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 § 102
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 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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claim(s) 1-4, 7-10, 21-24, 27-29 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Saffari et al. (“Fully Automated Breast Density Segmentation and Classification Using Deep Learning,” 23 November 2020, Diagnostics (Basel). 2020 Nov 23;10(11):988; hereinafter "Saffari").
With regards to Claim 1, Saffari discloses a method of determining density values from mammograms for subjects with unaffected breasts prior to diagnosis of breast cancer (Saffari Table I illustrates that the testing dataset {i.e. first mammograms} includes 27 images categorized as BI-RADS I which are normal breasts without any benign finding1), comprising:
identifying, by a computing system, a first mammogram corresponding to a first breast region in a first unaffected breast of a first subject prior to diagnosis of breast cancer in the first subject (as BI-RADS I images of the testing dataset are normal without any benign finding, i.e. prior to diagnosis of breast cancer), the first mammogram having a first region of interest (ROI) corresponding to a first dense area of the first breast region (FIG. 4 of Saffari illustrates the TESTING aspect in which the input {i.e. first mammogram} is applied to the generator network G(X));
applying, by the computing system, the first mammogram to a machine learning (ML) model to generate a first segmentation map identifying the first ROI within the first mammogram (FIG. 4 of Saffari illustrates the TESTING aspect in which the input {i.e. first mammogram} is applied to the generator network G(X) to output a segmented mask), the ML model established using a training dataset comprising a plurality of examples, each of the plurality of examples comprising (i) a respective second mammogram corresponding to a respective second breast region in a respective second unaffected breast of a corresponding second subject prior to diagnosis of breast cancer in the second subject (Saffari Table I illustrates that the training dataset {i.e. second mammograms} includes 108 images categorized as BI-RADS I which are normal breasts without any benign finding); and (ii) a respective second segmentation map identifying a second ROI in the respective second mammogram corresponding to a respective second dense area of the second breast region (during training in a supervised mode, the classifier learns to distinguish between fatty and dense pixels from manually annotated images {i.e. second mammogram of second breast region}; see Saffari pg. 9, ¶ 4 & FIG. 4; FIG. 4 of Saffari illustrates training cGAN where Y are the input segmentation maps from the manually annotated images);
determining, by the computing system, a density value for the first dense area of the first breast region based on the first segmentation map (determining a percent density of the fibroglandular breast tissue from the mammographic images based on the generated mask {i.e. first segmentation mask}; see Saffari pg. 10, §3.3.1); and
storing, by the computing system, using one or more data structures, an association between the first subject and the density value (the act of determining the percent density of the patient indicates that the computer system used had to store said information in memory).
Claim 21 recites similar limitations and are rejected under the same rationale as Claim 1 with the addition of the computer system (see Saffari Abstract).
With regards to Claim 21, Saffari discloses further comprising classifying, by the computing system, the first subject into one of a plurality of risk levels each associated with a predicted likelihood of occurrence of breast cancer based on the density value for the first dense area of the first breast region (thresholding the breast density based on the BI-RADS density standard (0% < BDE < 25%), (26% < BDE < 50%), (51% < BDE < 75%), (76% < BDE < 100%); see Saffari pg. 10, §3.3.1).
Claim 22 recites similar limitations and are rejected under the same rationale as Claim 2 with the addition of the computer system (see Saffari Abstract).
With regards to Claim 32, Saffari discloses further comprising categorizing, by the computing system, the first dense area of the first breast region into one of a plurality of density types based on a ratio of a first portion of the first segmentation map corresponding to the first ROI and a second portion of the first segmentation map outside the first ROI (thresholding the breast density based on the BI-RADS density standard (0% < BDE < 25%), (26% < BDE < 50%), (51% < BDE < 75%), (76% < BDE < 100%); see Saffari pg. 10, §3.3.1; Prevalence, relative risks of developing breast cancer based on Four classes of Breast Imaging and Reporting Data system (BI-RADS) density standard (i.e., fatty, scattered fibroglandular density, heterogeneously dense, and extremely dense; see also Saffari FIG. 1 & corresponding caption; one of ordinary skill in the art understands that the corresponding BI-RADS density standards I-IV are associated with almost entirely fat (negative), scattered fibroglandular densities (benign), heterogeneously dense (Probably Benign), & extremely dense (Suspicious for Malignancy), respectively).
