Model-guided imaging for mechanical ventilation

Abstract

A mechanical ventilation evaluation auxiliary device comprises at least one electronic processor; and a non-transitory storage medium storing instructions readable and executable by the at least one electronic processor to execute a mechanical ventilation assessment assistance method, the mechanical ventilation assessment assistance method comprising obtaining an image of a patient (P) undergoing mechanical ventilation; generating or updating a patient-specific lung model of at least one lung of the patient based on the obtained images; simulating a patient's response to mechanical ventilation therapy using the generated or updated patient-specific lung model; comparing the simulated response to an actual response of the patient to the mechanical ventilation treatment; based on the comparison, determining an imaging recommendation for acquiring an image of the at least one lung of the patient; and outputting the determined imaging recommendation.

Description

[0001] The following generally relates to the fields of pulmonology, mechanical ventilators, mechanical respiratory therapy, mechanical ventilator configuration or settings, respiratory modeling, ventilator-induced lung injury (VILI) prevention, and related fields.

[0002] CROSS-REFERENCE TO RELATED APPLICATIONS

[0003] This patent application claims priority under 35 U.S.C. §119(e) to U.S. Provisional Application No. 63 / 438,557, filed on January 12, 2023, the contents of which are incorporated herein by reference. Background Art

[0004] During a patient's mechanical ventilation treatment, clinicians determine the volume of air (e.g., per breath) that the mechanical ventilator will provide to the patient based on the patient's body size. This volume needs to provide adequate ventilation without causing damage to the lungs due to stress (barotrauma), strain (volumetric creation), or shearing due to the cyclical opening and collapse of the alveoli (atelectasis). This type of lung injury caused by the stress and strain imposed by the mechanical ventilator is known as ventilator-induced lung injury (VILI).

[0005] The problem with determining mechanical ventilator settings for preventing lung injury is that the lung is heterogeneous, either intrinsically due to locally varying structural, geometric, and mechanical properties, or secondarily due to localized injury or fluid accumulation caused by disease or infection, such as chronic obstructive pulmonary disease (COPD), pneumonia, edema, COVID-19, fibrosis, etc. This inhomogeneity of the lung can result in local stress and strain concentrations that are much higher than the apparent (e.g., mean or global) stress and strain estimated based on the patient's body dimensions or based on the lumped volume and compliance measured by the mechanical ventilator.

[0006] In some current approaches, solutions for assessing and preventing VILI include constructing a three-dimensional (3D) biophysical model of the patient's lungs based on the patient's computed tomography (CT) expiratory imaging information (see, e.g., Roth, J et al., 2017, "A comprehensive computational human lung model incorporating inter-acinar dependencies: Application to spontaneous breathing and mechanical ventilation", Int. J. Numer. Meth. Biomed. Engng. (2017); e02787). By using this model constructed using the patient's CT imaging, clinicians can virtually test various mechanical ventilator (MV) settings and see what happens in the lungs (e.g., strain distribution in the parenchymal tissue) via simulation. The mechanical properties of the lung tissue (i.e., the stiffness of the alveolar ducts and inter-alveolar connectors) are selected so that the lung model simulates experimental behavior. In this approach, the mechanical properties are not patient-specific and do not vary locally. The model of the patient's lung generated using the patient's CT images is sometimes called a digital twin because it is a digital representation of the patient's physical lung.

[0007] Certain improvements are disclosed below. Summary of the Invention

[0008] In one aspect, a mechanical ventilation assessment assistance device includes at least one electronic processor; and a non-transitory storage medium storing instructions that can be read and executed by the at least one electronic processor to perform a mechanical ventilation assessment assistance method, the mechanical ventilation assessment assistance method including: obtaining an image of a patient receiving mechanical ventilation; generating or updating a patient-specific lung model of at least one lung of the patient based on the obtained image; using the generated or updated patient-specific lung model to simulate the patient's response to mechanical ventilation therapy; comparing the simulated response with the patient's actual response to mechanical ventilation therapy; based on the comparison, determining an imaging recommendation for obtaining an image of at least one lung of the patient; and outputting the determined imaging recommendation.

[0009] In another aspect, a mechanical ventilation assessment assistance method includes, using at least one electronic processor: obtaining an image of a patient receiving mechanical ventilation; generating or updating a patient-specific lung model of at least one lung of the patient based on the obtained image; simulating the patient's response to mechanical ventilation therapy using the generated or updated patient-specific lung model; comparing the simulated response with the patient's actual response to the mechanical ventilation therapy; based on the comparison, determining an imaging recommendation for obtaining an image of the at least one lung of the patient; and outputting the determined imaging recommendation.

