FIELD

[0001] The following relates generally to the medical arts, medical education arts, medical imaging arts, medical image quality arts, and related arts.

BACKGROUND

[0002] For most medical imaging procedures, examination cards or imaging protocols describe different imaging scans or procedures to be performed and are most often re-used for general application. However, some scan settings are patient-dependent or procedure-dependent, and an imaging technologist adapts these settings before each scan, to accommodate for the needs of the patient, procedure or radiologist. This process is not always straightforward and is sometimes based on trial and error (rescans). Moreover, a significant level of experience (learning curve) is required to select optimal settings for specific exams as the purpose and patients differ.

[0003] An imaging technologist is responsible for selecting appropriate scan settings, while an imaging radiologist is responsible for diagnosing the condition based on the image. Sometimes the technologist is not satisfied with the image and will immediately perform a rescan. If the technologist is satisfied with the image and forwards the image to the radiologist, the radiologist still could be dissatisfied. If the radiologist cannot make a (correct) diagnosis on basis of the image(s), the patient has to come back for a rescan, which is undesirable for both the patient as well as the hospital. One of the reasons why a radiologists may not be able to make a (correct) diagnosis is due to poor image quality of the initial scan. In that case, the patient would need to be rescanned, typically with changed settings.

[0004] While the technologist is setting the parameters of a scan, limited guidance is provided on the most appropriate value(s) to be set. Also, protocol-specific imaging settings to conduct multi-site clinical trials are difficult to make consistent for comparison purposes (also against historical data) over sites with individual image improvements.

[0005] The following discloses certain improvements to overcome these problems and others. SUMMARY

[0006] In one aspect, a non-transitory computer readable medium stores instructions readable and executable by at least one electronic processor to perform a method for identifying scan setting errors. The method includes: analyzing medical imaging scans to generate a rescans database of time ordered first and second medical imaging scan pairs meeting a rescan criterion indicating the second medical imaging scan is a rescan replacing the first medical imaging scan; and analyzing the rescans database to identify scan setting error pairs for corresponding scan settings wherein each scan setting error pair comprises an erroneous value or range for the corresponding scan setting in a first medical imaging scan and a correct value or range for the corresponding scan setting in a second medical imaging scan.

[0007] In another aspect, the non-transitory computer readable medium of the immediately preceding paragraph further stores instructions readable and executable by the at least one electronic processor to perform an assistive method for assisting in setting up a current medical imaging scan on an electronic controller of a medical imaging device. The assistive method includes: providing a user interface (UI) on a display device of an electronic processing device via which values for scan settings of the current medical imaging scan are adjustable; matching a current value of a scan setting of the current medical imaging scan to the erroneous value or range of a matching scan setting error pair for the scan setting; and, on the UI, displaying assistive information related to the matching scan setting error pair.

[0008] In another aspect, a non-transitory computer readable medium stores instructions executable by at least one electronic processor to perform a method for identifying scan setting errors. The method includes: analyzing medical imaging scans to generate a rescans database of time ordered first and second medical imaging scan pairs meeting a rescan criterion indicating the second medical imaging scan is a rescan replacing the first medical imaging scan; and analyzing the rescans database to identify scan setting error pairs for corresponding scan settings wherein each scan setting error pair comprises an erroneous value or range for the corresponding scan setting in a first medical imaging scan and a correct value or range for the corresponding scan setting in a second medical imaging scan.

[0009] In another aspect, a non-transitory computer readable medium stores instructions readable and executable by at least one electronic processor to provide assistance in setting up medical imaging scans by performing an assistive method for assisting in setting up a current medical imaging scan on an electronic controller of a medical imaging device. The assistive method includes: providing a user interface (UI) on a display device of an electronic processing device via which values for scan settings of the current medical imaging scan are adjustable; matching a current value of a scan setting of the current medical imaging scan to the erroneous value or range of a matching scan setting error pair for the scan setting; and on the UI, displaying a recommendation to change the current value of the scan setting to the correct value or range of the matching scan setting error pair.

[0010] One advantage resides in reducing a number or frequency of rescans of a patent. [0011] Another advantage resides in increasing image quality of images acquired of a patient.

