Methods for processing dose information images

Abstract

Method for automatically controlling loading processes of dose information images (101), comprising the steps of: - Receiving (S101) a data record at a local data storage device (103) comprising a dose information image (101) and a scanner index, wherein the dose information image comprises a digital image file containing dose information, the dose information comprising at least a single dose, an average dose, a total dose and / or at least a scan length; - Matching (S102) the scanner index with a positive list (105) from the central data storage device (103); - Transferring (S103) the data record from the local data storage device (103) to a check directory of the central data storage device (107) if the scanner index is not included in the positive list (105);- Validating (S104) the dose information image (101) from the review directory using optical character recognition (OCR), comprising: scanning the dose information image with OCR to capture the dose information and checking the plausibility and correctness of the dose information; and - entering (S105) the scanner index (S105) into the positive list (105) if the dose information image (101) has been successfully validated.

Description

The present invention relates to a method and system for processing dose information images.In current systems, a user detects a fault in a generation system caused by certain proprietary dose information images. Several manual checks outside the production system are required until a remedy indicates correct support of a dose information image type. There is no controlled or automated support for detecting problems with dose information image types of CT devices of different manufacturers in a production system. In addition, there is no controlled or automated way to reuse dose information images of particular CT scanners once support is announced.There is no information for operators of supported and unsupported dose information images of CT scanners. This makes it difficult to automatically capture unsupported CT scanners and prevents loading of dose information images into the cloud. No operator information about supported or unsupported dose information images is provided for specific scanner types.For example, US 2015 / 0100572 A1 discloses a method for collecting and managing objective information about a plurality of procedures and subjective information about the result of these procedures.US 2013 / 0208966 A1 discloses a method for anonymousizing medical images based on Anonymisierungstemplates.It is the object of the present invention to provide a uniform mechanism which enables a reliable connection of new CT scanners of different manufacturers with regard to dose information image support to a central dose management system.This object is achieved by means of the subject matter according to the independent claims. Advantageous embodiments are the subject matter of the dependent claims, the description and the figures.According to a first aspect, the object is achieved by a method for automatically controlling loading operations of dose information images, comprising the steps of receiving a data record on a local data memory comprising a dose information image and a scanner index; matching the scanner index with a positive list from the central data memory; transmitting the data record from the local data memory into a checking directory of the central data memory if the scanner index is not comprised in the positive list; validating the dose information image from the checking directory by means of a font recognition; and entering the scanner index into the positive list upon successful validation of the dose information image. The checking directory can be formed by a container.In an advantageous embodiment of the method, the data record is transferred from the local data memory into a non-checking directory of the central data memory if the scanner index is included in the positive list. The non-verification directory may be formed by a container.In a further advantageous embodiment of the method, the positive list is stored in the central data memory.In a further advantageous embodiment of the method, the positive list is transmitted to the local data memory.In a further advantageous embodiment of the method, the checking directory is scanned cyclically after new data sets.In a further advantageous embodiment of the method, an uploading of data sets to a non-checking directory is blocked if the scanner index is not included in the positive list.In a further advantageous embodiment of the method, a validation state is assigned to the data record.According to a second aspect, the object is achieved by a system for automatically controlling loading processes of dose information images, comprising an uploader for receiving a data record on a local data memory, which comprises a dose information image and a scanner index, for matching the scanner index with a positive list from the central data memory and for transmitting the data record from the local data memory into a checking directory of the central data memory if the scanner index is not comprised in the positive list; and a validater for validating the dose information image by means of a font recognition and for entering the scanner index into the positive list if the dose information image is successfully validated.In another embodiment of the system, the system is suitable for entering the scanner index into the positive list if the validation of the dose information image was successful.According to a third aspect, the object is achieved by a computer program product loadable into