risk of people or other devices entering the work area.

In one embodiment, the spatial area recorded by the cameras or sensors is divided into spatial volumes of essentially equal size in the form of cuboids. From each measurement by a sensor that directly measures distance, it is immediately recorded in which cuboids the object causing the measurement or its surface is located and with what probability. In addition, a statement is obtained about which cuboids in the room are not occupied and with what probability. By calculating all of this information, which is recorded by the various active distance-measuring sensors in the room, a reliable statement is finally obtained about where objects are located in three-dimensional space. In order to save storage space, neighboring free cuboids (i.e. not occupied by objects) can be combined into a larger cuboid if necessary.

In a preferred embodiment of the invention, it is taken into account that persons and objects in the room are essentially known a priori. In the medical intervention room according to 1 For example, it is known in advance that the patient is lying on the patient couch, and areas are also known in which the treatment staff prefers to move. Models of the suspected people or objects from a model database can therefore be compared with the occupied spatial volumes in which corresponding people and objects are suspected. The occupied cells are thus analyzed to determine which model from the model database they match, so that during the integration of the measurements into the three-dimensional model, not only the occupied spatial volumes but also the objects characterized by them are recognized. The environmental model thus provides a scene description at a higher level, whereupon, for example, with continuous monitoring of the room, corresponding movements of the objects can be recorded and, if necessary, the movement of the objects can also be predicted.

The method according to the invention can also be used in particular for collision avoidance. For example, the evaluation unit 11 in the embodiment of the 1 recognize the C-arm as an object and determine the risk of collision with the other recognized objects based on predetermined criteria. If a collision risk exceeds a certain limit or a risk of collision is detected, appropriate countermeasures can be initiated, for example an alarm can be issued by the evaluation unit. The evaluation unit can also be connected to the control system of the C-arm, whereby the evaluation unit issues a corresponding command to the C-arm in the event of a risk of collision, which causes the C-arm to stop or initiates an evasive maneuver by the C-arm.

Claims (19)

Method for monitoring a spatial area (R), in particular the surroundings of a movable medical device (1), in which: - the spatial area is detected three-dimensionally with a plurality of actively distance-measuring detection means (7, 8, 9, 10), wherein the active distance-measuring detection means (7, 8, 9, 10) actively transmit and receive a signal, wherein the distance to objects is determined from the change in the signal due to reflection or scattering on objects or from the propagation time of the transmitted and received signal; - a three-dimensional environment model is created from the recorded spatial area, which specifies spatial volumes in the spatial area (R) occupied by one or more objects (1, 2, ..., 6, A, D). Procedure according to Claim 1 , in which the spatial area (R) is recorded with one or more actively distance-measuring 3D cameras and/or one or more pivoting laser scanners and/or one or more ultrasonic sensors. Procedure according to Claim 1 or 2 , in which at least a part of the spatial area (R), in particular the entire spatial area (R), is recorded by actively distance-measuring 3D cameras and additionally a partial area of the spatial area (R) is recorded by one or more pivotable laser scanners. Method according to one of the preceding claims, in which the actively distance-measuring detection means (7, 8, 9, 10) are arranged in a fixed position in the room and/or are attached to one or more movable devices (1). Method according to one of the preceding claims, in which the spatial region (R) is captured as an image stream of temporally successive images. Method according to one of the preceding claims, in which each actively distance-measuring detection means (7, 8, 9, 10) detects the spatial volumes occupied by one or more objects (1, 2, ..., 6, A, D) in a respective subsection of the spatial area (R) with a detection probability, wherein spatial volumes in the three-dimensional environment model are specified as occupied, taking into account the detection probabilities of all actively distance-measuring detection means (7, 8, 9, 10). Method according to one of the preceding claims, wherein in the three-dimensional environment model the spatial volumes are specified as cuboid cells. Method according to one of the preceding claims, in which adjacent, unoccupied spatial volumes are combined to form one spatial volume. Method according to one of the preceding claims, in which the occupied spatial volumes in the three-dimensional environmental model are compared with three-dimensional object types from a database and object types are assigned to the spatial volumes based on the comparison result. Method according to one of the preceding claims, in which a movable device (1) is monitored in the spatial area (R), the spatial volumes occupied by the device (1) being identified in the three-dimensional environment model. Procedure according to Claim 10 , in which a collision risk between the identified device (1) and one or more other objects (2, ..., 6, A, D) is determined based on the environmental model. Procedure according to Claim 11 , in which the determined collision risk is greater the smaller the distance between the device (1) and the other object(s) (2, ..., 6, A, D) in the room. Procedure according to Claim 11 or 12 , wherein when determining the collision risk, the future direction of movement of the device (1) and/or the other object(s) (2, ..., 6, A, D) in space is determined, wherein a collision risk is determined if, assuming further movement of the device and/or the object(s) (2, ..., 6, A, D), a collision occurs within a certain period of time. Procedure according to Claim 13 , in which the future direction of movement is determined from current and/or past directions of movement of the device (1) and/or the other object(s) (2, ..., 6, A, D) and/or from information about the movement to be carried out by the device and/or the other object(s) (2, ..., 6, A, D). Method according to one of the Claims 11 until 14 , in which, if the collision risk exceeds a predetermined level, an alarm is issued and/or the movement of the device (1) is stopped. Method according to one of the Claims 11 until 15 , in which, if the risk of collision exceeds a predetermined level, an evasive movement to avoid collision is carried out by the device (1). Method according to one of the preceding claims, in which a medical treatment room (R) with at least one movable medical device (1), in particular an X-ray device, arranged therein is monitored. Device for monitoring a spatial area, in particular the environment of a movable medical device, comprising: - a plurality of active distance-measuring detection means (7, 8, 9, 10) for three-dimensional detection of the spatial area (R), wherein the active distance-measuring detection means (7, 8, 9, 10) are designed such that they actively transmit and receive a signal, wherein the distance to objects is determined from the change in the signal due to reflection or scattering on objects or from the propagation time of the transmitted and received signal; - an evaluation unit (11) which creates a three-dimensional environment model from the detected spatial area, which specifies spatial volumes in the spatial area occupied by one or more objects (1, 2, ..., 6, A, D). Device according to Claim 18 , which is designed in such a way that the device can be used to carry out a method according to one of the Claims 1 until 17 is feasible.

Images