[0010] One advantage resides in providing a model of the lung of a patient undergoing mechanical ventilation therapy having mechanical properties of heterogeneous tissue stiffness, patient-specific, and calibrated mechanical properties.

[0011] Another advantage resides in providing a model of a lung of a patient undergoing mechanical ventilation therapy having a tissue map portion corresponding to the mechanical properties of the lung.

[0012] Another advantage resides in providing a map of the lungs of a patient undergoing mechanical ventilation therapy on a display device of a mechanical ventilator, thereby reducing the need for additional computers in the intensive care unit (ICU).

[0013] Another advantage resides in providing a map generated from a model of a patient undergoing mechanical ventilation therapy on a display device of a mechanical ventilator, the map overlaid with a radiographic image of the patient.

[0014] Another advantage resides in selecting a portion of the lung of a patient undergoing mechanical ventilation therapy to be imaged by an ultrasound device rather than a radiation imaging device.

[0015] Another advantage resides in reduced ionizing radiation exposure to patients undergoing mechanical ventilation therapy.

[0016] Another advantage lies in rapid lung ultrasound scanning.

[0017] Another advantage resides in allowing a sonographer, rather than a radiographer, to obtain images of a patient's lungs.

[0018] Another advantage is reduced imaging costs.

[0019] A given embodiment may provide none, one, two, more, or all of the aforementioned advantages, and / or may provide other advantages as will become apparent to one of ordinary skill in the art upon reading and understanding this disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The disclosure may take form in various components and arrangements of components, and in various steps and arrangements of steps.The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the disclosure.

[0021] Figure 1 and Figure 2 An illustrative mechanical ventilation system according to the present disclosure is schematically shown.

[0022] Figure 3 Shown by Figure 1 and 2 An example flow chart of operations suitable for performing the system is provided.

[0023] Figure 4 Schematically shows Figure 3 Example output of the flowchart. DETAILED DESCRIPTION

[0024] As used herein, the singular forms "a", "an" and "the" include plural references unless the context clearly dictates otherwise. As used herein, the statement that two or more parts or components are "coupled", "connected" or "engaged" shall mean that the parts are connected, operate or act together, either directly or indirectly (i.e., through one or more intermediate parts or components), so long as a linkage occurs. Directional phrases used herein, such as, but not limited to, top, bottom, left, right, up, down, front, back and their derivatives, refer to the orientation of elements shown in the accompanying drawings and do not limit the scope of the claimed invention unless expressly stated therein. The words "comprises" or "comprising" do not exclude the presence of elements or steps other than those described herein and / or listed in the claims. In an apparatus consisting of several components, several of these components may be implemented by the same item of hardware.

[0025] Digital twins based on CT imaging can support decisions about personalized protective ventilation scenarios. However, for example, when a patient has been in the ICU for some time, the condition of the patient's lungs may have changed (either improving or worsening, or improving in some aspects but worsening in another), and therefore a new CT scan may need to be performed for diagnostic reasons or mechanical ventilation therapy may need to be adjusted. However, CT scanning of ICU patients, especially when they are sedated, is challenging.

[0026] Disclosed herein are systems and methods for supporting bedside clinicians and care providers in providing safe mechanical ventilation, including guiding clinicians in selecting ventilator settings that will not produce VILI in a particular patient. However, obtaining CT or 3D or multi-view X-ray images of the patient may not be feasible or desirable, for example due to a desire to avoid ionizing radiation exposure of at-risk patients, difficulty locating unconscious patients, the urgency of the clinical emergency, etc. In these cases, the common approach of constructing a patient-specific digital twin for configuring a mechanical ventilator is not feasible due to a lack of access to CT or 3D or multi-view X-ray images of the patient.

[0027] In the embodiments disclosed herein, this type of situation is addressed by constructing a patient-specific digital twin of the patient's lungs based on CT or 3D or multi-view X-ray imaging data of similar patients. Alternatively, if a digital twin of a similar patient is available, the digital twin can be directly adapted to the current patient.

[0028] In some embodiments, the disclosed method uses a digital twin of a patient to determine mechanical ventilator settings. A map of the patient's lungs is generated from the digital twin, where specific locations in the lungs are selected for ultrasound imaging, rather than acquiring additional radiographic images of the patient.

