DETAILED ACTION
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 .
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claim(s) 1-11 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The limitations, under their broadest reasonable interpretation, cover mental processes (concepts performed in a human mind, including as an observation, evaluation, judgment, opinion, and/or mathematical concepts and calculations). Under step 2A, prong one, independent claims 1, 11, and 12 recite acquiring anonymized image information and clinical information, selecting target image information, selecting target clinical information, associating disease/organ information with the selected information, and formatting/standardizing the selected information. These limitations are directed to an abstract idea of mental processes and insignificant extra solutions of collecting, selecting, organizing, labeling/associating, and formatting/standardizing information. Under step 2A, prong two, the judicial exception is not integrated into a practical application. The claims disclose a computer, medical AI/drug discovery AI, processor, and CRM, which are recited with a high level of generality and are directed to generic computer components, and the medical AI/drug discovery AI are merely a field of use. The claims do not recite how the computer improves image processing or training data generation. Under step 2B, the additional elements, individually and as an ordered combination, amount only to generic computer implementation of routine data processing functions. These elements do not provide an inventive concept beyond the abstract idea.
The additional limitations of claims 2-9 merely specify data labeling, selection criteria, image association/alignment, blockchain recording, model generation, transmission of data, and fee allocation. These limitations remain directed to classifying, formatting, storing, transmitting, or monetizing information using generic computer functions. They do not recite a particular AI architecture, specific image processing algorithm, improved training technique, nonconventional data structure, or improvement to computer, medical imaging, blockchain, or AI technology.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-11 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
The term “selecting target image information suitable as data for medical AI/drug discovery AI” in claims 1, 10, and 11 is a relative subjective term which renders the claim indefinite. The term “suitable” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention.
The term “one or more contained lesions in each of the multiple image information have a complication” in claim 4 is a relative subjective term which renders the claim indefinite. The term “complication” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention.
Claims 2-19 are rejected by virtue of dependency.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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) 1-3, 5-6, 8, and 10-11 are rejected under 35 U.S.C. 103 as being unpatentable over Sommer (US 20230326598) in view of Poblenz (US 20200160963).
Regarding claim 1:
Sommer discloses: An information processing method causing a computer to execute operations to thereby generate data for medical AI/drug discovery AI (¶ [0017] – [0023] disclose generating medical training data for training AI algorithm including automatically linking annotations from findings reports and routine clinical data to medical image data, and using generated image + annotation datasets for AI training), the operations comprising:
acquiring multiple image information, in which each of the multiple image information is generated by a medical device having one or more imaging functions (¶¶ [0033] – [0037] teach acquiring medical image data generated by medical imaging modalities, including x-ray, CT, SPECT, PET, MRI, fMRI, ultrasound, OCT, MR-PET, PET-CT, pathology images, and ophthalmology images. Additionally, the image datasets can include 2D/3D images, image series, temporal series, 3D/4D data, video, or image elements such as organs, tissue, pathological areas, ROIs, and VOIs);
acquiring multiple clinical information related to medical practice for a patient (¶¶ [0040] – [0044] teach acquiring clinical information related to a patient, including laboratory findings, pathological findings, medical letter, personal data, other patient data or other elements of an electronic health records (HER) from PACS, RIS, and/or HIS.)
Sommer does not expressly teach: private information contained in such image information has been subjected to an anonymization process; private information associated with the multiple clinical information has been anonymized.
However, in the same field of endeavor, Poblenz teaches private information contained in such image information has been subjected to an anonymization process; and private information associated with the multiple clinical information has been anonymized. (¶ [0171] teaches de-identification system used to de-identify image data, medical report data, private fields of medical scan entries including patient identifier data. ¶¶ [0172] and [0210] further teaches receiving a medical scan and corresponding report, identifying patient identifier in the scan header, performing header anonymization, and generating a de-identified medical scan by replacing patient identifying header fields with anonymized fields. ¶¶ [0215] – [0218] also teach de-identifying patient identifiers in the image data itself using image analysis and image fiducials. Also see ¶ [0180]).