[0012] Another advantage resides in reducing differences between imaging settings set by a technologist and preferences of a radiologist.

[0013] Another advantage resides in improving a diagnosis of a patient based on improved images acquired of a patient.

[0014] A given embodiment may provide none, one, two, more, or all of the foregoing advantages, and/or may provide other advantages as will become apparent to one of ordinary skill in the art upon reading and understanding the present disclosure.

BRIEF DESCRIPTION OF THE DRAWINGS

[0015] 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.

[0016] FIGURE 1 diagrammatically illustrates an illustrative medical imaging system configured for providing assistance in setting up medical imaging scans in accordance with the present disclosure.

[0017] FIGURE 2 shows exemplary flow chart operations of the system of FIGURE 1.

DETAILED DESCRIPTION

[0018] As used herein, the term “medical imaging examination” or “imaging scan” (and variants thereof) refer to imaging scans performed on the same day on the same patient, usually as part of a single examination order or examination card.

[0019] The following discloses an approach for reducing rescans by leveraging historical rescans as data to determine what imaging scan parameters are typically entered incorrectly in an original scan to provide optimal parameter recommendations and/or automatically configure those optimal parameter recommendations.

[0020] Initially, the disclosed system identifies rescans. There are two types: in-exam or callback. An in-exam rescan is done during an imaging examination in response to poor image quality produced by the original scan, as recognized by the imaging technician or by an on call radiologist who reviews the images while the patient remains in the scanner. A call-back rescan is performed in a subsequent imaging session which is usually scheduled after the reading radiologist objects that the images from the initial exam are not of clinical quality.

[0021] In-exam rescans are generally easy to recognize, as there will be two scans within a single examination that are close to identical. Call-back rescans can be more challenging to recognize, because a second examination that is very similar to a first examination could be either a call-back rescan or a follow-up examination, for example performed after an oncology therapy regimen to assess effectiveness of the treatment. Some approaches for distinguishing these can include the time frame (i.e., a rescan will likely occur faster than a follow up exam), checking the “reason for exam” on the radiology examination order stored in a Radiology Information System (RIS), or checking whether the images of the second examination replace the images of the first examination (rescan) or not (follow-up exam) in a Picture Archiving and Communication System (PACS) storage.

[0022] With a scan and corresponding rescan identified, differences between scan parameters can be identified. There may be hundreds of scan parameters, but the expectation is that only one, two, or at most a few parameters will have been changed between the scan and rescan. There may also be noise, as some rescans may have been done for other reasons, such as a follow-up exam misidentified as a rescan, a scan redone due to patient movement, or so forth. One approach for addressing noise includes collecting enough scan/rescan pairs so that such noise is averaged out. The useful information is expected to include scan parameter changes from scan to rescan that occur frequently in the large scan/rescan dataset. The dataset can be variously segmented, for example pediatric imaging might exhibit more rescans that adjust the parameters to favor shorter scan length to combat patient motion, so analyzing only a pediatric segment of the scan/rescan dataset may be useful. Various statistical and/or artificial intelligence (Al) analyses of the scan/rescan dataset or segments thereof can be used to identify common nonoptimal scan settings in the scans and corresponding optimal scan settings in the rescans. [0023] In one approach the system monitors the scan settings prior to executing an imaging scan and any detected nonoptimal scan setting is flagged and the corresponding optimal scan setting is recommended. In another approach, the optimal scan settings can be used as default settings (i.e., the system can be used to fine tune the default settings).

[0024] While rescans are the predominant information source, other sources may be used. For example, images that are not uploaded to the PACS may be likely to be of low image quality, and so instead of the scan/rescan pairs the data could include (or be augmented by) unsaved/saved image scan settings.

