Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 8/14/2026 has been entered.
Response to Amendment
The Amendment filed August 14 2026 has been entered and considered. Claims 1, 10-14, and 20 have been amended. New claim 22 has been added. In light of the amendment the prior art rejections of claims 1, 10, 11, and 14 are withdrawn as moot. The new grounds of rejection set forth in the present action were necessitated by Applicants’ claim amendments.
Response to Arguments
Applicant’s arguments with respect to claim 1 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1-5, 7-16, and 20-22 are rejected under 35 U.S.C. 103 as being unpatentable over Foley (Previously cited) in view of Trautwein (Previously cited) further in view of Accomazzi et al. (US Patent Pub. No. 2020/0334810 A1, published 2020).
Regarding claim 1, Foley teaches a computer-implemented method (Para. 8, “The example system includes a computer vision processor to identify one or more second features in the image data”) for a post- acquisition check of an X-ray image dataset (Para. 28, “Imaging devices (e.g., X-Ray machine) generate medical images), the computer-implemented method comprising: receiving input data (Fig. 2, 210), the input data including the X-ray image dataset, the X-ray image dataset including an X-ray image and first metadata (Para. 64, “The example system 200 processes patient exam record data (e.g., one or more DICOM image files, etc.) using the input processor 210 to identify image data and metadata and separate the image data and metadata for processing.”), and the first metadata including a acquisition parameters of an examination region in the X-ray image (Para. 33, “For example, the automated QC process can rely on image metadata generated by the imaging device (e.g., image properties, acquisition parameters, etc.)”; Para. 34, “Deep learning model(s) check for features related to image quality (e.g., low level checks, etc.), patient anatomy and features…”); applying a trained function to the input data to generate output data, the output data including second metadata (Para. 65, “The metadata and/or image data can be processed by the AI modeler 220 to correlate image content and non-image information (e.g., features, etc.), verify patient identity, verify region of interest/anatomy of interest, etc.”); and providing the output data, in response to determining the first metadata does not match the second metadata (Para. 37, “In addition to checking the quality of the image, this information can be compared to the metadata to check for inconsistencies, enabling automatic quality control of metadata to reduce processing errors in QC. In certain examples, this QC analysis can be integrated into medical systems to generate notifications, corrections, etc.”).
Foley does not explicitly disclose the first metadata including a first view position of an examination region in the X-ray image or the trained function being trained on a plurality of classes including a corresponding body part and a corresponding view position. However, they do teach that the metadata includes information from the imaging device with non-limiting examples such as image properties and acquisition parameters, as well as training the trained function for classification.
Trautwein teaches the first metadata including a first view position of an examination region in the X-ray image (Para. 50, “The classification metadata may also provide the projection direction in which the image was acquired, and if necessary, which posture the patient has adopted during the image acquisition or in which position (e.g. lying, sitting, or standing) an image was acquired.”).
Trautwein does not explicitly disclose the trained function being trained on a plurality of classes including a corresponding body part and a corresponding view position.
Accomazzi teaches the trained function being trained based on a plurality of classes, and each among the plurality of classes including a corresponding body part and a corresponding view position (Para. 284, “In some embodiments, information corresponding to the orientation, body part and view position is manually entered or tagged by a clinician or technician when the medical image is acquired at a medical imaging device 104.”; Para. 288, “In one embodiment, the training data comprises approximately 500,000 thumbnail X-ray DICOM medical images from one or more PACS (Picture Archiving and Communication System) systems. A list is also generated comprising important body parts and view positions that had to be identified by the CNN. For example, body parts included skull, cervical spine, thoracic/lumbar spine, thorax, abdomen, pelvis and extremity while the view positions included lateral, anterior-posterior and oblique. Sample DICOM images for each kind of aforementioned body parts and view positions are also obtained.”; Para. 298, “Using a first alternative wherein a single CNN classifier that predicts a score for all possible combinations of orientation, body part and view position”, All possible combinations of body part and view position are scored, indicating that the CNN is trained on data that has corresponding body parts and view positions in order to score the combination).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Foley to incorporate the teachings of Trautwein and Accomazzi to include the first metadata including a first view position of an examination region in the X-ray image and the trained function being trained based on a plurality of classes, and each among the plurality of classes including a corresponding body part and a corresponding view position. Foley teaches a system of automatic quality control of medical images, as well as using image metadata derived from the imaging device and from features in the image. Trautwein teaches that the metadata may provide the projection direction in which the image was acquired and the posture of the patient. One of ordinary skill in the art would have understood that the view position can affect evaluative criteria such as anatomical appearance, feature location, and overlap. Modifying the system of Foley to include the view position in the metadata as taught by Trautwein would have predictably improved the accuracy and reliability of quality control determinations. Accomazzi further discloses using a trained function to score all possible combinations of body parts and view positions. One of ordinary skill in the art would have recognized that considering both the body part and view position together would predictably improve the speed of processing without sacrificing performance, as disclosed by Accomazzi (Para. 298).
