Prosecution Insights
Last updated: November 04, 2026
Application No. 18/781,271

CELL NUCLEI CLASSIFICATION WITH ARTIFACT AREA AVOIDANCE

Final Rejection §103§112
Filed
Jul 23, 2024
Priority
Apr 05, 2021 — provisional 63/170,677 +1 more
Examiner
SUMMERS, GEOFFREY E
Art Unit
2669
Tech Center
2600 — Communications
Assignee
NEC Laboratories America Way
OA Round
2 (Final)
72%
Grant Probability
Favorable
3-4
OA Rounds
1m
Est. Remaining
99%
With Interview

Examiner Intelligence

Grants 72% — above average
72%
Career Allowance Rate
263 granted / 367 resolved
+9.7% vs TC avg
Strong +35% interview lift
Without
With
+35.2%
Interview Lift
resolved cases with interview
Typical timeline
2y 5m
Avg Prosecution
26 currently pending
Career history
399
Total Applications
across all art units

Statute-Specific Performance

§101
10.9%
-29.1% vs TC avg
§103
41.9%
+1.9% vs TC avg
§102
14.1%
-25.9% vs TC avg
§112
29.7%
-10.3% vs TC avg
Black line = Tech Center average estimate • Based on career data from 367 resolved cases

Office Action

§103 §112
DETAILED ACTION Response to Amendment Claims 1-12 were previously pending. Applicant’s amendment filed August 14, 2026, has been entered in full. Claims 1-10 are amended. No claims are added or cancelled. Claims 1-12 remain pending. Response to Arguments Applicant’s summary of the interview (Remarks filed August 14, 2026, hereinafter Remarks: Page 6) is noted. Applicant argues that written description and indefiniteness issues have been overcome by amendments to the claims (Remarks: Pages 6-7). Examiner respectfully disagrees as multiple issues inadequate written description and indefiniteness remain – see the rejections below, which have been updated in view of Applicant’s amendment. Applicant traverses the previous rejections under 35 U.S.C. 103, arguing that the previously-cited prior art does not teach all elements of the amended claims (Remarks: Pages 7-9). In particular, Applicant notes that “Shaban and Chen are silent on the concept of a ‘loss function’” (Remarks: Page 8). As was discussed during the interview, it is unclear to what extent the “loss function” clause added by Applicant’s amendment actually limits the scope of the claim. See the rejection under 35 U.S.C. 112(b) below. In order to promote compact prosecution (MPEP 2173.06), two sets of prior art rejections are made below: one where the noted “loss function” phrase is given no weight and another where the phrase is given full weight. ‘Cokluk’ (“Kullback-Leibler divergence loss vs (weighted) cross entropy loss,” 2019) is applied to teach a weighted loss function in the rejections giving the phrase full weight. See below for further explanation. Admitted Prior Art In the Office Action dated May 18, 2026, Examiner took Official Notice of facts in the following instance(s): At Page 15: “However, Examiner takes Official Notice that it is old and well-known in the art of image analysis to implement an image processing method as a system comprising a hardware processor; and a memory that stores a computer program, which, when executed by the hardware processor, causes the hardware processor to perform the method. Such computer implementation advantageously allows the image processing method to be performed quickly and easily. Use of a memory storing a computer program is specifically advantageous because it allows the hardware processor to be instructed how to accomplish the desired image processing method and for those instructions to be retained so that the method can continue to be performed over time.” Regarding Official Notice, MPEP 2144.03(C) includes the following instructions: “To adequately traverse such a finding, an applicant must specifically point out the supposed errors in the examiner’s action, which would include stating why the noticed fact is not considered to be common knowledge or well-known in the art.” “A general allegation that the claims define a patentable invention without any reference to the examiner’s assertion of official notice would be inadequate.” “If applicant does not traverse the examiner’s assertion of official notice or applicant’s traverse is not adequate, the examiner should clearly indicate in the next Office action that the common knowledge or well-known in the art statement is taken to be admitted prior art because applicant either failed to traverse the examiner’s assertion of official notice or that the traverse was inadequate. If the traverse was inadequate, the examiner should include an explanation as to why it was inadequate.” In the reply filed August 14, 2026, Applicant generally alleges that the claims define a patentable invention without any reference to Examiner’s assertion of Official Notice, which is an inadequate traverse. Therefore, as required by the MPEP, Examiner clearly indicates that the Official Notice statement(s) noted above is/are taken to be admitted prior art because Applicant either failed to traverse it/them or inadequately traversed it/them. Priority Applicant’s claim for the benefit of a prior-filed application under 35 U.S.C. 119(e) or under 35 U.S.C. 120, 121, 365(c), or 386(c) is acknowledged. Applicant has not complied with one or more conditions for receiving the benefit of an earlier filing date under 35 U.S.C. 119(e) as follows: The later-filed application must be an application for a patent for an invention which is also disclosed in the prior application (the parent or original nonprovisional application or provisional application). The disclosure of the invention in the parent application and in the later-filed application must be sufficient to comply with the requirements