DETAILED ACTION
Claims 1-22 are currently pending.
The previous rejections to claims 1-22 under 35 U.S.C. 101 are maintained.
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 2/17/26 has been entered.
Response to Arguments
Applicant's arguments filed 2/17/26 relating to the rejections to claims 1-22 under 35 U.S.C. 101 have been fully considered but they are not persuasive.
The Applicant argues on page 8 of the response in essence that: The features of claim 1 cannot be practically performed in the human mind. In particular, claim 1 is directed to "training a segmentation algorithm for digital pathology." The human mind is not equipped to train a segmentation algorithm.
The involvement of a generic computer components does not provide additional elements that are sufficient to amount to significantly more than the judicial exception because the recitations to hardware involve no more than a generic computer performing generic computer functions that are well understood, routine and conventional activities previously known in the industry. That is, other than reciting “by a processor,” nothing in the claim precludes the steps from practically being performed in the human mind. See MPEP 2106.05(d)).
The Applicant argues on page 8 of the response in essence that: Further, even assuming that claim 1 did recite an abstract idea as alleged, claim 1 is surely a practical application of any alleged abstract idea. The Office Action admits that the segmentation algorithm is an additional element. Office Action, page 3. And, claim 1 is a method of training the segmentation algorithm. Thus, claim 1 requires the segmentation algorithm and the inclusion of the segmentation algorithm transforms and/or integrates any alleged abstract idea into a practical application.
The recitation of the segmentation algorithm is simply generally linking the use of a judicial exception to a particular technological environment or field of use. See MPEP § 2106.05(g). “[P]atents that do no more than claim the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models to be applied, are patent ineligible under § 101.” Recentive Analytics, Inc. v. Fox Corp., No. 2023-2437 (Fed. Cir. Apr. 18, 2025).
The Applicant argues on page 9 of the response in essence that: The segmentation algorithm is not merely linking claim 1 to a particular technological environment but is instead an integral component of claim 1 that transforms and/or integrates any alleged abstract idea of claim 1 into a practical application.
Segmenting an image is a step that does not preclude the limitation from being performed in the human mind. A human is capable of segmenting an image by using pen and paper. That the segmentation is performed by an algorithm is no more than linking the limitation to a generic computer performing generic computer function.
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.
Claims 1-22 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claims do not fall within at least one of the four categories of patent eligible subject matter because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more. The flow chart in MPEP 2106, Subject Matter Eligibility Test For Products and Processes, will be referenced to establish that the subject matter is ineligible.
Step 1: claim 1 recites a method, claim 19 recites a computer-readable medium and claim 20 recites a device. Thus, claims 1, 19 and 20 fall under one of the four recognized statutory categories.
Step 2A Prong One: However, claims 1, 19 and 20 are further directed to the abstract idea of examining an image of an annotated tissue sample to label regions of the tissue sample. See MPEP 2106.04(a)(2). Furthermore, the claims do not preclude the limitations from being performed in the human mind. The limitations are mental processes that can be performed by a human using pen and paper.
Step 2A Prong Two: Additional elements include a segmentation algorithm and a classification algorithm. The involvement of a generic computer components does not provide additional elements that are sufficient to amount to significantly more than the judicial exception because the recitations to hardware involve no more than a generic computer performing generic computer functions that are well understood, routine and conventional activities previously known in the industry. That is, other than reciting “by a processor,” nothing in the claim precludes the steps from practically being performed in the human mind. See MPEP 2106.05(d)). The recitation of a segmentation algorithm, a classification algorithm, class activation maps and pooling of values do not distinguish the claimed invention from a generic computer performing conventional machine learning.
Step 2B: The claims do not provide an inventive concept as they do not provide an improvement to any type of particular machine. The Applicant's Background describes inspection of tissue samples as predating the Applicant' s claimed invention. Merely automating the process of inspection of tissue samples does not constitute a patentable improvement in computer technology. Furthermore, “patents that do no more than claim the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models to be applied, are patent ineligible under § 101.” Recentive Analytics, Inc. v. Fox Corp., No. 2023-2437 (Fed. Cir. Apr. 18, 2025).
