US-12622658-B2 - Actuation method for X-ray device and X-ray device

US12622658B2US 12622658 B2US12622658 B2US 12622658B2US-12622658-B2

Abstract

A method for automatic actuation of an X-ray device includes acquiring at least one X-ray image of a body containing at least two objects and/or an object divisible into at least two object sections, segmenting and classify the at least two objects and/or the at least two object sections of the object, and determining an immediately upcoming or current workflow step of the intervention. Information relating to objects or object sections relevant to the determined workflow step is retrieved, at least one object of the at least two objects or at least one object section of the at least two object sections of the object is selected taking the information into account, and a collimator of the X-ray device is automatically adjusted for overlaying the selected object or selected object section taking the information into account. At least one X-ray image is acquired using the thus adjusted collimator and displayed.

Inventors

Marcus Pfister

Assignees

SIEMENS HEALTHCARE GMBH

Dates

Publication Date: 20260512
Application Date: 20221219
Priority Date: 20211220

Claims (17)

1 . A method for automatic actuation of an X-ray device during a medical intervention comprising at least two workflow steps on a body of a patient containing at least two objects, one object that is divisible into at least two object sections, or the at least two objects and the one object that is divisible into the at least two object sections, the method comprising: acquiring at least one X-ray image of the body containing the at least two objects, the object divisible into the at least two object sections, or the at least two objects and the object divisible into the at least two object sections; segmenting and classifying the at least two objects, the at least two object sections of the object, or the at least two objects and the at least two object sections of the object; determining an immediately upcoming or current workflow step of the medical intervention; retrieving information relating to objects or object sections relevant to the determined immediately upcoming or current workflow step, wherein the information includes an indication of which object or which object sections are relevant to the respective workflow step; selecting at least one object of the at least two objects or at least one object section of the at least two object sections of the object taking the information into account; automatically adjusting a collimator of the X-ray device for overlaying the selected at least one object or the selected at least one object section taking the information into account; and acquiring and displaying one or more X-ray images using the adjusted collimator.
2 . The method of claim 1 , wherein the medical intervention is formed by a robot-assisted navigation.
3 . The method of claim 1 , wherein the at least one X-ray image and the one or more X-ray images are formed at least to some extent by fluoroscopic X-ray images.
4 . The method of claim 1 , wherein the objects are formed by instruments, catheters, implants, or guidewires.
5 . The method of claim 1 , wherein the segmenting and the classifying are performed by at least one machine-learning algorithm.
6 . The method of claim 1 , wherein the immediately upcoming or current workflow step is determined automatically.
7 . The method of claim 6 , wherein the immediately upcoming or current workflow step is determined automatically using a machine-learning algorithm or a user query.
8 . The method of claim 1 , wherein indication of which objects or which object sections are relevant to the respective workflow step is prioritized according to relevance.
9 . The method of claim 1 , wherein retrieving the information comprises retrieving the information from a memory unit or from a database where the information is stored.
10 . The method of claim 9 , wherein the information is stored in the memory unit or the database in the form of a list or a lookup table.
11 . The method of claim 1 , wherein automatically adjusting the collimator of the X-ray device comprises automatically adjusting the collimator of the X-ray device taking the information into account, such that essentially only the at least one relevant object or the at least one relevant object section is inserted.
12 . The method of claim 1 , wherein automatically adjusting the collimator of the X-ray device comprises automatically adjusting the collimator of the X-ray device taking the information into account, such that a minimum bounding rectangle that contains a whole of the at least one relevant object or the at least one relevant object section is inserted.
13 . The method of claim 12 , wherein the minimum bounding rectangle is a bounding box.
14 . The method of claim 12 , wherein the minimum bounding rectangle and an adjustable peripheral region are inserted.
15 . The method of claim 1 , wherein the method is repeated depending on progress of the medical intervention, at a start of a new workflow step, at regular time intervals, or based on a user-trigger.
16 . An X-ray device configured to automatically actuate during a medical intervention comprising at least two workflow steps on a body of a patient containing at least two objects, one object that is divisible into at least two object sections, or the at least two objects and the one object that is divisible into the at least two object sections, the X-ray device comprising: an acquisition system comprising an X-ray detector and an X-ray source, the acquisition system being configured to acquire X-ray images; an image processing unit configured to process the X-ray images, the image processing unit having at least one algorithm for segmenting and classifying objects; a collimator configured to overlay image sections; a determination unit configured to detect a current or upcoming workflow step; a calculation unit configured to retrieve information relating to objects relevant to the current or upcoming workflow step, wherein the information includes an indication of which object or which object sections are relevant to the respective workflow step; a selection unit configured to select at least one object or object section taking the information into account; an input unit configured to receive user inputs; a memory unit configured to store data; a display unit configured to display X-ray images; and a system control unit configured to actuate the X-ray device.
17 . The X-ray device of claim 16 , wherein the X-ray device is assigned a robotic system comprising at least one robot control unit and a robot-assisted drive system having a drive and a drive mechanism, and wherein the robot-assisted drive system is configured to move at least one medical object in a hollow organ of a patient based on control signals of the at least one robot control unit.