of 35 U.S.C. 112(a) or the first paragraph of pre-AIA 35 U.S.C. 112, except for the best mode requirement. See Transco Products, Inc. v. Performance Contracting, Inc., 38 F.3d 551, 32 USPQ2d 1077 (Fed. Cir. 1994). The disclosure of the prior-filed application, Application No. 63/170,677, fails to provide adequate support or enablement in the manner provided by 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph for one or more claims of this application. See the rejections under 35 U.S.C. 112(a) below for examples of claim limitations that are not adequately described in the provisional application. Claim Rejections - 35 USC § 112 The following is a quotation of the first paragraph of 35 U.S.C. 112(a): (a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention. The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112: The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention. Claims 1-12 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention. Claim 1 recites a combination of elements that is not adequately described in the written description for at least one of the following reasons. First, claim 1 recites “rescanning the tissue sample to generate a rescanned tissue sample image to be processed by the neural network model” (emphasis added). Examiner finds no description of processing a rescanned tissue sample image by the neural network model, nor any description of this being the intended use of a rescanned image. For example, pars. [0018] and [0023] (as published) mention the re-scanning of a tissue sample, but do not describe any intended further processing of the rescanned image, such as processing by the neural network. Furthermore, Examiner notes that many of the types of artifacts noted in the specification (e.g., [0023], inked areas, tissue folds, overstained areas) are artifacts in the tissue specimen itself that would not be resolved by rescanning. I.e., rescanning a tissue fold does not resolve or remove the artifact. For at least these reasons, one of ordinary skill in the art would not understand the written description to adequately describe ““rescanning the tissue sample to generate a rescanned tissue sample image to be processed by the neural network model” as recited in claim 1. Second, claim 1 recites “rescanning … including excluding areas of the tissue sample image within the or more detected artifact regions.” The original disclosure describes rescanning (see above) and also describes excluding areas of the tissue sample image within the or more detected artifact regions (e.g., [0027], [0033]), but does not describe that the rescanning includes excluding areas of the tissue sample image within the or more detected artifact regions. Instead, the re-scanning generates a new image (e.g., [0023]), while the described “exclusion” is an exclusion of certain areas from analysis of an image (e.g., [0027]). The specification does not describe, for example, removing parts of the image itself during rescanning. Instead, it describes ignoring artifact areas during analysis of a given image. Third, regarding the same limitation noted at the second point above, the specification does not describe any case where both (a) rescanning and (b) excluding artifact regions occur. For example, [0018] describes: “However, the slide may include areas where TCR calculations should be avoided, for example in the case where the slide includes areas with artifacts. When these artifact areas occupy too much of the slide area, the slide may need to be re-scanned or replaced.” Rescanning is clearly an alternative to excluding artifact areas and continuing analysis (i.e., TCR) of a given slide. Indeed, one of ordinary skill in the art would recognize that there would be no need to exclude areas of the tissue sample image if that tissue sample image is being replaced by a rescanned image. Fourth, claim 1 recites “performing an analysis of the rescanned tissue sample image.” Examiner has reviewed the original disclosure but finds no description of performing an analysis of a rescanned tissue sample image. As noted above, the specification does indicate that a need for re-scanning may be identified (e.g., [0018]) and re-scanning may be prompted (e.g., [0018]), but there is no description of later performing an analysis of a rescanned tissue sample image. For example, Fig. 2 illustrates analyzing a “new slide image” at step 209, the new slide image having been collected at 206 and analyzed to identify cells, tumor cells and artifacts at step 208. The analysis at 209 is not performed for a distinct, rescanned image. Furthermore, as discussed above, many of the types of artifacts noted in the specification (e.g., [0023], inked areas, tissue folds, overstained areas) are artifacts in the tissue specimen itself that would not be resolved by rescanning. E.g., rescanning a tissue fold does not resolve or remove the artifact. This would have suggested to one of ordinary skill in the art that a rescanned image would not necessarily be appropriate for performing a tissue sample image analysis, which would further suggest that the description of such an embodiment is inadequate. Claim 7 recites similar limitations and also lacks adequate written description for substantially the same reasons as claim 1. Claims 2-6 and 8-12 include the limitations of claims 1 or 7 and therefore also adequate written description for at least the same reasons. Claim 2 further recites that “rescanning the tissue sample image further comprises generating an artifact segmentation mask to exclude areas within a marked area provided by an operator.” This also