[0029] Some embodiments also provide a user interface (UI) with an easy-to-read display. The model output is translated into actionable clinical decision support (CDS) information. The user interface (UI) of the mechanical ventilator shows the options for the clinician to decide.

[0030] A system for reducing the time required for lung ultrasound is disclosed. A digital patient model generates outputs to support the therapist in managing ventilator settings and deciding on imaging modality, and to guide the sonographer in performing lung ultrasound scans or patches. The model outputs can be displayed in a user interface on a computer or any device connected to the patient. The decision-making algorithm can also use cardiopulmonary model outputs and measurements as input.

[0031] refer to Figure 1 and 2 , shows a mechanical ventilation assessment aid or system 1 comprising a mechanical ventilator 2 for providing ventilation therapy to an associated current patient P. Figure 1 As shown, the mechanical ventilator 2 is connected to a patient breathing circuit 5 to deliver mechanical ventilation to the current patient P. The patient breathing circuit 5 includes typical components for a mechanical ventilator, such as an inlet line (i.e., inspiratory limb) 6, an optional outlet line (expiratory limb) 7 (which may be omitted if the ventilator employs a single-limb patient circuit), a connector or port 8 for connection to an endotracheal tube (ETT) or other patient respiratory interface device, and one or more respiratory sensors (not shown), such as a gas flow meter, a pressure sensor, an end-tidal carbon dioxide (etCO2) sensor, etc. The mechanical ventilator 2 is designed to deliver air, an air-oxygen mixture, or other breathable gas (supply not shown) to the inspiratory limb 6 of the patient breathing circuit 5 at a programmed (and typically time-varying) pressure and / or flow rate to ventilate the patient via the ETT. The mechanical ventilator 2 also includes an electronic controller (e.g., a microprocessor) 13 for controlling the operation of the mechanical ventilator 2, a display device 14 for displaying information about the current patient P and / or the settings of the mechanical ventilator 2 during mechanical ventilation of the current patient P, and a non-transitory computer-readable medium 15 storing instructions that can be executed by the electronic controller 13.

[0032] Figure 1Schematically, a current patient P is shown intubated with an ETT 16 (most of which is internal to the current patient P and is therefore shown in dashed lines). A connector or port 8 is connected to the ETT 16 for operative connection to a mechanical ventilator 2, thereby delivering breathable air to the current patient P via the inspiratory limb 6 via the ETT 16, and delivering exhaled air back to the ventilator (or to another exhaust port) via the expiratory limb 7. The mechanical ventilation provided by the mechanical ventilator 2 via the ETT 16 can treat a wide range of conditions in which the patient P requires breathing assistance or is completely unable to breathe spontaneously, such as various types of lung conditions, such as emphysema or pneumonia, viral or bacterial infections that affect breathing (such as COVID19 infection or severe influenza), cardiovascular conditions in which the current patient P receives oxygen-enriched breathable gas, and the like.

[0033] Figure 1 Also shown is a medical imaging device 18 (also referred to as an image acquisition device, imaging device, etc.). For example, the image acquisition device 18 can be a two-dimensional (2D) X-ray imaging device or an ultrasound (US) image acquisition device. As primarily described herein, the medical imaging device 18 includes a US medical imaging device 18 (i.e., a bedside US imaging device). Ultrasound imaging devices are a typical choice for patients who are poor candidates for CT or 3D or multi-view X-ray imaging due to the risk of ionizing radiation exposure, lack of mobility, or the need for emergency mechanical ventilation. Another option is a 2D X-ray imaging device such as a direct digital radiography (DDR) imaging device, which acquires a single planar X-ray image of the current patient P. Although portable, as primarily described herein, the medical imaging device 18 includes a US imaging device having an ultrasound transducer 20, which is configured to acquire a US image 22 of the diaphragm of the patient P. The US imaging device 18 can be a bedside imaging device. In some embodiments, the electronic controller 13 can be implemented in the US imaging device 18.

[0034] For bedside monitoring of patients to assess the effectiveness of mechanical ventilation and / or the onset of VILI, portable ultrasound advantageously provides low-cost imaging that is easily deployed in the ICU or other patient locations. Coupling the ultrasound imaging device 18 to the patient simply requires positioning the ultrasound transducer 18 (which can be a handheld device or strapped to the patient) on the torso of the patient P close to the lung to be imaged. This can be repeated to continuously image the left and right lungs (or right and left lungs). Advanced ultrasound imaging data analysis techniques, such as lung sliding analysis, can also provide additional clinically relevant information.

