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 .
Claims 1-22 are pending regarding this application.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 02/19/2025 and 06/26/2026 are considered and attached.
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.
Claims 1, 2, 4-7, 10, and 14-22 are rejected under 35 U.S.C. 103 as being unpatentable over Ocampo et al. (WO 2022/0241300 A1), hereinafter Ocampo in view of Ce-ougna et al. (U.S. Publication No. 2023/0056839 A1), hereinafter Ce-ougna.
Regarding claim 1, Ocampo teaches a method for predicting a gene alteration state in a sample, the method comprising:
receiving, at one or more processors, a plurality of image patches derived from at least one slide image of the sample (Ocampo teaches “a patch-generating module 211 can define a set of image patches for each digital pathology image. To define the set of image patches, the patch-generating module 211 can segment the digital pathology image into the set of image patches” in para. [0067], wherein the digital pathology image can be captured using an “image scanner 224 [which] can then capture an image of the sample slide” as shown in para. [0063]), wherein each image patch of the plurality is associated with a tissue phenotype classification label (Ocampo teaches that “the one or more features may comprise, e.g., one or more of a histologic feature, such as a cell type or cell grouping, a clinical feature, or a genomic feature. Accordingly, generating the label for each of the plurality of image patches may be based on the one or more features.” as shown in para. [0087]. These histologic features are interpreted as equivalent to the claimed tissue phenotype classification label);
identifying, using the one or more processors, a set of tumor region image patches from the plurality of image patches based on the tissue phenotype classification label for each image patch of the plurality of image patches (Ocampo teaches generating a label for each image patch defining whether the patch is a tumor region or not based on the tissue morphology (tissue phenotype classification label) as shown in para. [0087] and FIG. 3);
identifying, using the one or more processors, a set of buffer region image patches from the plurality of image patches (Ocampo teaches "generating, for each of the plurality of image patches, a label indicating a likelihood (e.g., a binary output or a percentage output) that the image patch depicts a cluster of tumor cells” as shown in para. [0024]. Here, the non-tumor labelled image patches (patches wherein the likelihood that the image patch depicts a cluster of tumor cells is low or non-existent) are interpreted as equivalent to the claimed set of buffer region image patches)
inputting, using the one or more processors, the set of tumor region image patches (Ocampo teaches “the digital pathology image processing system 210 may identify and predict pan-tumor or tumor-agnostic actionable gene fusion based on the digital pathology model” as shown in para. [0100]) to obtain a prediction of a presence of at least one gene alteration state for the tissue sample (Ocampo teaches “At step 340 in FIG. 3, the digital pathology image processing system 210 may determine, based on the labels generated for each image patch, that the digital pathology image comprises a depiction of an occurrence of gene fusion with respect to the cancer cells m the image” in para. [0088]. Here, since the labels for each image patch are used to determine whether the image (sample) depicts gene fusion (gene alteration state), it is interpreted that the tumor region image patches, and image patches which are labelled to not have tumors (buffer region image patches) are input into the model. Additionally, Ocampo teaches “aside from positively identifying the gene fusion, the digital pathology model may identify [or predict] more common mutations such as KRAS and EGFR and in doing so, rule out the presence of a gene fusion” in para. [0096]. See also “the remote computing system may then determine, based on the labels generated for each image patch, that the digital pathology image comprises a depiction of an occurrence of an actionable mutation with respect to the cancer cells” as shown in para. [0083]); and
outputting, using the one or more processors, the prediction of a presence of the at least one gene alteration state for the sample (Ocampo teaches outputting a gene alteration state for a sample in para. [0088] and [0096]-[0100]).
Ocampo fails to teach identifying, using the one or more processors, a set of buffer region image patches from the plurality of image patches based on the identified set of tumor region patches and a set of proximity criteria used to define buffer region image patches; and inputting, using the one or more processors, the set of tumor region image patches and the set of buffer region image patches into a classification model.