lacks adequate written description for substantially the same as the second and third points raised with respect to claim 1. I.e., the specification does not describe generating an artifact segmentation mask as part of the segmentation and does not describe a scenario where rescanning and generating an artifact segmentation mask to exclude areas within a marked area provided by an operator both occur. Claim 8 recites similar limitations and also lacks adequate written description for substantially the same reasons. Claim 3 further recites that “rescanning the tissue sample image further includes obtaining a new tissue sample and scanning the new tissue sample.” This lacks adequate written description because the specification describes obtaining a new tissue sample (i.e., replacing the tissue sample) as an alternative to rescanning. For example, par. [0018] states that “the slide may need to be re-scanned or replaced” (emphasis added). In another example, par. [0023] states “These areas may need to be excluded from the determination of the report 110, and may prompt the re-scanning or replacement of the tissue sample 102 if the artifact areas occupy more than a threshold amount of the slide image” (emphasis added). Nowhere does the specification describe that “rescanning the tissue sample image further includes obtaining a new tissue sample and scanning the new tissue sample”. Claim 9 recites similar limitations and also lacks adequate written description for substantially the same reasons. Claim 4 further recites that “the analysis includes determining a tumor cell ratio for a portion of the rescanned tissue sample image that does not include an artifact region,” which lacks adequate written description for substantially the same reasons as in the fourth point raised in the rejection of claim 1. Claim 10 recites similar limitations and also lacks adequate written description for substantially the same reasons. Claim 7 recites a hardware processor that is caused to “rescan the tissue sample image”. The specification does not describe a hardware processor that can rescan a tissue sample image. At par. [0021], with reference to Fig. 1, the specification does describe that “[a] tissue sample 102 is taken and put on a slide. The slide may be scanned by a whole-slide scanner 104, producing a slide image, which is stored in a slide database 106.” However, a whole-slide scanner is not a hardware processor. Furthermore, there is no description of any hardware processor that obtains a tissue sample. There is also no explanation of how a hardware processor may obtain a tissue sample or scan a tissue sample. Fig. 6 does describe a computer including a hardware processor and a memory, but the specification does not explain how this computer and its processor can obtain a tissue sample or scan a tissue sample. A hardware processor, by itself, cannot obtain a tissue sample, mount it on a slide, or image that slide. Claims 8-12 include the limitations of claim 7 and therefore lack adequate written description for substantially the same reasons. Claims 8-9 further recite rescanning and further lack adequate written description for substantially the same reasons as claim 7. 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 1-12 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. As explained above in the rejections under 35 U.S.C. 112(a), many different aspects of the claimed invention are inconsistent with the specification. “A claim, although clear on its face, may also be indefinite when a conflict or inconsistency between the claimed subject matter and the specification disclosure renders the scope of the claim uncertain as inconsistency with the specification disclosure or prior art teachings may make an otherwise definite claim take on an unreasonable degree of uncertainty.” MPEP 2173.03. Claims 1-12 are indefinite at least because they include so many inconsistencies with the specification that they make the claims take on an unreasonable degree of uncertainty. Claim 1 recites a method comprising the following step: “analyzing a tissue sample image using a neural network model that includes a first output that identifies cells, a second output that identifies tumor cells, and a third output that identifies an artifact region, the neural network model optimized with a loss function including weights for the first output, the second output, and the third output” (emphasis added). Claim 1 is indefinite because it is unclear to what extent, if any, the phrase “the neural network model optimized with a loss function including weights for the first output, the second output, and the third output” limits the scope of the claim. First, it is unclear whether the neural network optimization is required to be performed as part of the claimed method, or whether the claim is merely describing the neural network as already having been optimized. On the one hand, recitation of optimization within a list of elements that a method “comprises” suggests that the optimization may be part of the method. On the other hand, “optimized” may be read simply as describing something that has previously been done to the neural network. Second, if the optimization is not required to be performed as part of the method, then it is further unclear to what extent the noted phrase limits the scope of the claim. “Claim scope is not limited by claim language that suggests or makes optional but does not require steps to be performed, or by claim language that does not limit a claim to a particular structure.” MPEP 2111.04, Subsection I. If the optimization step is not required