Claim Rejections - 35 USC § 102
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.
Claims 1-3, 5-7, 13, 15, 16, 19, 20 and 22 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Dehaene et al. Non-Patent Literature Self-Supervision Closes the Gap Between Weak and Strong Supervision in Histology (hereafter “Dehaene”).
Referring to claims 1 and 20, Dehaene discloses a computer-implemented method of training a segmentation algorithm for digital pathology, the computer-implemented method comprising:
obtaining a whole-slide image depicting multiple types of tissue (page 1, One example of a routine task performed by pathologists is metastasis detection in hematoxylin and eosin (H&E) stained slides of lymph nodes, which can prove challenging as whole-slide images (WSI) can be over 25 mm wide while micro-metastases can be as small as 50µm wide);
obtaining a first annotation of at least a part of the whole-slide image, the first annotation having a first level of detail and including annotations for the multiple types of tissue depicted in the whole-slide image (page 1, However, in most existing cohorts, WSIs are only associated with a pathology report containing global labels, but that lacks detailed pixel-level annotations);
determining, based on the first annotation, a second annotation of the whole-slide image, the second annotation including annotations for the multiple types of tissue depicted in the whole-slide image and having a second level of detail which is higher than the first level of detail (page 1, Models that attempt to learn directly from global labels belong to the field of weakly-supervised learning, and have to propagate the slide-level information down to individual pixels of the WSI); and
setting parameters of the segmentation algorithm based on the second annotation, the segmentation algorithm configured to segment a whole-slide image depicting tissue of multiple types, in accordance with the multiple types of tissue based on the second level of detail (page 5, a patch-based classifier is trained using local tile annotations and tile predictions are combined in a post-processing stage to establish a strongly-supervised state of the art at 99.3% AUC on the test set).
Referring to claim 2, Dehaene discloses performing, based on the first annotation, a training of a classification algorithm configured to classify image regions of the whole-slide image in accordance with the multiple types; and
upon said performing of the training of the classification algorithm, using the classification algorithm to determine the second annotation (page 1, Models that attempt to learn directly from global labels belong to the field of weakly-supervised learning, and have to propagate the slide-level information down to individual pixels of the WSI).
Referring to claim 3, Dehaene discloses determining class activation maps of the classification algorithm for the image regions of the whole-slide image, wherein the second annotation is determined based on the class activation maps (page 6, Figure 2: (a) Tumor annotations (orange) displayed on a Camelyon16 test slide).
Referring to claim 5, Dehaene discloses extracting, at the second level of detail, features from the whole-slide image, and
wherein the second annotation is determined based on the features and the class activation maps (page 1, Models that attempt to learn directly from global labels belong to the field of weakly-supervised learning, and have to propagate the slide-level information down to individual pixels of the WSI).
Referring to claim 6, Dehaene discloses determining, based on the features extracted from the whole-slide image, a partitioning of the whole-slide image,
wherein the second annotation is determined based on a combination of the partitioning and the class activation maps (page 5, a patch-based classifier is trained using local tile annotations and tile predictions are combined in a post-processing stage to establish a strongly-supervised state of the art at 99.3% AUC on the test set).
Referring to claim 7, Dehaene discloses wherein the features include at least one of a color or a contrast gradient (page 4, The Camelyon16 dataset contains a total of 400 sentinel lymph node H&E-stained WSIs [analyzing stained slides requires color analysis]).
Referring to claim 13, Dehaene discloses wherein the first annotation is obtained from a manual annotation process (page 1, However, in most existing cohorts, WSIs are only associated with a pathology report containing global labels, but that lacks detailed pixel-level annotations).