Claim 23 recites similar limitations and are rejected under the same rationale as Claim 3 with the addition of the computer system (see Saffari Abstract).
With regards to Claim 41, Saffari discloses further comprising providing, by the computing system, information for presentation based on the association between the first subject and the density value (FIG. 2 of Saffari illustrates that the breast density estimation is outputted).
Claim 24 recites similar limitations and are rejected under the same rationale as Claim 4 with the addition of the computer system (see Saffari Abstract).
With regards to Claim 76, Saffari discloses wherein the predetermined threshold is based on a plurality of density values from a corresponding plurality of control subjects without breast cancer (thresholding the breast density based on the BI-RADS density standard (0% < BDE < 25%), (26% < BDE < 50%), (51% < BDE < 75%), (76% < BDE < 100%); see Saffari pg. 10, §3.3.1; one of ordinary skill in the art would understand that the BI-RADS density standard is established based on a cohort of samples).
Claim 27 recites similar limitations and are rejected under the same rationale as Claim 7 with the addition of the computer system (see Saffari Abstract).
With regards to Claim 91, Saffari discloses wherein determining the density value further comprises determining the density value based on a ratio of a first portion of the first segmentation map corresponding to the first ROI and a second portion of the first segmentation map outside the ROI (computing the ratio between the area of dense tissues and the area of the breast region to estimate the breast density in the input image; see Saffari pg. 10, §3.3.1).
With regards to Claim 101, Saffari discloses wherein the density value comprises a fibroglandular density value selected from among entirely fatty, scattered fibroglandular density, heterogeneously dense, or extremely dense (thresholding the breast density based on the BI-RADS density standard (0% < BDE < 25%), (26% < BDE < 50%), (51% < BDE < 75%), (76% < BDE < 100%); see Saffari pg. 10, §3.3.1; Prevalence, relative risks of developing breast cancer based on Four classes of Breast Imaging and Reporting Data system (BI-RADS) density standard (i.e., fatty, scattered fibroglandular density, heterogeneously dense, and extremely dense; see also Saffari FIG. 1 & corresponding caption; one of ordinary skill in the art understands that the corresponding BI-RADS density standards I-IV are associated with almost entirely fat (negative), scattered fibroglandular densities (benign), heterogeneously dense (Probably Benign), & extremely dense (Suspicious for Malignancy), respectively).
Claim 29 recites similar limitations and are rejected under the same rationale as Claim 10 with the addition of the computer system (see Saffari Abstract).
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 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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 5-6 & 25-26 are rejected under 35 U.S.C. 103 as being unpatentable over Saffari et al. (“Fully Automated Breast Density Segmentation and Classification Using Deep Learning,” 23 November 2020, Diagnostics (Basel). 2020 Nov 23;10(11):988; hereinafter "Saffari") in further view of Highnam et al. (US PGPUB 20240046472; hereinafter "Highnam") having an effective filing date of 21 April 2021.
With regards to Claim 52, while Saffari discloses all of the limitations of intervening claim 2 as shown above, it appears that Saffari may be silent to wherein the breast cancer is one of HER2-positive breast cancer, estrogen receptor-positive breast cancer, progesterone receptor-positive breast cancer, or triple negative breast cancer.
However, Highnam teaches of a system and method to characterize breast tissue in mammograms based on global density metrics {e.g. volumetric breast density, volume of fibroglandular tissue, breast volume} (see Highnam Abstract & ¶ [0050 & 0119]).
In particular, Highnam teaches of wherein the breast cancer is one of HER2-positive breast cancer, estrogen receptor-positive breast cancer, progesterone receptor-positive breast cancer, or triple negative breast cancer (A breast cancers, higher fibroglandular tissue volumes have shown positive associations with human epidermal growth factor receptor-2 (HER2)-positive, luminal B/HER2-negative and luminal B/HER2-positive subtypes. Conversely, ‘triple negative’ breast cancers are associated with smaller breast volumes and significantly lower non-dense volumes. Triple negative breast cancers account for 10-20% of breast cancers (triple negative breast cancers test negative for oestrogen- and progesterone-receptors, and HER2) and do not respond to targeted therapies, requiring chemotherapy as the primary treatment; see Highnam ¶ [0007]).