Description

This application claims the benefit of German Patent Application No. DE 10 2021 214 738.0, filed on Dec. 20, 2021, which is hereby incorporated by reference in its entirety. BACKGROUND The present embodiments relate to automatic actuation of an X-ray device and an X-ray device for performing such a method. Complex operative interventions and procedures are nowadays very frequently carried out within the framework of new technological developments based a minimally invasive approach (e.g., with image monitoring (fluoroscopy) using large C-arm X-ray devices, such as EVAR procedures by angiography systems). Also included therein are procedures in which a robotic system provided for moving objects (e.g., stents, catheters, guidewires, etc.) in the body of a patient is introduced between the hands of the treating practitioner and the patient (e.g., the Corindus CorPath GRX® system). The operational sequences and workflows associated with these complex interventions are likewise increasingly more complex, while the patient's exposure to radiation also increases with the length of the procedures. To provide support in complex workflows, systems that display the respective workflow step are known. Some of these methods recognize the respective workflow step automatically (e.g., from video images (“Machine and deep learning for workflow recognition during surgery” by N. Padoy, Minimally Invasive Therapy & Allied Technologies, Volume 28, 2019, p. 1ff.) or from X-ray images (“Workflow Phase Detection in Fluoroscopic Images Using Convolutional Neural Networks” by N. Arbogast et al., Bildverarbeitung für die Medizin (Image processing for medicine), 2019)). The choice of appropriate acquisition parameters (e.g., tube voltage or image frequency) is important in order to minimize the exposure to radiation for the patient and medical staff during the lengthy procedures, but the correct deployment of a collimator when collimating the irradiated field of view (FoV) is also important. In this connection, proposals for automatic collimation are known, for example, based on registered volumes (e.g., unexamined German application DE 10 2008 049 695 A1) or based on the totality of the detected objects (“A Machine Learning Framework for Context Specific Collimation and Workflow Phase Detection” by M. Alhrishy et al., 15th International Symposium on Computer Methods in Biomechanics and Biomedical Engineering, 2018). However, these approaches are of limited benefit for many procedures (e.g., complex aortic procedures) since the introduced objects (e.g., stents, catheters, guidewires, etc.) fill out the major part of the uncollimated X-ray image, and consequently, the collimation is not specific enough to significantly reduce the radiation exposure. Otherwise, a collimation is typically carried out by an operator (e.g., physician) by manually introducing the diaphragm elements of the collimator based on the current situation shown in the X-ray image. This is time-consuming and leads to an interruption of the clinical workflow. Further, the procedure is non-reproducibly dependent on the particular operator since the particular operator is first to manually identify a region of interest (RoI) around which the collimation is then applied. SUMMARY AND DESCRIPTION The scope of the present invention is defined solely by the appended claims and is not affected to any degree by the statements within this summary. The present embodiments may obviate one or more of the drawbacks or limitations in the related art. For example, within the scope of X-ray monitoring of interventional procedures, a method that provides that an operator may concentrate on the respective workflow step without distraction is provided. As another example, an X-ray device suitable for performing the method is provided. A method of the present embodiments for automatic actuation of an X-ray device during a medical intervention including at least two workflow steps on a body of a patient containing at least two objects and/or containing one object that is divisible into at least two object sections includes acquiring at least one X-ray image of the body containing the at least two objects and/or the object divisible into at least two object sections. The method also includes segmenting and classifying the at least two objects and/or of at least two object sections of the object, determining the immediately upcoming or current workflow step of the intervention, and retrieving information relating to objects or object sections relevant to the particular workflow step. The method also includes selecting at least one object of the at least two objects or at least one object section of the at least two object sections of the object taking the information into account. The method includes automatically adjusting a collimator of the X-ray device for overlaying the selected object or selected object sections taking the information into account, and acquiring and displaying at leas