the memory of a computer and comprising code portions for performing the method according to any of the first aspects, when the computer program product is executed by the computer.Exemplary embodiments of the invention are illustrated in the drawings and are described in more detail below.The following are shown: FIG. 1 shows a block diagram of a method; FIG. 2 shows a dose information image; FIG. 3 is a schematic diagram of a system; and FIG. 4 is a flow chart.FIG. 1 is a block diagram showing automatic control of loading operations of dose information images (black images). The dose information images are digital image files including information on a radiation dose, date, time and / or study ID.In a step S 101, an uploader (uploader) checks in a local data memory whether dose information images have been received from a known CT scanner, which is part of an entry in a positive list ("white list") of CT scanners. The CT scanner is identifiable, for example, by a scanner index that includes information about an operator, a manufacturer, a manufacturer model, a secondary capture type ("secondary capture type"), a software version, or a device serial number.In a step S 102, the uploader terminates an upload of dose information images to the central data store ("cloud") if these have not been received from a known CT scanner. Whether the CT scanner is known can be determined from a match between the positive list and the scanner index.In step S 103, the uploader transmits a single dose information image together with the associated scanner index to the central data memory and stores this in a checklist container "Greylist container" for unknown CT scanners. The checklist container may be filed per operator of the CT scanner or a global checklist container in the central data store that stores a single original DICOM file per scanner version. In step S 104, an automatic validation method is performed for the dose information images in the checklist container. For this purpose, the checklist container is scanned using additional tools. In this validation, the dose information image is scanned by means of automatic font recognition ("OCR").Each CT examination is based on two fundamental equations for dose calculation.The dose information images comprise all the components of these equations as well as the total dose in pictorial form. Therefore, the mechanism uses the following preconditions to check the plausibility and correctness of a dose information image with respect to the dose information.All the individual components for the two equations are present in the dose information image.The total dose is part of the dose information image.The number of individual doses is part of the dose information image.Consequently, there are two possibilities for mathematically checking the result of a previously performed text recognition with regard to plausibility.The sum of all individual doses must result in the total dose.The product of average dose and scan length of each single dose summed over all single doses must also give the total dose.These two checking possibilities make it possible to sufficiently check plausibility that the writing recognition mechanism has provided correct results for the dose information image.If the above-mentioned conditions are therefore checked and approximately the same value for the total dose is obtained mathematically, which value has been obtained on the basis of automatic character recognition, error-free character recognition can be assumed. However, if deviations are present which are above a predetermined value, it is assumed that the automatic identification of writing has produced erroneous results and the dose values cannot be used reliably. A positively validated CT scanner is entered into the positive list ("white list") and is visible in a positive list / negative list tool. Only positively validated CT scanners may upload dose information images to the central data store.FIG. 2 shows a dose information image 101 from a CT scanner. The dose information images 101 are generated by computed tomography (CT) scanners. Depending on the manufacturer of the computed tomography device, this can comprise different information in different arrangements and with different types of writing. The dose information image 101 is a secondary aspect ratio and is stored in a DICOM file with associated media type. The dose information image 101 is for the purpose of being read by humans to acquire the dose information.FIG. 3 shows a system 100 for automatically controlling loading operations of dose information images 101. The system comprises a CT scanner 111 for medical X-ray examination. The dose information image 101 including the aforementioned information is transmitted from the CT scanner 111 in a data set together with the scanner index for identifying the CT scanner 111 to a local data storage 103, such as a hospital image storage and communication system ("PACS").In the local data memory 103, the uploader 109 is implemented by a computer program which is capable of transmitting the data records of the CT scanners 111 to a central data memory 107 ("cloud"). Local data storage 103, such as a hard disk, may be part of a computer system having a processor and random access memory. However, this only occurs if a comparison of the scanner index with a positive list 109 reveals that the CT scanner is known.If this is not known, i.e. if the scanner index is not on the positive list, only a single data record is transmitted into a checking directory of the central data memory 107, which contains a dose information image and a scanner index. The dose information image 101 is