[0035] However, ultrasound imaging has several drawbacks for assessing the effectiveness of mechanical ventilation and / or the onset of VILI. Ultrasound can only image the superficial portions of the lungs, and some clinically relevant features may not have strong contrast in ultrasound images. Therefore, ultrasound images are often insufficient for constructing a digital twin of the lungs of a patient P for modeling mechanical ventilation.

[0036] Therefore, it may be desirable to perform (follow-up) X-ray imaging to provide an image of the entire lung and / or supplemental information about the condition of the lung. X-ray imaging can provide stronger contrast for some clinically significant features that may be missed in ultrasound imaging. However, X-ray imaging is less convenient than ultrasound imaging. Ultrasound imaging is local (i.e., it takes a considerable amount of time to obtain an image of the entire lung) and cannot penetrate lung tissue (i.e., it only provides information about the surface structure of the lung). Advanced X-ray imaging modalities such as computed tomography (CT) imaging scanners can provide sufficient 3D tomographic image data to construct a "digital twin" of the lung, but CT scanners are not portable, so bedridden patients must be transported to the radiology department and loaded into the CT scanner for CT imaging. Portable X-ray imaging systems are available that can avoid the need to transport the patient to the imaging device. However, portable X-ray imaging typically only acquires 2D X-ray images (or at most multiple discrete 2D views) and still requires the X-ray tube and X-ray detector to be positioned on opposite sides of the patient so that the X-rays pass from the X-ray tube through the patient to the X-ray detector. Therefore, bedside X-ray imaging typically involves lifting the patient to place a flat-panel X-ray detector under the patient, which is more inconvenient than using a handheld or strapped ultrasound transducer 20. In addition, a C-arm imaging system includes an X-ray tube and an X-ray detector mounted on opposite ends of a C-shaped suspension system. The suspension system can rotate around the patient. Therefore, images can be easily acquired from multiple directions. These can even be reconstructed into 3D images (similar to CT imaging). They are commonly used in surgical procedures. However, many of these systems are mobile and, in principle, can be used at the bedside. However, there are some disadvantages, such as being much larger and heavier than the portable X-ray systems used in today's ICUs. In addition, in order to allow good 3D reconstruction, the X-ray absorption of the bed must not be too high and is more expensive.

[0037] Therefore, it is desirable to use ultrasound imaging to assess mechanical ventilation and / or the onset of VILI daily (or more frequently); however, follow-up X-ray imaging should still be performed when clinically recommended. Furthermore, when using ultrasound imaging for daily assessment, it is desirable to use optimal settings and ultrasound transducer placement to ensure rapid and efficient ultrasound assessment. The embodiments disclosed herein provide automated assistance for determining and providing follow-up imaging recommendations for lung function assessment and early VILI onset detection, such as recommending follow-up ultrasound or X-ray imaging, and if the former, also recommending optimal ultrasound settings.

[0038] In one example, the medical imaging device 18 includes an electronic processor 21 configured to control the ultrasound imaging device 18 to acquire US images 22, and also includes a non-transitory computer-readable medium 23 storing instructions executable by the electronic processor 21 for determining imaging recommendations. The medical imaging device 18 may also include a display device 25 for presenting information including imaging recommendations. In another example, the electronic processor 13 of the mechanical ventilator 2 controls the ultrasound imaging device 18 to receive ultrasound imaging data 22, such as of the diaphragm of the patient P, from a US probe 20. The ultrasound probe 20 may allow for continuous and automatic acquisition of ultrasound imaging data 22.

[0039] The patient P may also be optionally monitored by a patient monitor 26 (only in Figure 2 2 ), such as a multi-function patient monitor that monitors vital signs such as heart rate, respiratory rate, blood pressure, blood oxygen (e.g., SpO2), capnography (i.e., carbon dioxide level in the breathed air), etc. Where available, patient monitoring data acquired by the patient monitor 26 may also be used to determine imaging recommendations. For example, if the vital signs indicate that the patient's heart rate is higher than expected, this may (along with other available data) indicate that lung function is insufficient and follow-up X-ray imaging should be recommended.

[0040] Figure 1 and 2 Also shown is a server computer or electronic processor 30 that stores previously acquired CT (or electrical impedance tomography (EIT), or X-ray, or ultrasound, magnetic resonance imaging (MRI), single photon emission computed tomography (SPECT), or positron emission tomography (PET)) images 32 of patient P, a plurality of previously acquired (i.e., historical) CT scans 34, and / or a plurality of patient-specific mechanical ventilation models 36 for other patients. The patient-specific mechanical ventilation models 36 may include, for example, anatomical models, artificial neural networks (ANNs), or other machine learning (ML)-based models, etc.