[0025] With reference to FIGURE 1, an illustrative imaging system or apparatus 10 configured for providing assistance in setting up medical imaging scans is shown. The imaging scans are performed by one or more medical devices 12 (e.g., an illustrative medical imaging device 12; or other suitable medical devices such as a radiation therapy device, an image guided therapy (IGT) device; or so forth). By way of some non-limiting illustrative examples, the medical imaging device 12 may be an interventional X-ray (IXR) or other interventional radiology (IR) system (used in combination with at least one interventional instrument in an IGT procedure), a magnetic resonance imaging (MRI) scanner, a computed tomography (CT) scanner, a positron emission tomography (PET) scanner, a gamma camera for performing single photon emission computed tomography (SPECT), or so forth. As shown in FIGURE 1, the system 10 includes, or is accessible by, a server computer 16 typically disposed remotely from the medical device(s) 12 used in the medical procedure for which content is to be generated.

[0026] The medical device 12 includes a non-transitory computer readable medium comprising a database 30 storing log data 32 related to operation of the device and imaging data 34 related to images of one or more patients. The log data 32 are generated by the medical device(s) 12, for example by sensors of the medical imaging device, by a controller of the imaging device, and/or so forth. In one commonly used format, the sensors, controller, and so forth generate timestamped messages that form the log data 32. The database 30 can also store quality metrics and thresholds for procedures that can be identified from the log data 32 to generate a “rescans” database of time ordered first and second medical imaging scan pairs meeting a rescan criterion indicating the second medical imaging scan is a rescan replacing the first medical imaging scan. The database 30 may also comprise multiple databases - for example, the illustrative medical imaging device 12 may generate machine log data as just described that is stored in a machine log database (not shown), and may also generate imaging examination data including the imaging data 34 comprising images and associated imaging device settings that are stored in a PACS database.

[0027] The server computer 16 comprises a computer or other programmable electronic device running software that analyzes the log data 32 in the database 30 to identify potential rescans. The server computer 16 includes or is in communication with a non-transitory computer readable medium 40 storing instructions executable by the server computer 16 to perform (i) a training method or process 100 implemented by the system 10 for training the server 16 to identify typical scan setting errors, and an assistive method or process 200 implemented by the system 10 for assisting in setting up a current medical imaging scan on an electronic controller 18 of the medical imaging device 12. In some examples, the methods 100, 200 may be performed at least in part by cloud processing (that is, the server computer 16 may be implemented as a cloud computing resource comprising an ad hoc network of server computers).

[0028] With reference to FIGURE 2, lefthand side, and with continuing reference to FIGURE 1, an illustrative embodiment of an instance of the training method 100 is diagrammatically shown as a flowchart. At an operation 102, the server computer 16 retrieves the log data 32 and imaging data 34 of medical procedures performed using the device 12 and optionally from other similar devices from the database 30, and the retrieved medical imaging scans are analyzed to generate the rescans database 30 of time ordered first and second medical imaging scan pairs meeting a rescan criterion indicating the second medical imaging scan is a rescan replacing the first medical imaging scan. The similar devices may, for example, be other medical imaging devices of the same (or similar) make and model. To provide a large amount of rescans data, the server computer 16 may, in some embodiments, utilize data from databases 30 of a fleet of medical imaging devices, including the illustrative imaging device 12 and similar imaging devices in the same hospital, or in a network of hospitals, or which are serviced by the same vendor, or so forth.

[0029] In one example, the rescan criterion indicating the second medical imaging scan is a rescan replacing the first medical imaging scan includes (i) a same-patient criterion that the first and second medical imaging scans are performed on the same imaging subject; and (ii) an inexamination rescan criterion in which the first and second medical imaging scans are performed in the same medical imaging examination and a similarity metric comparing the scan settings of the first medical imaging scan and the scan settings of the second medical imaging scan meets a threshold.

[0030] In another example, the rescan criterion indicating the second medical imaging scan is a rescan replacing the first medical imaging scan includes (i) a same-patient criterion that the first and second medical imaging scans are performed on the same imaging subject; and (ii) the second medical imaging scan is performed within a threshold time of the first medical imaging scan and a similarity metric comparing the scan settings of the first medical imaging scan and the scan settings of the second medical imaging scan meets a threshold.

[0031] In another example, the rescan criterion indicating the second medical imaging scan is a rescan replacing the first medical imaging scan includes (i) a same-patient criterion that the first and second medical imaging scans are performed on the same imaging subject; and (ii) a radiology examination order for the second medical imaging scan that indicates the second medical imaging scan is a rescan.