However, Ce-ougna teaches identifying, using the one or more processors, a set of buffer region image patches from the plurality of image patches based on the identified set of tumor region patches and a set of proximity criteria used to define buffer region image patches (Ce-ougna teaches “the pipeline first looks at only the tumor patches (FIG. 10). Then, the pipeline considers these patches as one object, and pipeline fills its holes to find the non-tumor patches among them. The pipeline also moves it by one pixel in each of the eight directions possible to encompass the area around our object FIG. 11” as shown in para. [0142], wherein “the pipeline could alternatively increase the number of pixels considered for the near region area” and “this yields the non-tumor patches inside and around the tumor area (FIG. 12)”. As such, the non-tumor patches derived from moving the pixel “in each of the eight directions possible to encompass the area around our object” as defined in para. [0142] are interpreted as equivalent to the claimed buffer region patches wherein the pixel defined surrounding area is interpreted as equivalent to the claimed set of proximity criteria. See also para. [0145]. See that Ce-ougna additionally teaches utilizing numerical features of both the tumor region image patches and the set of buffer region image patches as variables for the prediction model in para. [0139]-[0145]. Ce-ougna additionally teaches that “the deep learning model may take as input the genomic data, the clinical data and the maps/masks (e.g., directly)” as shown in para. [0181], wherein the maps include both the non-tumor and tumor patches).
Ocampo and Ce-ougna are both considered to be analogous to the claimed invention because they are in the same field of determining tumor and non-tumor image patches from pathology slides of sample tissue. 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 the teachings of Ocampo to incorporate the teachings of Ce-ougna and include “identifying, using the one or more processors, a set of buffer region image patches from the plurality of image patches based on the identified set of tumor region patches and a set of proximity criteria used to define buffer region image patches”. The motivation for doing so would have been to combine “visual data with genomic data and clinical data” in order to “improve accuracy of the prognosis” and “output patient-wise information relative to patient's future with respect to the cancer disease of the patient”, as suggested by Ce-ougna in para. [0048] and para. [0055], respectively. Therefore, it would have been obvious to one of ordinary skill at the time the invention was filed to combine Ocampo with Ce-ougna to obtain the invention specified in claim 1.
Regarding claim 2, Ocampo and Ce-ougna teach the method of claim 1, further comprising generating the plurality of image patches by
inputting, using the one or more processors, the plurality of image patches into a tissue phenotype classification model, wherein the tissue phenotype classification model is configured to classify each image patch and output a tissue phenotype classification label for the image patch (Ocampo teaches “the digital pathology image processing system may generate, for each of the plurality of image patches, a label indicating a likelihood (e.g., a binary output or a percentage output) that the image patch depicts a cluster of tumor cells” in para. [0008] and [0024]. Here, “generating the label for each of the plurality of image patches may be based on the use of one or more trained machine-learning models” as shown in para. [0087]. This trained model is interpreted as equivalent to the claimed tissue phenotype classification model. The classification here is whether or not the image patch contains a cluster of tumor cells).
Regarding claim 4, Ocampo and Ce-ougna teach the method of claim 1,
wherein the set of proximity criteria comprises
an adjacency criterion,
a distance criterion (Ce-ougna teaches “the pipeline also moves it by one pixel in each of the eight directions possible to encompass the area around our object FIG. 11). As mentioned above, the pipeline could alternatively increase the number of pixels considered for the near region area. The pipeline then subtracts the image of FIG. 10 to the image of FIG. 11, and this yields the non-tumor patches inside and around the tumor area (FIG. 12)” as shown in para. [0142]. Here, since the pipeline considers patches within a specific number of pixels surrounding the tumor patches, this number of pixels in each of the eight directions possible is interpreted as equivalent to the claimed distance criterion. It should be noted that this may also be interpreted as equivalent to the claimed adjacency criterion), or
any combination thereof.
Similar motivations as applied to claim 1 can be applied here to claim 4.