automatically validated in the central data memory 107. If validation by the validater 113 is successful, the scanner index is entered into the positive list. Central data store 107, such as a hard disk system, may be part of a computer system having processors and random access memory. The central data store comprises a verification directory 117 and a non-verification directory 119. A) Greylist uploader in the application (receiver) a) Generating a scanner index for an operator's CT scanner ("tenant") from a particular scanner manufacturer. b) Checking (reading) the scanner index (if present) for a manufacturer's y CT scanner x and running at an operator of the CT apparatus with a particular software version against a positive list / negative list ("whitelist / blacklist") of stored scanner indices that can assume different states, for example "unknown"-"unsnown", "in process"-"in work", "unsupported"-"bad", "Fully supported" - "ok". c) set the scanner index to "unknown" if the scanner index is not present in the positive list in the central memory and simultaneously one-time upload of a dose information image 101 to the verification directory in the central data memory 107. d) stop the upload of dose information images 101 coming directly from CT scanners or from the operator's PACS system to the regular non-verification directory in the central data memory 107 as long as the scanner index is not in the status "ok". e) Upon each new arrival of a dose information image 101, this check is carried out again, so that a change to the status "ok" is immediately noted, thereby storing the dose information image 101 in the regular non-verification directory ("study container") is (resumed). The application's greylist uploader has no influence on it, but registers changes immediately and then immediately switches the processing of the dose information image 101 for the corresponding scanner index. B) Greylist downloader as central service or in separate computer a) Cyclic checking of the central positive list whether the scanner index is present, which is in the state "unknown". b) If so, the scanner index is read and its state is set to "in work", and the data associated with the scanner index is loaded from the central data memory 107 by means of a reference in the scanner index (reference) into a memory which can be accessed by a subsequent validation. C) Validation of the dose information image 101 a) Accessing the dose information image 101 loaded from the verification directory and performing automatic dose verification. b) The dose verification reads the data of the dose information image 101 and generates by means of OCR the dose data contained therein. c) By horizontal and vertical checksums plausibility checks are performed on the dose data to the effect that their correctness can be verified. d) As a result of this verification, the status of the scanner index in the central data memory 107 is set either to "ok" or "bad". e) Furthermore, the results of the verification are stored as text file back in the verification directory of the central data memory 107 under the entry referenced by the scanner index, so that they can be tracked at any time from another side.The individual actions are performed by the independent software components as follows.0) The application / receiver receives a dose information image 101 directly from CT scanner 111 or requests it actively from a PACS system at the operator and forms therefrom the scanner index from the DICOM data of the dose information image 101, which forms the CT scanner 111 of the scanner manufacturer with a serial number and a software version because it is in operation at this uniquely defined operator and from this operator, either directly or indirectly via PACS dose information images 101, are obtained periodically, for example 50 per day.1) The application / recipient checks, by means of a central positive list / negative list, whether the scanner index resulting from the currently processed dose information image 101 is in the state "ok" or is present at all. If this is not present, the scanner index is newly entered under the state "unknown". The dose information image is stored in the central data memory 107 and a reference thereto is likewise stored in the scanner index in the central index memory. In addition, the storage of dose information images 101 of this CT scanner 111 and operator is suppressed until after each new arrival of a dose information image 101 of this CT scanner is in the status "ok" when checking in the central index memory. From this point in time, all dose information images 101 of the CT scanner 111 of this operator are then loaded again into the regular non-checking directory ("study container") of the central data memory 107, because it can now be assumed that the dose information images 101 contain correctly determinable dose values.2) Regardless of the operator's processing at the application / receiver, there is a downloader 115 which permanently checks the central index store to see if there are scanner index entries in the "unknown" state that are pending checking by a central service. If such entries are found, the associated data of the dose information image 101 of the checking directory 117 in the central data memory 107 can be accessed by means of the reference stored in the scanner index entry and this can be copied-if necessary-into an environment which can be accessed by the validation step for checking the dose. This may be either in the central data store 107 or in another computer or virtual machine.3) The dose verification of a dose information image 101 is then performed by a validation program. This