selecting target image information suitable as data for medical AI/drug discovery AI from among the multiple image information (Sommer ¶¶ [0021] – [0023] teach generating AI training data from image data and annotation obtained from routine clinical findings reports, and that the generated datasets are suitable for training diagnostic AI algorithm. ¶¶ [0054] – [0058] further teach that the contents of selected findings report elements are linked to image datasets and that these linked data serve as training data, and that the entire image dataset, individual images, slices, or image elements may be linked to report element contents, and that this linking is annotation of the image data. Poblenz also teach selection of suitable medical scans for training (see ¶¶ [0101] – [00102]));
selecting target clinical information to be used as the data for medical AI/drug discovery AI from among the multiple clinical information (Sommer ¶¶ [0042] – [0044] teach structure findings reports constructed from findings report elements, and these findings reports include semantic features describing diagnosis or patient health status. ¶ [0060] also teaches selecting findings report contents for linking with image data, image elements, segmentation data and/or classification data., and that the contents of finding reports or findings report elements considered for linking can be determined in advance or during the creation of the report, and that the selection may be based on guidelines, textbooks, recommendations, user experience, study guidelines, or requests from those responsible for AI training. Also see ¶¶ [0065] – [0056], [0063], [0070] – [0072]. Poblenz also teaches using clinical/training features such as patient history data, diagnosis data, medical code data, annotation data, scan type, and other image associated data with features selected based on administrator instructions or to reduce classification error (see ¶ [0102)) ;
associating disease information related to a disease and organ information related to an organ with the target image information and the target clinical information (Sommer ¶¶ [0004] – [0008] teach that annotations include classification and segmentation information. Classification includes categories such as tumor vs. cyst, determination of a risk score, etc. Segmentation information recognizes organs, tissue types, blood vessels, cells and areas affected by a specific pathology. ¶¶ [0037] – [ 0038] further teach that image elements include organs or other coherent tissue, pathological areas such as cysts, tumors or altered tissue areas, (ROIs or VOIs). Also see ¶¶ [0061], [0063], and [0070] – [0071]) ; and
performing standardization on the target image information and the target clinical information (Sommer ¶¶ [0043] teaches standardization of clinical/report information. It specifically teaches that a structured report is machine-readable, has a fixed structure and contains standardized elements, wording and layout. In addition, pre-generated report templates can be used. These may provide case-specific structure and include recommended reporting steps. ¶¶ [0047] – [0049] further teach automatically linking report elements to unique identifiers, where each identifier corresponds to a specific medical semantic content, the identifiers may be controlled terminology, lexicon, standard, or a medical ontology like RadLex, SNOMED-CT, LOINC or DICOM, and also teach a synoptic report is a medical findings report in which the individual findings elements are assigned to concrete database entries. Synoptic findings are completely machine-readable and the individual findings elements can be uniquely assigned a meaning Electronic synoptic findings reports use templates with coded values to record interoperable data in discrete fields. Synoptic findings provide semantic features. ¶¶ [0054] – [0056] and [0070] – [0071] further teach linking image/image elements to unique identifier so report contents can be lined to image datasets, images, and/or image elements via identifiers. Poblenz also teaches standardization of image and clinical/annotation information by teaching that medical scan entries may be formatted in DICOM (see ¶ [0042] and ¶ [0102])).
Regarding claim 2:
Sommer further teaches: wherein associating the disease information related to a disease and the organ information related to an organ with the target image information includes specifying one or more locations of and/or classifying one or more lesions and/or one or more organ structures contained in the target image information (Sommer ¶¶ [0006], [003] and [0061] disclose specifying locations of lesions, locations of organ structures, and classifying lesions. ¶¶ [0063], [0070] – [0071] teach linking the segmentation/classification data to the image data and finding report/clinical information).