Regarding claim 5, Ocampo and Ce-ougna teach the method of claim 1, wherein identifying the set of buffer region image patches based on the identified set of tumor region patches and the set of proximity criteria comprises
identifying a set of n nearest neighboring, non-tumor region image patches adjacent to a tumor region indicated by the identified set of tumor region patches (Ce-ougna teaches that “the pipeline first looks at only the tumor patches (FIG. 10). Then, the pipeline considers these patches as one object, and pipeline fills its holes to find the non-tumor patches among them. The pipeline [] moves it by one [or more] pixel[s] in each of the eight directions possible to encompass the area around our object FIG. 11). […] The pipeline then subtracts the image of FIG. 10 to the image of FIG. 11, and this yields the non-tumor patches inside and around the tumor area (FIG. 12)” as shown in para. [0142]. Here, Ce-ougna teaches identifying non-tumor patches in an area adjacent to the tumor area. Here, the total number of non-tumor patches (total tissue patches) may be equivalent to the claimed set with n non-tumor patches as defined above).
Similar motivations as applied to claim 1 can be applied here to claim 5.
Regarding claim 6, Ocampo and Ce-ougna teach the method of claim 1, wherein identifying the set of buffer region image patches based on the identified set of tumor region patches and the set of proximity criteria comprises
identifying a set of n neighboring, non-tumor region image patches that are within a specified number of image pixels, m, of an edge of a tumor region indicated by the identified set of tumor region patches (Ce-ougna teaches that “the pipeline first looks at only the tumor patches (FIG. 10). Then, the pipeline considers these patches as one object, and pipeline fills its holes to find the non-tumor patches among them. The pipeline [] moves it by one [or more] pixel[s] in each of the eight directions possible to encompass the area around our object FIG. 11). […] The pipeline then subtracts the image of FIG. 10 to the image of FIG. 11, and this yields the non-tumor patches inside and around the tumor area (FIG. 12)” as shown in para. [0142]. Here, Ce-ougna teaches identifying non-tumor patches around the tumor area which are within a specific number of pixels from the edge of the tumor patches (object). Here, the total number of non-tumor patches (total tissue patches) may be equivalent to the claimed set with n non-tumor patches as defined above).
Similar motivations as applied to claim 1 can be applied here to claim 6.
Regarding claim 7, Ocampo and Ce-ougna teach the method of claim 1, wherein identifying the set of buffer region image patches based on the identified set of tumor region patches and the set of proximity criteria comprises
identifying a set of n neighboring, non-tumor region image patches that are within a predefined distance, d, of an edge of a tumor region indicated by the identified set of tumor region patches (Ce-ougna teaches that “the pipeline first looks at only the tumor patches (FIG. 10). Then, the pipeline considers these patches as one object, and pipeline fills its holes to find the non-tumor patches among them. The pipeline [] moves it by one [or more] pixel[s] in each of the eight directions possible to encompass the area around our object FIG. 11). […] The pipeline then subtracts the image of FIG. 10 to the image of FIG. 11, and this yields the non-tumor patches inside and around the tumor area (FIG. 12)” as shown in para. [0142]. Here, Ce-ougna teaches identifying a set of non-tumor patches around the tumor area which are within a specific number of pixels from the edge of the tumor patches (object). Here, since the process involves determining non-tumor patches within a specific number of pixels from the tumor patches (object), this specified number of pixels is interpreted as equivalent to the claimed predefined distance, as finding patches within a certain amount of pixels can be interpreted as broadly equivalent to the claimed predefined distance, d. Additionally, the total number of non-tumor patches (total tissue patches) may be equivalent to the claimed set with n non-tumor patches as defined above).
Similar motivations as applied to claim 1 can be applied here to claim 7.