check is fully automated. The validater determines whether the dose check on this dose information image was ok or bad.4) The result of this automatic check is entered by the validation mechanism into the central index under the entry of the scanner index. Furthermore, it stores the data which have been generated during the dose checking in the checking directory 117 (known by reference in the scanner index) of the central data memory 107.5) Then, the application / recipient at the operator determines, at the next received dose information image 101 of this CT scanner 101, that upon routine examination of the scanner index, its state has changed. If the state has changed to "ok", the uploader of the operator's procedure application / recipient will change in such a way that from now on dose information images 101 for this CT scanner 111 are sent into the regular non-checking directory 119 of the central data memory 107 and are enabled for further processing and evaluation of the dose information.6) If, in the further course of the operator, the relevant CT scanner 111 is provided with a new software version, then the scanner index changes automatically since the software version of the CT scanner is part of the scanner index.This automatically results in the described method beginning again at the first step. This is helpful because by changing the software version of the CT scanner 101, the font associated with the dose information image could also be changed, or dose information images 101 could be switched from black / white to color. This would then no longer ensure that the dose check carried out is still valid. The described automatic method can therefore continue to operate correctly even in changed environmental conditions, in that a new dose check via the checking directory 117 temporarily (until the new dose check succeeds) temporarily prevents the regular processing via the non-checking directory.FIG. 4 shows a flow chart. In step S401, the scanner index is searched. In step S402, it is checked whether the scanner index is in the positive list. If the scanner index is in the positive list, normal PHI dose information ("PHI") images are continued to be normally uploaded from this scanner index into the central data memory in step S403. If the scanner index is in the positive list, then in step S404, an upload of dose information images for that scanner index is stopped and the dose information image from the study is loaded as an original into the checklist container.S405 denotes a PHI dose information image. S 406 denotes a central data store comprising a data store S 407 and a check list container S 408. In step S409, the scanner index is set in the negative list.S410 denotes a database having a positive list S411 with scanner indices and a negative list S412 with scanner indices. In step S 413, customer information (App+Email) is generated. In step S414, customer information is also generated. S415 denotes an HTML application for the positive list. In step S416, a platform web page is created.Step S417 is started upon request. In step S417, original / phi dose information images are loaded from the checklist container into a file system via a checklist download tool. S418 denotes a job with a check list download operation. S419 and S420 denote original dose information images. S421 denotes a PHI dose information image. S 422 denotes a local file system. S423 and S424 denote a Radiation Dose Structured Reports ("RDSR"). S 425 denotes an original dose information image.In step S426, PHI dose information images are generated (optionally) from the original dose information images using a PHI validation tool. In step S427, defects are validated by a recipient team. In step S428, a PHI validation is performed. In step S429, errors are checked.If there are errors, the defect is sent to the recipient team in step S430. If there are no errors, in step S 431 a structured radiation dose report is generated (optional) from the original dose information images by means of an RDSR validation tool. S432 denotes a job with an RDSR validation process. In step S434, the defect is validated by a DM team.In step S433, errors are checked. If there are errors, the defect is sent to the DM team in step S435. If errors are present, a structured radiation dose report is generated from the PHI dose information images using an RDSR validation tool in step S436, and the results are uploaded. S438 denotes a job with an RDSR validation process.In step S439, the scanner index is set to the negative list. In step S440, errors are checked. If there are no errors, the scanner index is inserted into the positive list and deleted from the negative list in step S441. In step S442, the defect is obtained from and solved by an RCV team. S443 denotes a TFS defect. In step S444, the defect is obtained from and solved by a DM team.The section A1 denotes an evaluation method, the section A2 denotes a preparation method, the section A3 denotes a validation method, and the section A4 denotes a correction method.Scanner index DICOM tags for unique CT scanner dose information image variants may be defined as follows:Manufacturers ManufacturersScanner Index-1 (SI-1)Manufacturer ModelScanner Index-2 (SI-2)Station nameScanner Index-3 (SI-3)Photometric InterpretationScanner Index-4 (SI-4)Device serial numberScanner Index-5 (SI-5)Software versionScanner Index-6 (SI-6)The scanner index is obtained, for example, by:Scanner Index ='@'+'SI-1'+'@'+'SI-2'+'@'+'SI-3'+'@'+'SI-4'+'@'+'SI-5'+'@'+'SI-6'+'@'Example = =@Sie@Definition Flash@CTAW3@Monochrome2@@Syngo VA30A@The scanner index may have the following state values:-"To check"Set by the