Regarding claim 3:
Poblenz further teaches: wherein selecting the target image information from among the multiple image information is performed (Poblenz ¶ [0029] teaches a medical scan annotator system that selects a medical scan for transmission to client devices);
based on an operation on a terminal apparatus that is used by a user (¶ [0029] also teaches that the selected medical scan is transmitted to a first and second client device and displayed via an interactive interface, and that annotation data is received from the client devices in response. ¶¶ [0023] – [0025] teach that client devices are associated with users, including hospitals, medical institutions, medical professionals, employees, admins, radiologists, and other users, and that a medical scan is presented for review and that scan review data is generated based on user input to the interactive interface.)
Regarding claim 5:
Sommer further teaches: selecting the target image information from among the multiple image information includes selecting, from among the multiple image information, first target image information generated by a first medical device and second target image information generated by a second medical device (Sommer ¶¶ [0033], [0035], [0054] – [0056] teach medical image data maybe acquired using different medical imaging modalities (from different medical devices), image datasets may be composed of images from these different modalities, and that the entire image dataset, individual images of the image dataset, slices, or individual image elements can be linked to finding reports contents, and that such linking is annotating image data); and
performing standardization on the target image information includes aligning the first target image information and the second target image information (Sommer ¶¶ [0054], [0055], [0101], and [0103] – [0106] teach aligning the first and second target image information by linking images/image elements to unique identifiers, storing related image/annotation/semantic feature data in connected databases with unique links, and combining image data with segmentation/classification information and semantic features into multi-dimensional AI training datasets. The instant specification defines aligning and states that “aligning the first target image information and the second target image information” may refer to any of associating the first target image information and the second target image information with each other, combining the first target image information and the second target image information with each other, and comparing the target image information and the second target image information” in ¶ [0037]).
Regarding claim 6:
Sommer further teaches: selecting the target image information from among the multiple image information includes selecting, from among the multiple image information, first target image information generated based on photographing by a pre-determined medical device at a first point in time (Sommer ¶ [0034] teaches medical image datasets that may form a temporal series, including 3D/4D data. This supports medical images captured at a first time point as part of a time series imaging dataset);
and second target image information generated based on photographing by the pre-determined medical device at a second point in time different from the first point in time (Sommer ¶ [0034] expressly gives examples of temporal image data showing changes over time, including temporal accumulation of a contrast agent, growth of a tumor, and temporal neuronal signal courses. These examples require image/data from different times) and
performing standardization on the target image information includes aligning the first target image information and the second target image information (Sommer ¶¶ [0054], [0055], [0101], and [0103] – [0106] teach aligning the first and second target image information by linking images/image elements to unique identifiers, storing related image/annotation/semantic feature data in connected databases with unique links, and combining image data with segmentation/classification information and semantic features into multi-dimensional AI training datasets. The instant specification defines aligning as ““aligning the first target image information and the second target image information” may refer to any of associating the first target image information and the second target image information with each other, combining the first target image information and the second target image information with each other, and comparing the target image information and the second target image information” in ¶ [0037]).
Regarding claim 8:
further causing the computer to execute generating a training model which has been trained on the data for medical AI/drug discovery AI (Poblenz ¶¶ [0097], [0101] teach generating NN or other learning models trained on medical scan classification data, metadata, patient data, and related scan data. Sommer ¶¶ [0021] – [0023] also teach that the generated medical image + annotation datasets are suitable for training diagnostic AI algorithm).
Regarding claims 10-11: the claims limitations are similar to those of claim 1; therefore rejected in the same manner as applied above. Sommer discloses an apparatus in FIGS. 3 and 4, and a CRM in ¶ [0010].
Claim(s) 7 and 9 are rejected under 35 U.S.C. 103 as being unpatentable over Sommer (US 20230326598) in view of Poblenz (US 20200160963) and Diedrich (US 20220270146).
Regarding claim 7:
Sommer in view of Poblenz does not specifically teach: further causing the computer to execute recording information related to transmission and receipt of the data for medical AI/drug discovery AI in Blockchain.
In a related field, Diedrich teaches: further causing the computer to execute recording information related to transmission and receipt of the data for medical AI/drug discovery AI in Blockchain (¶¶ [0017] - [0019], [0042] – [0045], and [0061] teach recording in a blockchain ledger, references to medical image/annotation training data, transactions, ownership, access history, and use of the data for model training testing).