[0041] The patient-specific mechanical ventilation model 36 is also referred to herein as a "digital twin" of the lungs. In some mechanical ventilation treatment sessions, there may be rare circumstances when it is desired to develop and apply a protective ventilation regimen with a digital twin model. Provided there is a sufficiently large library of available and accessible imaging and other clinical data, the server computer 30 is configured to search for similar patients, obtain digital twins of similar patients, and personalize the model for the current patient P using patient images 32 adjusted to the current patient P. The patient-specific mechanical ventilation model 36 can be a grid-based model, but does not need to be a grid-based model. The patient-specific mechanical ventilation model 36 includes an imaging-based patient-specific biophysical model that provides an understanding of heterogeneous lung deformation and flow. This can include the use of gridless methods, mathematical models, machine learning, data analysis, etc.

[0042] The non-transitory computer-readable medium 15 of the mechanical ventilator 2 and / or the non-transitory computer-readable medium 23 of the US imaging device 18 and / or the server computer 30 store instructions that are executable by the electronic controller 13 (and / or the electronic processor 21 and / or the server computer 30) to perform a mechanical ventilation assessment assistance method or process 100. Although primarily described in terms of the electronic controller 13 / non-transitory computer-readable medium 15 of the mechanical ventilator 2, the method 100 may similarly be executed by the electronic processor 21 / non-transitory computer-readable medium 23 of the US imaging device 18 and / or the server computer 30.

[0043] refer to Figure 3 , and continue to refer to Figure 1 and 2 An illustrative embodiment of a mechanical ventilation assessment assistance method 100 is illustrated as a flow chart. To begin method 100, patient P is intubated with an ETT 16. At operation 102, while patient P is receiving mechanical ventilation via mechanical ventilator 2, an image 32 of patient P is obtained. To this end, image 32 is obtained via an imaging device (not shown) or from a server computer 30. In some embodiments, the obtained image 32 includes a spectral CT or EIT image. Image 32 may include, for example, a 3D chest CT scan of patient P.

[0044] At operation 104, a patient-specific lung model 36 of at least one lung of patient P is generated or updated based on the acquired image 32. This can be accomplished in various ways. In some example embodiments, one or more lung lesions are identified in the acquired image 32, and the patient-specific lung model 36 of patient P can be generated or updated to include the identified one or more lung lesions. In other example embodiments, fluid in at least one lung is identified in the acquired image 32, and the patient-specific lung model 36 of patient P can be generated or updated to include the identified fluid. In other example embodiments, at least one value of at least one biophysical parameter (e.g., lung volume, airflow, etc.) of at least one lung region of patient P is determined using the acquired image 32, and the patient-specific lung model 36 of patient P can be generated or updated to include the determined at least one value. In other example embodiments, the acquired image 32 (i.e., a spectral CT or EIT image) is used to identify a perfusion distribution in at least one lung of patient P, and the patient-specific lung model 36 of patient P can be generated or updated to include the identified perfusion distribution.

[0045] The model 36 may also include a cardiopulmonary model of the patient to simulate physiological parameters such as gas exchange and vital signs (see, for example, Raju, BI et al., 2021, “Enhanced acute care management combining imaging and physiological monitoring”, U.S. Patent No. 11,166,666). The patient-specific lung model 36 may have multiple “layers”, such as (i) an anatomical model with segmentation (e.g., airways, bronchi, lobes, lesions, etc.). If necessary, a mesh can be created from these segmentations, which in finite element modeling are triangles or polygons for 2D surfaces, or polyhedral meshes for 3D bodies; (ii) a functional model with, for example, ventilation, strain, compliance, or perfusion distribution. The functional model can be constructed using biophysical simulations or image processing techniques such as deformable image registration (Jacobian matrix); and (iii) a 3D map with diagnostic findings such as lesions, pneumothorax, water, etc., as determined by radiological examinations (digital annotations and labeling), CT physics, and / or artificial intelligence (AI).