Regarding claim 2, Foley as modified teaches all of the elements of claim 1, as stated above, as well as wherein the first metadata and the second metadata each include information regarding a respective body part of the examination region; and the second metadata includes a second view position of the examination region (Para. 33, “The metadata can include important information that can be configured according to one or more criterion, such as patient details, scanner details, scanning parameters, and contextual/clinical information (contrast, anatomy scanned, etc.). For example, the automated QC process can rely on image metadata generated by the imaging device (e.g., image properties, acquisition parameters, etc.)”; Para. 34, “Image information involves low-level features such as presence of noise, contrast enhancement, artifact, etc.; medium-level features such as presence of anatomy region”; Trautwein; Para. 50, “The classification metadata may also provide the projection direction in which the image was acquired, and if necessary, which posture the patient has adopted during the image acquisition or in which position (e.g. lying, sitting, or standing) an image was acquired.”).
Regarding claim 3, Foley as modified teaches all of the elements of claim 1, as stated above, as well as wherein the X-ray image dataset is a DICOM image dataset (Para. 33, “For example, metadata is analyzed from one or more image files (e.g., DICOM files, etc.)”).
Regarding claim 4, Foley as modified teaches all of the elements of claim 1, as stated above, as well as wherein the second metadata is automatically corrected (Para. 67, “The quality controller 250 can determine a quality result based on the automated analysis of the incoming patient data by the AI modeler 220, computer vision processor 230, and results evaluator 240.”; Para. 94, “At block 1150, when the patient data does not satisfy the quality criterion(-ia), the patient data is rejected. In certain examples, at block 1160, corrective action may be applied. When corrective action is to be applied, at block 1170, corrective action can be taken to adjust the image data and/or metadata, and the patient data can be re-evaluated to determine compliance”).
Regarding claim 5, Foley as modified teaches all of the elements of claim 1, as stated above, as well as wherein the providing includes displaying a suggestion to a user to correct the first metadata with the second metadata for confirming or declining the suggestion (Para. 95, “Incorrect and/or uncorrelated content can be flagged for user review and/or triggered for adjustment/corrective action (e.g., metadata can be adjusted to correctly identify image content, metadata can be deidentified, new image data can be acquired and associated with the metadata, etc.).”).
Regarding claim 7, Foley as modified teaches all of the elements of claim 1, as stated above, as well as wherein the second metadata includes information regarding at least one of a body part or a second view position of the examination region (Para. 34, “Deep learning model(s) check for features related to image quality (e.g., low level checks, etc.), patient anatomy and features (e.g., medium level checks), and checks about the patient such as obesity (e.g., high level checks), etc. Image information involves low-level features such as presence of noise, contrast enhancement, artifact, etc.; medium-level features such as presence of anatomy region, organ, landmark, etc.; and high-level features such as presence of obesity, extreme pathology, and implant detection, etc.”; Para. 33, “For example, the automated QC process can rely on image metadata generated by the imaging device (e.g., image properties, acquisition parameters, etc.)”; Trautwein; Para. 50, “The classification metadata may also provide the projection direction in which the image was acquired, and if necessary, which posture the patient has adopted during the image acquisition or in which position (e.g. lying, sitting, or standing) an image was acquired.”).
Regarding claim 8, Foley as modified teaches all of the elements of claim 1, as stated above, as well as wherein the second metadata is available when the X-ray image is reviewed (Para. 102, “At block 1220, image information is extracted from the DICOM file. For example, deep learning model(s) can be used to instantiate algorithms to check for features including image quality (low level checks), patient anatomy (medium level checks), and checks about the patient such as obesity (high level checks).”; Para. 104, “At block 1240, the analyzed image information is compared with the metadata to preserve the integrity of a large dataset.”).
Regarding claim 9, Foley as modified teaches all of the elements of claim 1, as stated above, as well as wherein the trained function is based on a convolutional neural network (Para. 71, “FIG. 4 illustrates a particular implementation of the example neural network 300 as a convolutional neural network 400.”).
Regarding claim 10, Foley as modified teaches a computer-implemented method for providing a trained function (Fig. 7), the computer-implemented method comprising: receiving input training data, the input training data including an X-ray image dataset, the X-ray image dataset including an X-ray image and first metadata, and the first metadata including a first view position of an examination region in the X-ray image (Para. 88, “Thus, health data provided to the AI modeler 220 from the input processor 210, which identified the image data and non-image data (e.g., meta data, other meta information, etc.)…”; Para. 33, “For example, the automated QC process can rely on image metadata generated by the imaging device (e.g., image properties, acquisition parameters, etc.)”; Trautwein; Para. 50, “The classification metadata may also provide the projection direction in which the image was acquired, and if necessary, which posture the patient has adopted during the image acquisition or in which position (e.g. lying, sitting, or standing) an image was acquired.”); receiving output training data, the output training data being related to the input training data, and the output training data including second metadata (Fig. 9B), training a function based on the input training data, the output training data and a plurality of classes to obtain the trained function (Para. 86, “In the example of FIG. 9B, a plurality of training inputs 911 are provided to a network 921 to develop connections in the network 921 and provide an output to be evaluated by an output evaluator 931. Feedback is then provided by the output evaluator 931 into the network 921 to further develop (e.g., train) the network 921.”), each among the plurality of classes including a corresponding body part and a corresponding view position (See analysis of claim 1 above); and providing the trained function (Para. 87, “FIG. 9C depicts an example deployed device 903. Once the training device 901 has learned to a requisite level, the training device 901 can be deployed for use”).