checklist uploader when a new scanner index for an operator is inserted into the table."In Processing"Set by the checklist downloader when a checklist container has been downloaded to the file system."OK"is set by the dose information image verifier when a validation result was positive."Non OK"is set by the dose information image verifier when a validation result was negative.The following software modules run as independent programs. "Checklist uploaders":It is called by the recipient whenever a new study is received by the image placement and communication system PACS. "Checklist downloader":It is called cyclically by a virtual machine or cloud service to process "to-be-checked" positive list entries. "Dose information image validation":Is called by the checklist downloader during processing of a positive list entry to be checked. "whitelist browserHTML5 app":Mobile application to review current positive list / negative list entry states.The checklist uploader ("Greylist uploader") performs the following steps:U0) The remote operator's playlist uploader "greylist uploader" is started cyclically, for example every hour.U1) Each local study directory in the recipient's query study pool is searched (obtained from the remote recipient).U2) Each dose information image within a local, found study directory is searched (obtained from the remote recipient).U3) The scanner index of this dose information image is obtained (obtained from the remote recipient).U4) The scanner index of a new available dose information image (obtained from the remote recipient) is compared to a positive list in the central data store whether or not it comprises the scanner index from that operator.U5) The full directory of 1..n files is uploaded from a local file system to the central data store to a checklist container "Greylist container" (if the checklist container does not exist, it is created).U6) The status of the new entry is set to the "check" status and the scanner index is included in the positive list.U7) The container in the recipient's query study pool directory is deleted.U8) A predetermined time is maintained and the process is continued again with step U1).The checklist downloader ("Greylist downloader") performs the following steps:D0) The checklist downloader is started cyclically, for example every hour.D1) A scanner index line with the status = "to check" is searched in the positive / negative list.D2) For each entry found, the checklist container reference (name) is obtained.D3) The complete container of 1..n files is downloaded from the central data store to the local file system directory (if the directory does not exist, it is created).D4) Dose information image validation is performed and the results are entered into both the checklist container and the positive list / negative list.D5) The full local folder with 1..n files is uploaded from the local file system to the centre store's checklist container (this always has updated status content in the positive / negative list entries).D6) The status of the new entry is set to the status "OK" or "not OK" depending on the check result in step D4), and this is entered in the positive list.D7) The local folder is deleted if the container is in the query study pool directory.D8) The corresponding scanner index line is updated in the checklist container with the new in-process state.D9) Proceed to step D2) until no values to be checked are found.D10) A predetermined time is maintained and the process is continued again with step D1).Dose Information Image VerifierV0) Begin dose information image validation, for example as a result of activation of a checklist download mockup ("for review" or "non-OK" on demand activation).V1) Search the number of dose information images available in the local directory with study content downloaded from the checklist container ("in-work" or "non-OK"). If original dose information images are available, steps D2-D7 may be applied to the previously converted PHI-converted dose information images using a PHI transformation tool and for all available dose information images, a Radiation Dose Structured Reports ("RDSR") is called via a Clunie tool and the RDSR is stored in the same local folder.V2) The structured radiation dose report generated in step V1) is used and this is converted into a text file via a dcm-dump DdmK tool and stored in the same local folder. Then, the text file "*_dcmtk_dcmdump.txt" generated in step V3) and an RDSR dose verification script are applied, and the result is stored as a text file in the same local folder.V3) The result "*_Dose_Check.txt" is returned to the local folder ("OK" or "non-OK") by the RDSR dose checkThe software mechanism enables a reliable and secure connection of CT scanners of different operators and different manufacturers to a cloud-based dose management system. This solves the problem of unnormalized dose information images being unified by different manufacturers' CT scanners. It can be ensured that dose information images of CT scanners of different manufacturers are either supported or an uploading of dose information images is blocked until problems are solved. In addition, information about supported or unsupported dose information images is provided by CT scanners based on uniform criteria from a plurality of manufacturers.The advantages are a cloud enabled continuous-time mechanism to support association of CT scanners that generate dose information images in an automated and controlled manner. A central cloud enabled positive list of CT scanner indices with dose information images tested is provided. A central cloud-enabled negative list of CT scanner indices with dose information