[0032] In another example, the rescan criterion indicating the second medical imaging scan is a rescan replacing the first medical imaging scan includes (i) a same-patient criterion that the first and second medical imaging scans are performed on the same imaging subject; and (ii) the second medical imaging scan replacing the first medical imaging scan in a PACS storage.

[0033] In another example, Potential imaging “rescans” (i.e., subsequent or repeat imaging examinations) can be determined using a similarity metric, such as for example according to Equation 1: where El and E2 refer to first and second imaging examinations and i runs over the (possibly hundreds) of scan settings, values of the scan settings in the corresponding first and second imaging exams, w _t refers to a weighting factor, and N refers to a total number of scan settings.

[0034] These rescan criterion are merely illustrative examples, and should not be construed as limiting.

[0035] At an operation 104, the rescan database 30 is analyzed to identify scan setting error pairs for corresponding scan settings. Each scan setting error pair comprises an erroneous value or range for the corresponding scan setting in a first medical imaging scan and a correct value or range for the corresponding scan setting in a second medical imaging scan. (Note that the terms “erroneous” and “correct” in this context may be inferred based on scan setting pairs satisfying a setting error criterion). In one embodiment, the scan setting error pairs for corresponding scan settings are identified based on occurrences of the scan setting error pairs in the rescans database 30 meets such a setting error criterion. For example, the setting error criterion comprises a ratio or other quantitative comparison of (i) a count of time ordered first and second medical imaging scan pairs in the rescans database 30 having the erroneous value or range for the corresponding scan setting in the first medical imaging scan and the correct value or range for the corresponding scan setting in the second medical imaging scan to (ii) a count of time ordered first and second medical imaging scan pairs in the rescans database 30 having the erroneous value or range for the corresponding scan setting in the first medical imaging scan meeting a threshold. In another embodiment, the setting error criterion comprises an artificial intelligence (Al) analysis (performed by an Al component 42 implemented in the server computer 16) applied to the rescans database to identify the scan setting error pairs for corresponding scan settings. The Al component 42 may be trained for example on labeled training data comprising scan/rescan setting pairs in which a human annotator has annotated some setting pairs as scan setting error pairs. Other embodiments for identifying scan setting error pairs may utilize techniques such as clustering within the database against imaging system type, imaging usage specialization, patient types, or so forth against the correct setting range.

[0036] In the above examples, it is assumed that rescans correspond to medical imaging scan pairs, where the first scan includes the erroneous scan setting (or settings), and the second scan includes the corrected scan setting (or settings). In some situations, there could be multiple rescans, for example if the first rescan does not resolve the problem there may be a second rescan. This situation can be handled in various ways. In one approach, the first scan in time of such a set of three or more scans can be taken as the scan with the erroneous setting(s), while the last scan in time of the set of three or more scans is taken as the correct scan. This approach is motivated by the expectation that any intermediate scans (e.g. the second scan of a set of three scans) still may have errors that were then corrected in the later (e.g. third) scan.

[0037] Referring back to FIGURE 1, the server computer 16 is in communication with an electronic processing device 18, such as a medical device controller of the medical imaging device 12. The illustrative medical device controller 18 includes typical components, such as an electronic processor 20 (e.g., a microprocessor), at least one user input device (e.g., a mouse, a keyboard, a trackball, and/or the like) 22, and a display device 24 (e.g. an LCD display, plasma display, cathode ray tube display, and/or so forth). In some embodiments, the display device 24 can be a separate component from the medical device controller 18 or may include two or more display devices. The electronic processor 20 is operatively connected with one or more non- transitory storage media 26. The non-transitory storage media 26 stores instructions executable by the at least one electronic processor 20 to perform a portion of the assistive method 200. The instructions optionally include instructions to generate a graphical user interface (GUI) 28 for display on the display device 24 that show the scan setting error pairs, the acquired medical images, and so forth.