Highnam also teaches that the global and localized levels of characterization {i.e. volumetric breast density, volume of fibroglandular tissue, breast volume as established above} help predict… response to treatments or risk-reducing medications (see Highnam ¶ [0029]).
Saffari and Highnam are both considered to be analogous to the claimed invention because they are in the same field of breast density characterization. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Saffari to incorporate the above teachings of Highnam to provide at least the limitations of Claim 5. Doing so would aid in providing predictive information specific to different breast cancer or benign lesion subtypes (see Highnam ¶ [0007]).
Claim 25 recites similar limitations and are rejected under the same rationale as Claim 5 with the addition of the computer system (see Saffari Abstract).
With regards to Claim 61, Saffari teaches of further comprising administering one or more of: a radiation therapy, immunotherapy, chemotherapy or surgery to the first subject, when the density value of the first subject is elevated relative to a predetermined threshold (that the global and localized levels of characterization {i.e. volumetric breast density, volume of fibroglandular tissue, breast volume as established with relation to Claim 5} help predict… response to treatments or risk-reducing medications (see Highnam ¶ [0029]; it should be appreciated that one of ordinary skill in the art would understand that a highly efficacious response to risk-reducing medications {i.e. chemotherapy or immunotherapy} would inform a practitioner to administer said treatment)
Claim 26 recites similar limitations and are rejected under the same rationale as Claim 6 with the addition of the computer system (see Saffari Abstract).
Claims 30-32 are rejected under 35 U.S.C. 103 as being unpatentable over Saffari et al. (“Fully Automated Breast Density Segmentation and Classification Using Deep Learning,” 23 November 2020, Diagnostics (Basel). 2020 Nov 23;10(11):988; hereinafter "Saffari") in further view of Ning et al. (US PGPUB 20170020474; hereinafter "Ning")
With regards to Claim 301, Saffari discloses further comprising:
selecting, by the computing system, based on thresholding the first mammogram, a first area of the first mammogram corresponding to at least a portion of the first breast region excluding a pectoral in the first breast region (region growing segmentation to remove pectoral muscle from the image; see Saffari pg. 6, ¶ 5-7; it should be appreciated that region growing techniques rely on minimum area & similarity threshold),
wherein applying the first mammogram to the ML model further comprises applying the first area of the first mammogram to the ML model to generate the first segmentation map identifying the first ROI within the first area of the first mammogram (the pre-processed are then fed into the trained cGAN network as illustrated in FIG. 2).
While Saffari discloses pre-processing the mammograms to remove pectoral muscle to reduce false positive, it appears that Saffari may be silent to excluding a first outer epidermis layer. However, Ning teaches of a method of breast density measurement (see Ning Abstract) in which a histogram thresholding method is applied to the image to remove the skin from the image (see Ning ¶ [0062-0065]).
Saffari and Ning are both considered to be analogous to the claimed invention because they are in the same field of breast density quantification. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Ning to incorporate the above teachings of Ning to provide at least excluding a first outer epidermis layer. Doing so would aid radiologists to make more efficient and accurate decisions (see Ning ¶ [0029]).
With regards to Claim 3120, wherein the ML model is established by:
identifying, from a training dataset comprising a plurality of examples, an example comprising (i) the second mammogram of the second breast region of the second subject (Saffari Table I illustrates that the training dataset {i.e. second mammograms} includes 108 images categorized as BI-RADS I which are normal breasts without any benign finding); and (ii) the second segmentation map identifying the second ROI in the second mammogram corresponding to the second dense area of the second breast region (during training in a supervised mode, the classifier learns to distinguish between fatty and dense pixels from manually annotated images {i.e. second mammogram of second breast region}; see Saffari pg. 9, ¶ 4 & FIG. 4; FIG. 4 of Saffari illustrates training cGAN where Y are the input segmentation maps {i.e. second segmentation map} from the manually annotated images),
selecting, based on thresholding the second mammogram, a second area of the second mammogram corresponding to at least a portion of the second breast region excluding a second outer epidermis layer in the second breast region (a histogram thresholding method is applied to the image to remove the skin from the image (see Ning ¶ [0062-0065]),
applying the second area of the second mammogram to the ML model to generate a third segmentation map identifying a third ROI corresponding to the second dense area of the second breast region (FIG. 4 of Saffari illustrates a cGAN in which a loss function minimizes a mapping between a ground-truth {i.e. second segmentation map} and a predicted output {i.e. third segmentation map}; see Saffari pg. 7, ¶ 3),
determining a loss metric based on a comparison between the second segmentation map of the example and the third segmentation map generated by the ML model, and updating at least one weight of the ML model based on the loss metric (training of the cGAN is based on the optimization of weighting factor, 𝐺(𝑥,𝑧) and 𝐷(𝑥,𝐺(𝑥,𝑧) as show in EQ. 1; see Saffari pg. 8, ¶ 2-3).