images that are not currently supported is provided. Defining a scanner index as a table entry with a defined set of states ("for review", "in progress", "non-OK", or "OK") within a cloud-enabled positive list or negative list is enabled.An event driven or cyclic runtime state / event engine that checks a local study folder with dose information images not currently supported for scanner indexes in uploads of a positive list / negative list is on a check list container in the cloud and sets a state of a scanner index to "check.".An event driven or cyclic runtime state / event engine that checks a playlist container in the cloud for scanner indices in uploads of a positive list / negative list is provided to a local study folder with dose information images to be validated for support and sets a scanner index to either "ok" or "non-ok".The above object is achieved because the positive list mechanism whitelisting mechanism for dose information images is useful for a large base of different manufacturers' CT scanners to handle reliable and effective operator connectivity in cloud-based systems.Support for non-standardized dose information images for CT scanners of different manufacturers is unified. Dose information images from different manufacturers can either be reliably supported or an uploading of dose information images can be blocked until problems have been solved. Information about supported and unsupported dose information images from different manufacturers may be provided based on uniform criteria.All features explained and shown in connection with individual embodiments of the invention can be provided in different combinations in the subject matter according to the invention in order to realize their advantageous effects at the same time.All method steps may be implemented by devices suitable for carrying out the respective method step. All functions executed by subject features can be a method step of a method.The scope of the present invention is given by the claims and is not limited by the features explained in the specification or shown in the figures.

Claims

Method for automatically controlling loading operations of dose information images (101), comprising the steps of: - receiving (S101) a data set on a local data memory (103) comprising a dose information image (101) and a scanner index, wherein the dose information image comprises a digital image file comprising dose information, wherein the dose information comprises at least a single dose, an average dose, a total dose and / or at least a scan length; - matching (S102) the scanner index with a positive list (105) from the central data memory (103); - transmitting (S103) the data set from the local data memory (103) into a verification directory of the central data memory (107) if the scanner index is not comprised in the positive list (105); validating (S 104) the dose information image (101) from the verification directory by means of a font recognition, comprising: scanning the dose information image with the font recognition to acquire the dose information and verifying the plausibility and correctness of the dose information; and - entering (S 105) the scanner index (S 105) into the positive list (105) upon successful validation of the dose information image (101).The method of claim 1, wherein the data set is transferred from the local data store (103) into a non-checking directory of the central data store (107) if the scanner index is comprised in the positive list (105).The method of claim 2, wherein the positive list (105) is stored in the central data store (107).Method according to Claim 2 or 3, wherein the positive list (105) is transmitted to the local data memory (103).The method of any preceding claim, wherein the verification record is scanned cyclically for new records.The method according to any of the preceding claims, wherein an upload of data sets to a non-checking directory is disabled if the scanner index is not comprised in the positive list (105).Method according to one of the preceding claims, wherein a validation state is assigned to the data record.A system (100) for automatically controlling loads of dose information images (101), comprising: - an uploader (109) for receiving a data set on a local data storage (103) comprising a dose information image and a scanner index, wherein the dose information image comprises a digital image file comprising dose information, wherein the dose information comprises at least a single dose, an average dose, a total dose and / or at least one scan length, for matching the scanner index with a positive list (105) from the central data storage (103) and for transmitting (S103) the data set from the local data storage (103) into a verification directory of the central data storage (107) if the scanner index is not comprised in the positive list (105); a validater (113) for validating the dose information image (101) by means of a font recognition, wherein the validating comprises scanning the dose information image with the font recognition to acquire the dose information and checking the plausibility and correctness of the dose information, and for entering the scanner index into the positive list (105) upon successful validation of the dose information image (101).The system (100) of claim 8, wherein the system (100) is adapted to enter the scanner index into the positive list (105) if the validation of the dose information image (101) was successful.A computer program product loadable into the memory of a computer and comprising code portions for performing the method of any of claims 1 to 7 when the computer program product is executed by the computer.

Images