[0038] Once the scan setting error pairs for corresponding scan settings are identified at the operation 104, the training method 100 ends, and the assistive method 200 begins. (In some variant adaptive training embodiments, the training method 100 may include adaptive training in which results of the assistive method may be fed back to the training method to tune the identification of scan setting pairs). The assistive method 200 provides real-time assistance for a user performing an imaging examination (i.e., a technician) with the medical device 12 by recommending one or more parameters for the medical device 12. With reference to FIGURE 2, and with continuing reference to FIGURE 1 , an illustrative embodiment of an instance of the assistive method 200 is diagrammatically shown as a flowchart. At an operation 202, the server computer 16 is programmed to provide the GUI 28 on the display device 24 of the medical device controller 18 via which values for scan settings of the current medical imaging scan are adjustable based on the scan setting error pairs for corresponding scan settings. At an (optional) operation 204, a default value for a scan setting of the current medical imaging scan is assigned based on the correct value or range of a scan setting error pair for the scan setting being initialized. That is, as the correct value of the scan setting error pair is frequently used in the rescan, it is likely to be the correct value and hence in the optional operation 204 the scan setting is set to this as the default value.

[0039] In addition to, or alternative to, this default setting operation 204, the system can provide real-time assistance to the user in editing the values of the settings of the current medical imaging scan. At an operation 206, a current value of a scan setting of the current medical imaging scan is matched to the erroneous value or range of a matching scan setting error pair for the scan setting. At an operation 208, a recommendation is displayed on the GUI 28 to change the current value of the scan setting to the correct value or range of the matching scan setting error pair. This approach advantageously recognizes in operation 206 that the user is entering a likely incorrect scan setting (since it is the incorrect value, or in the incorrect value range, of the scan setting error pair) and responds by recommending the correct value of the scan setting error pair.

[0040] The default setting operation 204 and the real-time assistance operational sequence 206 and 208, can optionally be combined. In such a combination, the default setting operation 204 provides the user with the likely correct starting (i.e. default) value of a scan setting. If it is detected in the operation 206 that the user then changes this default value to another value (designated here without loss of generality as the value <V>) that falls in the incorrect value or value range of the scan setting error pair, then the operation 208 can for example provide a confirmatory check on this. For example, the operation 208 could present a pop-up notification such as: “The value <V> you are entering often leads to the need for a rescan. Are you sure you want to change this parameter to the value <V>?”. The user’s response to this question provides the desired confirmatory check, and can also serve as feedback for update-training the assistive method 200. For example, each time a user decides to revert to the default value assigned in the operation 204, this is a positive example supporting that the scan setting error pair is reliable; conversely, each time a user decides to proceed with the value <V>, this is a negative example indicating the scan setting error pair may be unreliable. If the balance of such feedback over multiple scans weighs toward negative examples, the scan setting error pair may be deemed to be unreliable and removed from further use in the assistive method 200.

[0041] In some embodiments, the assistive method 200 can include an optional operation in which the user performing the imaging examination with the medical device 12 can provide feedback based on the recommended scan settings. For example, the feedback can comprise a selection of course material (stored in the server computer 16) or other educational material reflective of issues demonstrated in the recommended scan settings. For example, occurrences of a correction of a scan parameter value using the assistive method 200 can be tracked, with various granularities (e.g., per imaging technician, per work shift, per radiology group, et cetera). Individual technicians, or work shifts, et cetera that have a large number of corrections can then be assigned educational material designed to remediate these frequent errors.

[0042] The recommended scan setting changes can also allow a radiologist reviewing the imaging examination to determine if their preferences have a significant weight in determining scan parameters. For example, one radiologist might think a rescan is needed with a selected parameter is at a first value, while a different radiologist might not think rescan is needed, or might order rescan when the selected parameter is at a different value. To implement this approach, radiology department work schedule or other available information is referenced to determine which radiologist is on-call and will be reviewing the images before the patient is released from the imaging examination. (Such reviews are done in some radiology department workflows - such a “real-time” radiologist review is not typically the clinical review by which the radiologist the results in the final radiology report for the exam, but rather is a quality control check to ensure the images that have been obtained are of diagnostic quality before releasing the patient). The assistive method 200 keeps track of the scan setting error pairs on a per-radiologist basis, and thus detects an erroneous value 206 only if the value is erroneous for a scan setting error pair specific to the on-call radiologist. This can not only help the technician determine preferences of different radiologists, but also distinguish between preferences and training requirements.