With regards to Claim 3230, further comprising determining, by the computing system, a risk level indicating a predicted likelihood of breast cancer in the first subject based on the density value (the accuracy values depicted in Table 2 along with the corresponding BI-RADS classification can inform one of ordinary skill in the art of the corresponding patient risk.).
Response to Arguments
Applicant's arguments filed 15 January 2026 have been fully considered but they are not persuasive.
In particular, Applicant contends that Saffari does not anticipate the amended claims. In support, Applicant argues that:
“The INbreast dataset referred to in Saffari as the dataset used to train its generator (the supposed "ML model") includes "410 images [(the alleged 'first mammogram' and 'second mammogram')] ... from women with both breasts affected ... and ... from mastectomy patients" in the Abstract. 1 Nowhere does Saffari ever even insinuates at the use of mammograms in connection unaffected breasts, never mind unaffected breasts prior to diagnosis of breast cancer in a subject. Saffari at best mentions classification of breast density into "normal fatty breast" at § 2.1 and a classifier "to distinguish between fatty and dense pixels from manually annotated images" at § 3 .2. Even then, Saffari never considers establishing such a generator using images corresponding to unaffected breasts, much less generating a segmentation map identifying an ROI in a mammogram of such an unaffected breast.”
The Office respectfully disagrees. It should be appreciated that during the interview of 13 January 2026, the Office agreed that the proposed claims my distinguish over Saffari; however, upon further consideration that is not the case. More specifically, Applicant interpretation of Saffari’s use of the INbreast dataset to train and test their cGAN-UNet framework does not take the entirety of the Saffari reference into consideration, let alone the cited training embodiment as illustrated in FIG. 4. Firstly, Applicant relies on the claims being limited to first/second “unaffected breast(s).” However, the instant specification fails to establish a special definition of “unaffected,” thus, one of ordinary skill in the art would rely the plain an ordinary meaning: not influenced or changed mentally, physically, or chemically2 (emphasis added). Moreover, BI-RADS classification category I is define as normal or not even a benign finding3. This definition is further supported by Applicant’s own admission that “unaffected breasts lack apparent anatomical features, such as detectable tumors, that affected breasts have, and are often composed of dense fibroglandular tissue” which is commensurate with the BI-RADS I definition of not even a benign finding. Accordingly, one of ordinary skill in the art would recognize that BI-RADS I classification image would meet the plain an ordinary meaning of an “unaffected breast.”
Returning to the Saffari, Saffari discloses in Table 1 that the training dataset includes 108 images classified as BI-RADS I, i.e. unaffected breasts. Similarly, Table 1 also discloses that the test subset includes 27 images classified as BI-RADS I, i.e. unaffected breasts. While further considering FIG. 7 of Saffari, one of ordinary skill in the art would conclude that Saffari does teach establishing a generator network with unaffected breasts which is used to generate a corresponding segmentation mask {i.e. map} and does teach of identifying a first breast region in a first unaffected breast in a first mammogram {i.e. Saffari’s testing} and applying said first mammogram to a ML model trained based on second breast region in a second unaffected breast {i.e. Saffari’s training}.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ASHISH S. JASANI whose telephone number is (571)272-6402. The examiner can normally be reached M-F 8:00 am - 4:00 pm (CST).
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, Keith M. Raymond can be reached on (571) 270-1790. 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.
/ASHISH S. JASANI/Examiner, Art Unit 3798
/KEITH M RAYMOND/ Supervisory Patent Examiner, Art Unit 3798
1 https://radiopaedia.org/articles/breast-imaging-reporting-and-data-system-bi-rads-assessment-category-1
2 https://www.merriam-webster.com/dictionary/unaffected
3 https://radiopaedia.org/articles/breast-imaging-reporting-and-data-system-bi-rads-assessment-category-1