Referring to claim 15, Dehaene discloses obtaining a further whole-slide image (page 1, One example of a routine task performed by pathologists is metastasis detection in hematoxylin and eosin (H&E) stained slides of lymph nodes, which can prove challenging as whole-slide images (WSI) can be over 25 mm wide while micro-metastases can be as small as 50µm wide);
determining a segmentation result using the segmentation algorithm for the further whole-slide image (page 2, Most approaches thus rely on a pre-processing step to reduce the dimensionality of the input, often separating the slide into smaller images (tiles) on which conventional architectures, such as convolutional neural network, can operate); and
detecting, based on the segmentation result, tumor-type tissue in the further whole-slide image (page 5, a patch-based classifier is trained using local tile annotations and tile predictions are combined in a post-processing stage to establish a strongly-supervised state of the art at 99.3% AUC on the test set).
Referring to claim 16, Dehaene discloses extracting, at the second level of detail, features from the whole-slide image (page 1, Models that attempt to learn directly from global labels belong to the field of weakly-supervised learning, and have to propagate the slide-level information down to individual pixels of the WSI),
wherein the second annotation is determined based on the features and the class activation map (page 6, Figure 2: (a) Tumor annotations (orange) displayed on a Camelyon16 test slide).
Referring to claim 19, Dehaene discloses a non-transitory computer-readable storage medium storing computer-executable instructions that, when executed by at least one processor, cause the at least one processor to perform the computer-implemented method of claim 1 (page 1, Models that attempt to learn directly from global labels belong to the field of weakly-supervised learning, and have to propagate the slide-level information down to individual pixels of the WSI [the model must be run on a computer with a computer-readable medium]).
Referring to claim 22, Dehaene discloses wherein a resolution of labels of the second annotation is higher than a resolution of labels of the first annotation (page 1, Models that attempt to learn directly from global labels belong to the field of weakly-supervised learning, and have to propagate the slide-level information down to individual pixels of the WSI).
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.
Claims 4, 9 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Dehaene et al. Non-Patent Literature Self-Supervision Closes the Gap Between Weak and Strong Supervision in Histology as applied to claims 1 and 6 above, and further in view of Nazeri et al. Non-Patent Literature Paper Two-Stage Convolutional Neural Network for Breast Cancer Histology Image Classification (hereafter” Nazeri”).
Referring to claim 4, Dehaene discloses the class activation maps, but does not disclose expressly wherein the class activation maps have a third level of detail, which is smaller than the second level of detail.
Nazeri discloses wherein the class activation maps have a third level of detail, which is smaller than the second level of detail (page 5, For the proposed image-wise network we follow a similar pattern. Series of 3 x 3 convolutional layers are followed by a 2 x 2 convolution with stride of 2 for downsampling. Each layer is followed by batch normalization and ReLU activation function. We use the same 1 x 1 convolutional layer as before to obtain the spatial average of activation maps before the classifier).
Before the effective filing date of the claimed invention, it would have obvious to a person of ordinary skill in the art for the class activation maps to have a smaller level of detail than the second annotation. The motivation for doing so would have been to use a network architecture that efficiently prevents overfitting and not introduce any dropout or weight decay. Therefore, it would have been obvious to combine Nazeri with Dehaene to obtain the invention as specified in claim 4.
Referring to claims 9 and 17, Dehaene determining the second annotation, but does not disclose expressly wherein the second annotation is determined based on a pooling of values of the class activation maps for each type across partitions of the partitioning.
Nazeri discloses wherein the second annotation is determined based on a pooling of values of the class activation maps for each type across partitions of the partitioning (page 7, The image label is obtained using one of the three different patch probability fusion methods. These methods include, majority voting, where the image label is selected as the most common patch label, maximum probability, where the patch with higher class probability decides the image label, and sum of probabilities, where the patch class probabilities are summed up and the class with the largest value is assigned).