EXAMPLE

[0043] The following describes the system 10 and the methods 100, 200 in more detail. The system 10 takes a “big-data perspective” on historical rescans in the context of current imaging tasks in order to improve image quality and minimize the need for further rescans, whilst improving diagnostic consistency. This is done by offering guidance to the technologist during the configuration stage of the imaging task.

[0044] The system 10 is configured to analyze the context of each individual imaging exam being scheduled or configured by a user at a particular site. The system 10 is configured to analyze from a (distributed) database of historical rescans, possible root-causes of rescans that are similar to the context of the current imaging task of a user. The root-causes are investigated in the form of (combinations of) parameter values used. The system 10 is configured to determine the likelihood of artefact manifestations based on the currently set settings that could result in the need for a rescan. The system 10 is configured to determine the expected image quality of a scan, given the settings used. This can be determined by an Al algorithm (i.e., the Al component 42), trained on historical data comprising scan parameters/image quality metric. The system 10 is configured to determine what setting changes were historically made during the rescans to avoid a rescan in the current situation. The system 10 is configured to guide the user, via the GUI 28, with recommendations to adapt scan settings to reduce the possibility of rescans and improve image quality and/or the consistency of the quality of the resulting image. The system 10 is configured to take the historical settings and individual settings of imaging devices across multiple sites in order to calibrate to the site settings that provide the best consistency and image balance across the other sites in multi-site clinical trials. When a rescan is required, e.g., due to an unfamiliar context, the system 10 is configured to assist the user to provide feedback to the system in order to help in determining the reason for the rescan (such as poor image quality), which the system can use to learn to better intervene in a similar situation in the future (i.e., a self-learning system). [0045] In one embodiment, the Al component 42 is configured to analyze all the moments when a rescan was made and analyze the changes to the scan settings. This can be used to identify a possible issue root cause, and, in the future to pre-emptively notify the staff before a scan, in order to avoid such re-scans. Rescans are identified from looking at imaging sessions of the same patient within a short period of time, where the same exam with the same scans has been performed, where the individual scans differ in some parameter settings between the two sessions. Such differences include, for example, changes in exam card settings; coil usage; use of breath guidance; the time and amount of contrast fluid and the uptake-time before the exam takes place; the environmental context (i.e., temperature); patient movement during the scan, and so forth.

[0046] In another embodiment, additional context matching parameters are used to compare with historical rescans, including patient-specific aspects such as weight, age, implants, mobility, ability to he still; hospital-specific aspects such as type, size, imaging models and software versions. These parameters help to cluster the context of the rescans better when there are multiple possible root causes to choose from when matching against the context of the current imaging task. [0047] In another embodiment, an expected image quality indicator is used to determine poor choice of parameter settings. The expected image quality indicator can be the output of a machine learning model, which has been trained on historical data comprising scan parameters with good/bad image quality. The image quality for the training data can be determined directly by a software component that directly analyses the image quality and possible artifacts (e.g., the Al component 42), or by looking at the data around images that apparently are not appreciated, such as, for example, image which are not send to a PACS System, re-scans of same sequence/exam card on same patient, radiologist rated images in PACS labeled as insufficient, and so forth. [0048] In some embodiments, the user can be guided to prevent the possibility of rescans and to improve the consistency of the scan, such as, for example, by notifying the technologist just before the start of the scan when the system expects a likely need for a rescan based on similar historic data, an automatic recommendation of what could be changed which the technologist still needs to confirm), limiting what values can be chosen by the technologist, an option to automatically create an extra exam card if the situation occurs often, external sensors that monitor possible external factors that could lead to artifacts and poor image quality, like the usage of a wrong type of light bulb in the imaging room (interference) or if door is not closed properly, and so forth.

[0049] The disclosure has been described with reference to the preferred embodiments. Modifications and alterations may occur to others upon reading and understanding the preceding detailed description. It is intended that the exemplary embodiment be construed as including all such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.