[0046] At operation 106, the generated or updated patient-specific lung model 36 is used to simulate the response of the patient P to mechanical ventilation therapy using the mechanical ventilator 2. To this end, the server computer 30 is configured to simulate the operation of the mechanical ventilator 2, which delivers mechanical ventilation with current ventilator settings to the lung model 36 to provide the patient P with an expected patient response to the mechanical ventilation therapy. In some examples, the patient-specific lung model 36 simulates ventilation in the lungs and blood flow in the cardiopulmonary system as a function of the mechanical ventilator settings. The simulated MV settings can be changed to simulate the expected response of the patient P to the changed settings (advantageously, without actually applying those "test" settings to the patient P), and the therapist selects the optimal settings by evaluating the simulation output. In another embodiment, the electronic processor 30 can automatically select the optimal settings. For example, a positive end-expiratory pressure (PEEP) setting can be selected so that the local strain in the patient's lungs indicated by the ventilation simulation does not exceed 200%.

[0047] At operation 108, the electronic controller 13 controls the operation of the mechanical ventilator to provide mechanical ventilation therapy to the patient P using the optimal settings selected by the therapist (or automatically selected) to determine the patient's actual response, which can be compared with the simulated response from operation 106. In some embodiments, when the actual response of the patient P to the applied mechanical ventilation therapy does not meet the predetermined criteria, the ventilator settings of the mechanical ventilator 2 applying the mechanical ventilation therapy to the patient P can be automatically updated. The mechanical ventilation therapy with the updated settings can then be applied to the patient P. These operations 102-108 can be repeated until the actual response of the patient P exceeds the predetermined criteria. For example, the patient P is treated (i.e., ventilated) with the selected PEEP, and the response of the patient P is monitored using the patient monitoring system 26.

[0048] At operation 110, an imaging recommendation 40 for acquiring an image of at least one lung of the patient P may be determined based on the results of the comparison operation 108. The determined imaging recommendation 40 may be (i) a recommendation to acquire an ultrasound image 22 of the patient P if the comparison of the simulated response to the patient's actual response to the mechanical ventilation therapy meets predetermined criteria, or (ii) otherwise a recommendation to acquire a different image type (e.g., CT, X-ray, EIT, etc.) of the patient P. This is based on the expectation that if the actual patient response agrees sufficiently well with the simulation, then it is likely that the patient-specific lung model 36 will still be a reasonably accurate representation of the patient's lungs. On the other hand, if the actual patient response deviates significantly from the simulation, then it is likely that the patient-specific lung model 36 will need to be updated, and to this end, the recommendation may be to acquire new X-ray images, which may then be used to update the patient-specific lung model 36 (which is determined by Figure 3In some embodiments, one or more patient-specific ultrasound imaging settings (e.g., spatial locations in the lung to be imaged) may be determined, and the recommendation 40 may include a recommendation to acquire the ultrasound image 22 of the patient using the determined one or more patient-specific ultrasound imaging settings.

[0049] In a specific example, the electronic controller 13 implements a decision algorithm that uses the simulated and measured results as inputs. If the patient P is stable or improving, and there is no significant difference between the simulated and measured outputs (e.g., less than 10%), follow-up with the patient P using ultrasound is recommended. If the patient P is not responding as expected (i.e., as determined / input by the therapist), or if the simulated and measured differ by, for example, more than 10%, follow-up with the patient P using radiographic imaging (X-ray or CT) is recommended to see if the model 36 and / or the therapist is overlooking or missing unexpected events in the patient's lungs, such as exacerbations or infections.

[0050] At operation 112, the determined imaging recommendation 40 may be output (e.g., on the display device 14 of the mechanical ventilator 2 and / or the display device 25 of the US imaging device 25). In some embodiments, the determined imaging recommendation 40 may include a map including one or more regions of interest (ROIs) in the acquired image 32, the map being generated when the actual response of the patient P to the applied mechanical ventilation therapy meets predetermined criteria. The map 40 may include a 3D map with ROIs for lung ultrasound examination (i.e., existing lesions, water or fluid, risk areas (e.g., areas with simulated strain greater than 200%), atelectasis, etc.).

[0051] At operation 114, the generated map 40 is then displayed on the display device 14 of the mechanical ventilator 2 and / or the display device 25 of the US imaging device 25. In some examples, the acquired image 32 may be displayed on the display device 14 of the mechanical ventilator 2 and / or the display device 25 of the US imaging device 25, and the generated map 40 is then overlaid on the acquired image 32. One or more ROIs in the map 40 may be highlighted. The clinician may then provide one or more inputs indicating a selection of one or more of the ROIs (i.e., the clinician may tap the display device screen to select one of the ROIs), and additional information related to the selected ROI may be displayed.