Regarding claim 10, Ocampo and Ce-ougna teach the method of claim 1, wherein the tissue phenotype classification label corresponds to one of a plurality of tissue phenotypes comprising
a tumor tissue phenotype (Ocampo teaches “the digital pathology image processing system may generate, for each of the plurality of image patches, a label indicating a likelihood (e.g., a binary output or a percentage output) that the image patch depicts a cluster of tumor cells” in para. [0008] and [0024]. Here the tumor classification is interpreted as equivalent to the claimed tumor tissue phenotype),
a normal tissue phenotype,
a necrotic tissue phenotype,
a stromal tissue phenotype,
an immune tissue phenotype, or
any combination thereof.
Regarding claim 14, Ocampo and Ce-ougna teach the method of claim 1,
wherein the prediction of at least one gene alteration state for the sample comprises a prediction of a presence of at least 1 (Ocampo teaches “at step 340 in FIG. 3, the digital pathology image processing system 210 may determine, based on the labels generated for each image patch, that the digital pathology image comprises a depiction of an occurrence of gene fusion with respect to the cancer cells m the image” in para. [0088]. This depiction is interpreted as equivalent to the prediction of a presence of at least 1 gene alteration),
2,
3,
4,
5,
6,
7,
8,
9, or
10 gene alterations in the sample.
Regarding claim 15, Ocampo and Ce-ougna teach the method of claim 1,
wherein the prediction of at least one gene alteration state comprises a prediction of a presence of at least 1 (Ocampo teaches “aside from positively identifying the gene fusion, the digital pathology model may identify more common mutations such as KRAS and EGFR and in doing so, rule out the presence of a gene fusion” in para. [0096], wherein “FIG. 10B is a schematic illustration of an example representation of mutational context 1030— an oncogene driver mutation (e.g., an EGFR mutation)— and its corresponding visual signature 1040” as shown in para. [0121]),
2,
3,
4,
5,
6,
7,
8,
9, or
10 driver mutations in the sample.
Regarding claim 16, Ocampo and Ce-ougna teach the method of claim 14, wherein the prediction of at least one gene alteration state for the sample comprises a prediction of a presence of an
EGFR alteration (Ocampo teaches “aside from positively identifying the gene fusion, the digital pathology model may identify more common mutations such as KRAS and EGFR and in doing so, rule out the presence of a gene fusion” in para. [0096]),
ALK alteration,
ROS1 alteration,
NTRK1 alteration,
NTRK2 alteration,
NTRK3 alteration,
KRAS alteration,
G12C alteration, or
any combination thereof in the sample.
Regarding claim 17, Ocampo and Ce-ougna teach the method of claim 2, further comprising generating the image patches of the plurality by processing the at least one slide image of the sample (Ocampo teaches that the “digital pathology image processing system 210 can process digital pathology- images, including WSIs [whole slide images]” as shown in para. [0067], which are interpreted as equivalent to the claimed slide image of the sample) using an image segmentation algorithm (Ocampo teaches “to define the set of image patches, the patch-generating modul =e 211 can segment the digital pathology image into the set of image patches” as shown in para. [0067]. See further para. [0067] through [0068] which defines a set of rules to be followed in calculating the segmented image patches. This set of rules for generating the image patches is broadly interpreted as equivalent to the claimed image segmentation algorithm).
Regarding claim 18, Ocampo and Ce-ougna teach the method of claim 1,
wherein each image patch of the plurality of image patches is of a uniform, predetermined shape and size (Ocampo teaches “each pathology slide image may he cropped into image patches with width and height of certain number of pixels”, wherein “an image patch size [] can be determined […] by selecting an image patch size [] associated with one or more performance metrics above a predetermined threshold and/or associated with one or more performance metric(s)” as shown in para. [0067]).
Regarding claim 19, Ocampo and Ce-ougna teach the method of claim 1,
wherein the at least one slide image of the sample comprises a pathology slide image of the sample (Ocampo teaches “the digital pathology image processing system 210 can process WSIs of tissue samples” in para. [0067]. Here, WSIs stands for “whole slide images”).