Before the effective filing date of the claimed invention, it would have obvious to a person of ordinary skill in the art to determine the second annotation by based on a pooling of values of the class activation maps. The motivation for doing so would have been to increase the accuracy of the model to detect tumors. Therefore, it would have been obvious to combine Nazeri with Dehaene to obtain the invention as specified in claims 9 and 17.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Dehaene et al. Non-Patent Literature Self-Supervision Closes the Gap Between Weak and Strong Supervision in Histology as applied to claim 6 above, and further in view of Bhatt et al. US Publication 2022/0122347 (hereafter “Bhatt”).
Referring to claim 8, Dehaene discloses partitioning the whole-slide image, but does not disclose expressly wherein the partitioning includes super-pixels of the whole-slide image.
Bhatt discloses wherein the partitioning includes super-pixels of the whole-slide image (paragraph 30, In accordance with an embodiment of the present disclosure, the neural network self-trains in order to segment the image into certain number of clusters. After making the inference at every iteration, a normalized cut method is applied on the input image to obtain a superpixel output).
Before the effective filing date of the claimed invention, it would have obvious to a person of ordinary skill in the art to partition an image into superpixels. The motivation for doing so would have been to increase the effectiveness of the image classification by reducing the number of iterations required. Therefore, it would have been obvious to combine Bhatt with Dehaene to obtain the invention as specified in claim 8.
Claims 10-12 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Dehaene et al. Non-Patent Literature Self-Supervision Closes the Gap Between Weak and Strong Supervision in Histology as applied to claims 1-3 above, and further in view of Wang et al. Non-Patent Literature Weakly Supervised Deep Learning for Whole Slide Lung Cancer Image Analysis (hereafter “Wang”).
Referring to claim 10, Dehaene discloses obtaining the first annotation, but does not disclose expressly wherein the first annotation is restricted to a fraction of the whole-slide image.
Wang discloses wherein the first annotation is restricted to a fraction of the whole-slide image (page 3951, Patch-level labels or even pixelwise annotation masks can be feasibly provided by pathologists for designing the effective algorithms at the training phase),
wherein the image regions of the whole-slide image for which the class activation maps are determined are at least partially outside of the fraction of the whole-slide image (page 3956, We implement different feature aggregation methods that are commonly used in the previous works to obtain the image-level prediction of each WSI).
Before the effective filing date of the claimed invention, it would have obvious to a person of ordinary skill in the art to obtain patch level annotations for the whole-slide image. The motivation for doing so would have been to increase the accuracy and effectiveness of the image classification. Therefore, it would have been obvious to combine Wang with Dehaene to obtain the invention as specified in claim 10.
Referring to claim 11, Dehaene discloses obtaining the first annotation, but does not disclose expressly not including tissue fractions of other types of the multiple types larger than a length threshold.
Wang discloses wherein any given image region of the whole-slide image having a label of the first annotation that is associated with a first type of the multiple types does not include tissue fractions of other types of the multiple types larger than a size threshold, and
wherein a length scale of the image regions corresponds to the size threshold (page 3951, Patch-level labels or even pixelwise annotation masks can be feasibly provided by pathologists for designing the effective algorithms at the training phase).
Before of the effective filing date of the claimed invention, it would have obvious to a person of ordinary skill in the art to obtain patch level annotations for the whole-slide image. The motivation for doing so would have been to increase the accuracy and effectiveness of the image classification. Therefore, it would have been obvious to combine Wang with Dehaene to obtain the invention as specified in claim 11.
Referring to claim 12, Dehaene discloses obtaining the first annotation, but does not disclose expressly wherein the first annotation is restricted to a fraction of the whole-slide image.
Wang discloses wherein the first annotation is restricted to a fraction of the whole-slide image (page 3951, Patch-level labels or even pixelwise annotation masks can be feasibly provided by pathologists for designing the effective algorithms at the training phase), and
wherein the second annotation covers the whole-slide image (page 3956, We implement different feature aggregation methods that are commonly used in the previous works to obtain the image-level prediction of each WSI).