[0052] The 3D map 40 can be projected along with one or more 2D projections / tables of the chest (with locations) and a scan plan to guide the sonographer. The projections / tables can have 4, 6, 8, 10, 12, 24, or even more regions. Regions with ROIs (i.e., lesions or areas of high risk) are highlighted, and ultrasound settings are mentioned.

[0053] In some embodiments, a 3D map 40 with the ROI is projected onto a 2D projection of the chest, with one or more trajectories covering the ROI. The sonographer is then required to wipe the ultrasound probe 20 along the predetermined trajectory. Other methods of displaying the trajectory to the sonographer (e.g., augmented reality, virtual reality, or other graphical means) may also be used to display the trajectory.

[0054] Now refer to Figure 4 , shows an example of the displayed recommendations and map 40. The map 40 may include, for example, a simplified ultrasound scanning plan (8 ROIs), one of which may be highlighted (indicated by a dotted line). From left to right, the recommendations 40 may include: (a) a 2D projection map with (multiple) highlighted regions, (b) a predetermined trajectory for moving the US probe 20, (c) a table with regions and patterns for the ultrasound imaging data 22, and (d) a step-by-step indication of regions and patterns for the ultrasound imaging data 22.

[0055] Return Reference Figure 3 At operation 116, the US imaging device 18 is controlled to acquire one or more US images 22 of at least one of the ROIs displayed on the display device 14 (or the display device 25). In some examples, one or more ROIs in the map 40 may be selected, and the US imaging device 18 may be controlled to acquire US images 22 of the patient P corresponding to the selected ROIs. In other examples, the generated map 40 may be used to update the patient-specific mechanical ventilation model 36. The acquired US images 22 may then be used as a follow-up US examination for updating the model 104, such as Figure 3 Indicated by the return flow arrow 117.

[0056] For example, the sonographer operator scans the highlighted area (or predetermined track) and reports the findings in a table (manually or automatically) in a predetermined format for further processing. For example, the cell "Front Lower" contains three data: "Lung Sliding: Yes;", "Pleural Line: Yes;" and "Lesion: 10 mm". Alternatively, in the case of swabbing, the ultrasound probe 20 is aligned to the patient's chest, and then the image analysis algorithm (e.g., based on machine learning (ML)) automatically analyzes the ultrasound image stream 22 and generates updates for the model 36, such as the new lesion size. The model 36 can also be updated based on the data table. For example, the lesion size increases from 8 mm to 10 mm.

[0057] Over time, as the model 36 becomes more mature, the use of X-ray based imaging (CT, X-ray) can be reduced and the use of ultrasound can be increased. This can be done for each new patient or for new users who are using X-ray as a clinical reference.

[0058] Method 100 can be applied to other clinical fields, such as oncology (i.e., tracking the effects of treatment (chemotherapy, radiotherapy)), long-term lung diseases (e.g., COPD), cardiology, and neurology (i.e., tracking blood vessels and blood flow, such as vascular stenosis, hemodynamics, and structural information), whenever there is a need to track patient progress over time.

[0059] A specific example of method 100 is described below. A patient is admitted to the ICU with COVID-19. A CT scan 32 shows lesions (i.e., bilateral multiple peripheral ground-glass opacities) that can be identified by a radiologist, CT physics, or machine learning (ML). A model 36 is constructed that shows these lesions are free of flow and deformation. The model 36 then simulates the strain distribution in the lungs. Based on the simulation, the therapist sets the PEEP pressure and other settings (volume, oxygen) on the mechanical ventilator 2. The patient responds well (i.e., stable). Based on the simulation (e.g., ROI type 1 - lesions known from the CT scan 32, and ROI type 2: risk areas with no lesions but strain values greater than 2), an ROI map 40 is projected in a 2D scanning scheme and displayed on the display device 14 or 25. The sonographer scans the highlighted chest areas (e.g., ROI type 1: lesion size that may have changed, ROI type 2: lesions that may have been initiated, etc.). The model 36 is updated (i.e., to include existing lesions and new lesions). The model simulation results are compared with measurement data from the mechanical ventilator 2 and the optional patient monitor 26. A decision is made regarding the next imaging modality. If the patient is stable or as predicted, ultrasound imaging is selected. If the patient is unstable or the DT output varies by more than 10% from the measured value, X-ray or CT is selected because a new lesion may be present in an unexpected location. The decision (i.e., recommendation) is displayed on display device 14 or 25.

[0060] The present disclosure has been described with reference to preferred embodiments. Modifications and alterations may occur to others after reading and understanding the preceding detailed description. The exemplary embodiments are intended to be interpreted as including all such modifications and alterations as long as they come within the scope of the appended claims or their equivalents.