Regarding claim 20, Ocampo and Ce-ougna teach the method of claim 1,
wherein the sample comprises a tissue sample (Ocampo teaches “the digital pathology image processing system 210 can process WSIs of tissue samples” in para. [0067]).
Regarding claim 21, Ocampo and Ce-ougna teach a system comprising:
one or more processors (Ocampo teaches “one or more processors” in para. [0023]); and
a memory communicatively coupled to the one or more processors and configured to store instructions that, when executed by the one or more processors, cause the system to perform the method of claim 1 (Ocampo teaches “one or more processors; and a non- transitory memory including instructions that, when executed by the one or more data processors, cause the one or more data processors to perform part or all of one or more methods disclosed herein” as shown in para. [0023]).
Regarding claim 22, Ocampo and Ce-ougna teach a non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by one or more processors of a system, cause the system to perform the method of claim 1 (Ocampo teaches “one or more processors; and a non- transitory memory including instructions that, when executed by the one or more data processors, cause the one or more data processors to perform part or all of one or more methods disclosed herein” as shown in para. [0023]).
Claim 3 is rejected under 35 U.S.C. 103 as being unpatentable over Ocampo et al. (WO 2022/0241300 A1), hereinafter Ocampo in view of Ce-ougna et al. (U.S. Publication No. 2023/0056839 A1), hereinafter Ce-ougna in view of Prabhudesai et al (U.S. Publication No. 2020/0074146 A1), hereinafter Prabhudesai.
Regarding claim 3, Ocampo and Ce-ougna teach the method of claim 1.
While Ocampo teaches that “the digital pathology image processing system 210 may use a binary classifier to classify image patches and then determine slide-level prediction by combining (e.g., averaging) all image patch predictions” as shown in para. [0108], Ocampo and Ce-ougna fail to specifically teach applying a binary mask.
However, Prabhudesai teaches wherein identifying the set of tumor region image patches comprises applying a binary mask based on tissue phenotype classification label to the plurality of image patches (Prabhudesai teaches “the process for tumor classification by tumor classification module 122 starts with using a tumor classification model for the type of tissue identified by tissue classification module 120 to create a binary mask to identify ROI” in para. [0020]. While Prabhudesai does not specifically teach applying the binary masks to image patches, Ocampo’s teaching of applying binary classifiers to the image patches can be combined with Prabhudesai’s teaching of applying binary masks to digitized tissue slides based on a tissue phenotype classification as shown in para. [0020], [0028], [0035] and FIGS. 3 and 4).
Ocampo, Ce-ougna, and Prabhudesai are all considered to be analogous to the claimed invention because they are in the same field of determining tumor and non-tumor image areas from pathology slides of sample tissue. 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 the teachings of Ocampo (as modified by Ce-ougna) to incorporate the teachings of Prabhudesai and include “wherein identifying the set of tumor region image patches comprises applying a binary mask based on tissue phenotype classification label to the plurality of image patches”. The motivation for doing so would have been to “create[] a binary mask that isolates the tumor cells shown in the digitized tissue slide by filtering out the non-tumorous cells. The result is an image which includes only the regions of continuous tumor cells, which are ROI”, as suggested by Prabhudesai in para. [0035]. Therefore, it would have been obvious to one of ordinary skill at the time the invention was filed to combine Ocampo and Ce-ougna with Prabhudesai to obtain the invention specified in claim 3.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Ocampo et al. (WO 2022/241300 A1), hereinafter Ocampo in view of Ce-ougna et al. (U.S. Publication No. 2023/0056839 A1), hereinafter Ce-ougna and Bredno et al. (U.S. Publication No. 2019/0038310 A1), hereinafter Bredno.
Regarding claim 8, Ocampo and Ce-ougna teach the method of claim 2.
Ocampo and Ce-ougna fail to teach wherein the tissue phenotype classification model comprises a supervised machine learning model.