Before of the effective filing date of the claimed invention, it would have obvious to a person of ordinary skill in the art to obtain patch level annotations for the whole-slide image. The motivation for doing so would have been to increase the accuracy and effectiveness of the image classification. Therefore, it would have been obvious to combine Wang with Dehaene to obtain the invention as specified in claim 12.
Referring to claim 14, Dehaene discloses obtaining the first annotation, but does not disclose expressly not including tissue fractions of other types of the multiple types larger than a length threshold.
Wang discloses wherein any given image region of the whole-slide image having a label of the first annotation that is associated with a first type of the multiple types does not include tissue fractions of other types of the multiple types beyond at least one of a quota or larger than a size threshold (page 3951, Patch-level labels or even pixelwise annotation masks can be feasibly provided by pathologists for designing the effective algorithms at the training phase).
Before of the effective filing date of the claimed invention, it would have obvious to a person of ordinary skill in the art to obtain patch level annotations for the whole-slide image. The motivation for doing so would have been to increase the accuracy and effectiveness of the image classification. Therefore, it would have been obvious to combine Wang with Dehaene to obtain the invention as specified in claim 14.
Claim 18 is rejected under 35 U.S.C. 103 as being unpatentable over Dehaene et al. Non-Patent Literature Self-Supervision Closes the Gap Between Weak and Strong Supervision in Histology and Bhatt et al. US Publication 2022/0122347 as applied to claim 8 above, and further in view of Nazeri et al. Non-Patent Literature Paper Two-Stage Convolutional Neural Network for Breast Cancer Histology Image Classification (hereafter” Nazeri”).
Referring to claim 18, Dehaene determining the second annotation, but does not disclose expressly wherein the second annotation is determined based on a pooling of values of the class activation maps for each type across partitions of the partitioning.
Nazeri discloses wherein the second annotation is determined based on a pooling of values of the class activation maps for each type across partitions of the partitioning (page 7, The image label is obtained using one of the three different patch probability fusion methods. These methods include, majority voting, where the image label is selected as the most common patch label, maximum probability, where the patch with higher class probability decides the image label, and sum of probabilities, where the patch class probabilities are summed up and the class with the largest value is assigned).
Before the effective filing date of the claimed invention, it would have obvious to a person of ordinary skill in the art to determine the second annotation by based on a pooling of values of the class activation maps. The motivation for doing so would have been to increase the accuracy of the model to detect tumors. Therefore, it would have been obvious to combine Nazeri with Dehaene to obtain the invention as specified in claim 18.
Claim 21 is rejected under 35 U.S.C. 103 as being unpatentable over Dehaene et al. Non-Patent Literature Self-Supervision Closes the Gap Between Weak and Strong Supervision in Histology as applied to claim 1 above, and further in view of Yip et al. US Publication 2021/0166785 (hereafter “Yip”).
Referring to claim 21, Dehaene discloses the second level of detail, but does not disclose expressly wherein the second level of detail has at least one of a size threshold or a quota of minority type instances that is smaller than for the first level of detail.
Yip discloses wherein the second level of detail has at least one of a size threshold (paragraph 185, From the determinations made by the modules 304, 306, and 307, the post-processing controller 308 can determine whether the image data contains an amount of tissue that exceeds a threshold and/or satisfies a criterion, for example enough tissue for genetic analysis, enough tissue to use the image data as training images during a learning phase of the deep learning framework, or enough tissue to combine the image data with an already existing trained classifier model).
Before the effective filing date of the claimed invention, it would have obvious to a person of ordinary skill in the art to have a size threshold for a level of detail. The motivation for doing so would have been to increase the accuracy and effectiveness of the image classification by ensuring sufficient data in the sample. Therefore, it would have been obvious to combine Yip with Nazeri to obtain the invention as specified in claim 21.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to PETER K HUNTSINGER whose telephone number is (571)272-7435. The examiner can normally be reached Monday - Friday 8:30 - 5:00.
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, Benny Q Tieu can be reached at 571-272-7490. 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.
/PETER K HUNTSINGER/ Primary Examiner, Art Unit 2682
Prosecution Insights