Therefore, it would have been obvious to a person of ordinary skill in the art prior to the effective filing date of the claimed invention to have modified Sommer in view of Poblenz to incorporate the teachings of Diedrich by including blockchain recording because the references address the same recognized problem of making medical image/annotation data available for AI models training while preserving data control, traceability, privacy, and accountability, This combination would predictably improve traceability and security of shared medical training datasets, especially where the data are provided to external AI developers or model trainers.
Regarding claim 9:
Diedrich further teaches: transmitting the data for medical AI/drug discovery AI to a user (¶¶ [0039] – [0043], [0055], [0060], [0066] – [0068] teach an online marketplace where images, annotations, and models are made available for user, including model trainers. Images and annotations are uploaded to cloud storage/marketplace, user can view images/annotations before purchasing, and model trainers may purchase images, annotations, or both for model training); and
determining at least part of a usage fee for the data for medical AI/drug discovery AI which has been received from the user (¶¶ [0042] – [0044] teach that the marketplace facilitates buying/selling access to images, annotations, and expertise, and that the blockchain ledger manages purchases, sales, and financial transactions between marketplace users. ¶¶ [0021], [0057], [0060], [0066] – [0068] further teach that data quality determines marketplace prices, higher quality image/annotation data receive higher marker price, and model trainers can purchase image/annotation data),
as a data provision fee to be paid to a provider of the target image information (¶¶ [0020], [0043] – [0045] teach that the marketplace users include image owners/providers, annotators, model trainers, service/data providers, and clients, and that the blockchain/smart contracts tie sets of image and annotations to image owners, annotators, or model trainers. ¶¶ [0067] –[0068] teach payment back to contributors: annotators may be paid by usage of annotations sold in the marketplace, providing a rate back to the annotator based on the quality of the annotations, and image/model/annotation prices are based on data quality).
Claim(s) 4 is rejected under 35 U.S.C. 103 as being unpatentable over Sommer (US 20230326598) in view of Poblenz (US 20200160963) and Iguchi (US 20230012527).
Regarding claim 4:
Sommer in view of Poblenz does not specifically teach: wherein selecting the target image information from among the multiple image information includes determining whether or not one or more contained lesions in each of the multiple image information have a complication
In a related field, Iguchi teaches: determining whether or not one or more contained lesions in each of the multiple image information have a complication (¶¶ [0007] – [0009) disclose acquiring a medical image before treatment, inputting the image into a trained model, and outputting complication information on a complication that is likely to occur after the treatment. ¶¶ [0038] – [0044] teach that the model outputs probability values for types of complications. Specifically ¶¶ [0041] – [0043] tie the complications to lesions by teaching that when a ballon ins inflated in a lesion where unevenly distributed calcified tissue is present, the blood vessel may rupture, and then stent is expanded at a plaque portion, vessel wall dissociation or plaque rupture may occur, and then treating unstable plaque, thrombus/peripheral embolism that may occur. ¶¶ [0031], [0047], [0048], [0054] – [0057], [0075] – [0076] also teach that the trained model detects a dangerous region in the medical image where the complication is likely to occur, and outputs a second medical image indicating the dangerous region).
Therefore, it would have been obvious to a person of ordinary skill in the art prior to the effective filing date of the claimed invention to have modified Sommer in view of Poblenz to incorporate the teachings of Iguchi by including complication determination process in order to improve evaluating medical images containing lesions/plaque/calcified tissue to output complication information. This combination predictably improve the quality of the selected AI training dataset by identifying images with complication relevant lesions and distinguishing them from images without complications.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Prakash (WO 2022269626) teaches generating annotated dataset for machine learning implementation for segmenting lesions in medical images.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to WASSIM MAHROUKA whose telephone number is (571)272-2945. The examiner can normally be reached Monday-Thursday 8:00-5:00 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, Stephen Koziol can be reached at (408) 918-7630. 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.
/WASSIM MAHROUKA/Primary Examiner, Art Unit 2665
Prosecution Insights