However, Bredno teaches wherein the tissue phenotype classification model comprises a supervised machine learning model (Bredno teaches “a soft classification module 114 may be executed to classify each patch using a “soft” classification, such as SVM, and generating a confidence score and a label for each patch” in para. [0056] wherein “the plurality of tissue types may comprise any of normal tissue, tumor, necrosis, stroma, and lymphocyte aggregates” as shown in para. [0032]).
Ocampo, Ce-ougna, and Bredno are all considered to be analogous to the claimed invention because they are in the same field of determining tumor and non-tumor image patches from pathology slides of sample tissue. 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 the teachings of Ocampo (as modified by Ce-ougna) to incorporate the teachings of Bredno and include “wherein the tissue phenotype classification model comprises a supervised machine learning model”. The motivation for doing so would have been that “for each low-confidence patch, neighboring high-confidence patches make larger contributions towards refining the labels for each patch, which improves the segmentation accuracy in the low-confidence patches” and that “treating each test image independently while adaptively improv[es] the classification accuracy based on the labeling confidence information for the image under analysis”, as suggested by Bredno in para. [0057] and para. [0005], respectively. Therefore, it would have been obvious to one of ordinary skill at the time the invention was filed to combine Ocampo and Ce-ougna with Bredno to obtain the invention specified in claim 8.
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Ocampo et al. (WO 2022/0241300 A1), hereinafter Ocampo in view of Ce-ougna et al. (U.S. Publication No. 2023/0056839 A1), hereinafter Ce-ougna and Cheng et al. (CN 114332037 A, see attached English translation for citations), hereinafter Cheng.
Regarding claim 9, Ocampo and Ce-ougna teach the method of claim 2,
wherein the tissue phenotype classification model is trained on a plurality of annotated image patches (training image patches) derived from a plurality of slide images of samples (Ocampo teaches “Generating the label for each of the plurality of image patches may be based on the use of one or more trained machine-learning models. In particular embodiments, the digital pathology image processing system 210 may tram the one or more machine-learning models based on a plurality of training data comprising one or more labeled depictions of a sample comprising, e.g., a tumor region or tumor nest structure, and one or more labeled depictions of a sample that does not include a tumor region or tumor nest structure” as shown in para. [0087])
Ocampo and Ce-ougna fail to teach training the tissue phenotype classification model using a plurality of slide images of samples from a cohort of subjects diagnosed with a disease of interest.
However, Cheng teaches wherein the tissue phenotype classification model is trained on a plurality of annotated image patches (training image patches) derived from a plurality of slide images of samples from a cohort of subjects diagnosed with a disease of interest (Cheng teaches “dividing the classification data set into a training data set and a testing data set, wherein the training data set comprises 8 tissue labeling type image blocks” as shown in para. [0024], and “extracting the tissues marked in the digital pathological image marking library in the form of image blocks, wherein the size of each image block is set to be 224 multiplied by 224 pixels and is used for training an 8-class tissue classifier” as shown in para. [0067], wherein “pathological sections of 87 pancreatic cancer patients are collected and organized” to be used in creating the digital pathological image marking library as shown in para. [0061]-[0065] . Here, the trained tissue classification model, which contains the 8-class tissue classifier, is interpreted as equivalent to the claimed tissue phenotype classification model. Furthermore, since the training data set is derived from annotated image patches of images from a digital library of image data of pancreatic cancer patients, it is interpreted that the training images are derived from samples from a cohort (87 pancreatic cancer patients) of subjects diagnosed with a disease of interest (pancreatic cancer)).
Ocampo, Ce-ougna, and Cheng are all considered to be analogous to the claimed invention because they are in the same field of determining tumor and non-tumor image patches from pathology slides of sample tissue. 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 the teachings of Ocampo (as modified by Ce-ougna) to incorporate the teachings of Cheng and include “training the tissue phenotype classification model using a plurality of slide images of samples from a cohort of subjects diagnosed with a disease of interest”. The motivation for doing so would have been “automatically segmenting multiple classes of tissues in a pathological image of a pancreas of the present invention”, as suggested by Cheng in para. [0039]. Therefore, it would have been obvious to one of ordinary skill at the time the invention was filed to combine Ocampo and Ce-ougna with Cheng to obtain the invention specified in claim 9.