Claims

1. A mechanical ventilation assessment assist device, comprising: at least one electronic processor; as well as A non-transitory storage medium storing instructions, the instructions being readable and executable by the at least one electronic processor to perform a mechanical ventilation assessment assistance method, the mechanical ventilation assessment assistance method comprising: Images of patients (P) receiving mechanical ventilation were obtained; generating or updating a patient-specific lung model of at least one lung of the patient based on the acquired images; simulating the patient's response to mechanical ventilation therapy using the generated or updated patient-specific lung model; comparing the simulated response to the patient's actual response to the mechanical ventilation therapy; determining an imaging recommendation for acquiring an image of at least one lung of the patient based on the comparison; and The determined imaging recommendation is output.

2. The apparatus according to claim 1, wherein said generating or updating comprises: identifying one or more lung lesions in the acquired image; as well as The patient-specific lung model of the patient (P) including the identified one or more lung lesions is generated or updated.

3. The apparatus according to claim 1, wherein said generating or updating comprises: identifying fluid in at least one lung in the acquired image; as well as A patient-specific lung model of the patient (P) is generated or updated, the patient-specific lung model including an identified fluid in at least one lung.

4. The apparatus according to claim 1, wherein said generating or updating comprises: determining at least one value of at least one biophysical parameter of at least one lung region of the patient (P) using the acquired images; as well as The patient-specific lung model of the patient including the at least one determined value is generated or updated.

5. The apparatus of claim 1 , wherein the acquired image comprises a spectral computed tomography (CT) or electrical impedance tomography (EIT) image, and the generating or updating comprises: identifying a perfusion distribution in at least one lung of the patient (P) using the acquired spectral CT or EIT image; as well as The patient-specific lung model of the patient including the identified perfusion distribution is generated or updated.

6. The apparatus of claim 1 , wherein the determined imaging recommendation is one of: recommending obtaining an ultrasound image of the patient (P) if the comparison of the simulated response and the patient's actual response to the mechanical ventilation therapy meets predetermined criteria; or Otherwise it is recommended to obtain a radiographic image of the patient.

7. The apparatus of claim 1 , wherein the determining of the imaging recommendation comprises: determining one or more patient-specific ultrasound imaging settings; as well as A recommendation to acquire an ultrasound image of the patient is determined using the determined one or more patient-specific ultrasound imaging settings.

8. The apparatus of claim 7, wherein the determining of the one or more patient-specific ultrasound imaging settings comprises one or more of: One or more spatial locations to be imaged in an acquired ultrasound image of the patient (P) are determined.

9. The apparatus of claim 1 , wherein the instructions further comprise: updating a setting of a mechanical ventilator applying the mechanical ventilation therapy to the patient (P) when the actual response of the patient to the applied mechanical ventilation therapy does not meet a predetermined criterion; as well as The mechanical ventilation therapy with the updated settings is applied to the patient.

10. The apparatus of claim 1, wherein the instructions further comprise: generating a map comprising one or more regions of interest (ROIs) in the acquired image when the actual response of the patient (P) to the applied mechanical ventilation therapy meets a predetermined criterion; as well as The generated map is displayed on a display device.

11. The apparatus according to claim 10, wherein the instructions further comprise: An imaging device is controlled to acquire one or more images of at least one of the ROIs displayed on the display device.

12. The apparatus of claim 11, wherein the instructions further comprise: receiving one or more inputs on the display device, the one or more inputs indicating a selection of one of the ROIs; as well as An imaging device is controlled to acquire one or more images of the selected ROI.

13. The apparatus of claim 10, wherein the instructions further comprise: A patient-specific mechanical ventilation model for the patient (P) is updated using the acquired one or more images of at least one of the ROIs.

14. The apparatus of claim 10, wherein the acquired one or more images of at least one of the ROIs comprises an ultrasound image.

15. A method for assisting in mechanical ventilation assessment, comprising using at least one electronic processor to: Images of patients (P) receiving mechanical ventilation were obtained; generating or updating a patient-specific lung model of at least one lung of the patient based on the obtained image; simulating a response of the patient to mechanical ventilation therapy using the generated or updated patient-specific lung model; comparing the simulated response to the patient's actual response to the mechanical ventilation therapy; determining an imaging recommendation for acquiring an image of at least one lung of the patient based on the comparison; as well as The determined imaging recommendation is output.

Images