Claim 11 corresponds to claim 1 and is rejected under the same analysis.
Claim 12 corresponds to claim 1 and is rejected under the same analysis.
Claim 13 corresponds to claim 10 and is rejected under the same analysis.
Claim 14 corresponds to claim 10 and is rejected under the same analysis.
Claim 15 corresponds to claim 1 and is rejected under the same analysis.
Claim 16 corresponds to claim 5 and is rejected under the same analysis.
Claim 20 corresponds to claim 1 and is rejected under the same analysis.
Claim 21 corresponds to the recited matching elements of claim 1 and body part/view position elements of claim 2. It is rejected under the same analysis.
Regarding claim 22, Foley as modified teaches all of the elements of claim 1, as stated above, as well as wherein the trained function is trained to output a vector of tuples, each of the tuples representing a different class among the plurality of classes (Accomazzi; Para. 298, “Using a first alternative wherein a single CNN classifier that predicts a score for all possible combinations of orientation, body part and view position”, Although Accomazzi describes generating a score for each possible combination rather than explicitly disclosing the combination as tuples, one of ordinary skill in the art would have recognized that each combination defines a respective class and could predictably be represented as a tuple comprising the attributes defining the class without changing the underlying classification operation.)
Claim(s) 6 and 17-19 are rejected under 35 U.S.C. 103 as being unpatentable over Foley as modified in view of Trautwein and Accomazzi and further in view of DCMTK (NPL, “dcmodify: Modify DICOM files”, published 2010).
Regarding claim 6, Foley as modified in view of Trautwein and Accomazzi teaches correcting the first metadata using the second metadata; and modifying a private DICOM tag in response to the correcting (Para. 37, “In addition to checking the quality of the image, this information can be compared to the metadata to check for inconsistencies, enabling automatic quality control of metadata to reduce processing errors in QC. In certain examples, this QC analysis can be integrated into medical systems to generate notifications, corrections, etc.”; Para. 36, “Removal of identifying information and/or other deidentification involves detection and modification of metadata (e.g., erase from DICOM header, etc.)”; Para. 95, “ Incorrect and/or uncorrelated content can be flagged for user review and/or triggered for adjustment/corrective action”; Para. 106, “Patient-identifying meta information can be detected and modified (e.g., erase from the DICOM header, replaced with dummy information such as 0000 or xxxx, etc.), for example.”).
Foley as modified in view of Trautwein and Accomazzi does not explicitly teach adding a private DICOM tag in case the first metadata is corrected using the second metadata. However, they do modify a DICOM header to erase, replace, and otherwise alter information as well as flag content when it may need corrective action. Accomazzi also discloses using private DICOM tags.
DCMTK teaches wherein a private DICOM tag is added (Pg. 6, “If you wish to insert a private tag (not a reservation with gggg,00xx), be sure, that you've listed it in your dictionary”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Foley, Trautwein, and Accomazzi to incorporate the teachings of DCMTK to include wherein a private DICOM tag is added in case the first metadata is corrected using the second metadata. Foley discloses a method for reviewing metadata and performing correction processing if it does not pass the quality control criteria. They also disclose the ability to modify the DICOM header in order to erase, replace, or otherwise alter information. DCMTK discloses an ability to insert a private tag into a DICOM file. One of ordinary skill in the art would understand that modifying the DICOM information of Foley to include a private DICOM tag when the metadata is corrected after review would be a routine implementation of the known DICOM tag insertion techniques taught by DCMTK. Foley further discloses flagging content when it is incorrect in order to trigger corrective action. Applying a private DICOM tag once this flagged content is corrected would be a straightforward method of improving organization of the stored data after correction, consequently providing more robust information which would be desirable for version control/auditing and could even be used as feedback for the training of the automated system.
Claim 17 corresponds to claim 6 and is rejected under the same analysis.
Claim 18 corresponds to claim 8 and is rejected under the same analysis.
Claim 19 corresponds to claim 9 and is rejected under the same analysis.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to DAVID A WAMBST whose telephone number is (703)756-1750. The examiner can normally be reached M-F 9-6:30 EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Gregory Morse can be reached at (571)272-3838. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/DAVID ALEXANDER WAMBST/Examiner, Art Unit 2663
/GREGORY A MORSE/Supervisory Patent Examiner, Art Unit 2698
Prosecution Insights