Claim 11 is rejected under 35 U.S.C. 103 as being unpatentable over Ocampo et al. (WO 2022/0241300 A1), hereinafter Ocampo in view of Ce-ougna et al. (U.S. Publication No. 2023/0056839 A1), hereinafter Ce-ougna and Zhang et al. (CN 111180013 A, see attached English translation for citations), hereinafter Zhang.
Regarding claim 11, Ocampo and Ce-ougna teach the method of claim 1.
Ocampo and Ce-ougna fail to teach wherein the gene alteration state classification model comprises a supervised machine learning model.
However, Zhang teaches wherein the gene alteration state classification model comprises a supervised machine learning model (Zhang teaches “the fusion gene confidence prediction module comprises: the characteristic selection submodule is used for setting characteristics; and the random forest quantitative model submodule is used for training a random forest quantitative model by using the characteristics set in the characteristic selection submodule, predicting the reliability of the fusion gene by using the trained random forest quantitative model” as shown in para. [0016]. Here the fusion gene confidence prediction module “identif[ies] multiple fusion variation types of multiple genes” as shown in para. [0009]. Additionally, the trained random forest quantitative model is interpreted as equivalent to the claimed supervised machine learning model, as all random forest models are supervised machine learning models).
Ocampo, Ce-ougna, and Zhang are all considered to be analogous to the claimed invention because they are in the same field of determining tumor and non-tumor image patches from pathology slides of sample tissue. 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 the teachings of Ocampo (as modified by Ce-ougna) to incorporate the teachings of Zhang and include “wherein the gene alteration state classification model comprises a supervised machine learning model”. The motivation for doing so would have been that “the attribute selection is carried out by the information gain of the random forest quantitative model in the random forest quantitative model submodule, so that the accuracy of the prediction of the fusion gene reliability prediction module can be improved” and that “various fusion variation types of a plurality of genes can be accurately identified”, as suggested by Zhang in para. [0043] and para. [0023], respectively. Therefore, it would have been obvious to one of ordinary skill at the time the invention was filed to combine Ocampo and Ce-ougna with Zhang to obtain the invention specified in claim 11.
Allowable Subject Matter
Claims 12-13 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
The following is a statement of reasons for the indication of allowable subject matter.
The best prior art of record is Ocampo, Ce-ougna, Cheng, Zhang, Bredno, and Prabhudesai. Prior art applied alone or in combination with fails to anticipate or render obvious claims 12-13.
Claim 12
Regarding claim 12, Ocampo and Ce-ougna teach the method of claim 1.
Ocampo further teaches training the gene alteration state classification model.
Ce-ougna further teaches buffer region image patches.
Cheng further teaches a plurality of samples from a cohort of subjects diagnosed with a disease of interest.
However, neither Ocampo, nor Ce-ougna, nor Cheng, nor Zhang, nor Bredno, nor Prabhudesai teaches wherein the gene alteration state classification model is trained on tumor region and buffer region image patches derived from slide images and associated gene alteration state data for a plurality of samples from a cohort of subjects diagnosed with a disease of interest.
Claim 13 includes allowable subject matter by virtue of being dependent upon claim 12.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
(JP 2023510915 A) teaches non-tumor segmentation to aid tumor detection and analysis
Any inquiry concerning this communication or earlier communications from the examiner
should be directed to KYLA G ALLEN whose telephone number is (703)756-5315. The examiner can
normally be reached M-F 7:30am - 4:30pm 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,
John Villecco can be reached on (571) 272-7319. 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.
/Kyla Guan-Ping Tiao Allen/
Examiner, Art Unit 2661
/JOHN VILLECCO/Supervisory Patent Examiner, Art Unit 2661
Prosecution Insights