to be performed as part of the method, then this limitation would only limit the scope of the claim if it limited the claim to a particular structure. One could argue that the structure of a neural network model is affected by the loss function used to train it. However, use of a specific loss function does not, in itself, limit a neural network model to a particular structure. In one example, use of a specific loss function does not affect a model’s architecture (e.g., numbers, sizes, types, and arrangement of layers). In another example, use of a specific loss function does not result in particular weight values or other parameters. Even if the same architecture and loss function are used, the parameters of an optimized neural network model will still vary based on what training dataset is used, any number of randomizations introduced during the training (e.g., random initialization, random data augmentation, stochastic gradient descent), etc. Furthermore, the act of using of an already-trained/optimized neural network to perform analysis (i.e., inference) is substantially the same regardless of what type of loss function was used to train/optimize it. This would all suggest to one of ordinary skill in the art that the noted clause of claim 1 does not limit the claim to a particular structure. On the other hand, the presence of the noted clause within the claim would suggest to one of ordinary skill in the art that Applicant is attempting to limit the claim in some way. In summary, claim 1 is indefinite because it is unclear whether (a) the recited optimization is required to be performed as a step in the method and (b), if not, to what extent the recited optimization limits the claim to a particular structure or otherwise limits the scope of the claim. Examiner notes that this issue may be resolved by amending claim 1 to recite a distinct step of “optimizing a neural network model that includes a first output that identifies cells, a second output that identifies tumor cells, and a third output that identifies an artifact region with a loss function including weights for the first output, the second output, and the third output;” prior to an amended step of “analyzing a tissue sample image using the neural network model;” because such an amendment would clarify that the optimization is required to be performed. Claim 7 recites a clause that is substantially the same as the noted clause of claim 1 and is also indefinite for substantially the same reasons. Claims 2-6 and 8-12 are also indefinite at least because they depend from one of indefinite claims 1 and 7. In order to promote compact prosecution (MPEP 2173.06), two sets of prior art rejections are made below: one where the noted “loss function” phrase is given no weight and another where the phrase is given full weight. 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. Rejections giving the “loss function” phrase no weight – see § 112(b) rejection Claim(s) 1-4, 6-10, and 12 is/are rejected under 35 U.S.C. 103 as being unpatentable over ‘Shaban’ (“A Novel Digital Score for Abundance of Tumour Infiltrating Lymphocytes Predicts Disease Free Survival in Oral Squamous Cell Carcinoma,” 2019) in view of ‘Chen’ (US 2022/0318979 A1)1. Regarding claim 1, Shaban teaches a computer-implemented method for analyzing a tissue sample (e.g., Figure 2; Page 7, Implementations), comprising: analyzing a tissue sample image (e.g., Fig. 2, WSI and/or sub-images) using a neural network model (e.g., Fig. 2, Deep Neural Network; Pg. 4, Tissue region classifier) that includes a first output that identifies cells (Fig. 2; Pg. 4, Tissue region classifier, Lymphocyte cell identification output), a second output that identifies tumor cells (Fig. 2; Pg. 4, Tissue region classifier, Tumour cell identification output), and a third output that identifies an artifact region (Fig. 2; Pg. 4, Tissue region classifier, Non-ROI identification output, which indicates the presence of slide preparation and scanning artifacts, such as tissue folding and blurring), the neural network model optimized with a loss function including weights for the first output, the second output, and the third output (This limitation has been interpreted as not limiting the scope of the claim – see rejection under 35 U.S.C. § 112(b) above); and excluding areas of the tissue sample image within one or more detected artifact regions (e.g., Pg. 4, Tissue region classifier, 1st par., “slide preparation and scanning artefacts … need to be ignored”; e.g., Pgs. 4-6, TIL quantification, only lymphocyte and tumor regions are considered for co-localization score analysis, with non-ROI/artifact areas being excluded from the co-localization score calculation). Shaban teaches ignoring artifact regions during analysis of its whole slide image (WSI) tissue sample regions (Pg. 4, Tissue region classifier) and the artifact regions in its example images are relatively minor and take up a small proportion of the image as a whole (e.g., Fig. 6, black non-ROI region is relatively small; Note that the Figure may best be seen in a color version of the reference). Shaban does not explicitly teach: rescanning the tissue sample image to generate a rescanned tissue sample image to be processed by the neural network model responsive to a determination that a proportion of the tissue sample image within one or more detected artifact regions exceeds a threshold; and performing an analysis of the rescanned tissue sample image. However, Chen does teach a WSI inspection technique (e.g., Fig. 1) that, like Shaban, identifies artifact regions (e.g., [0045]; ‘028 provisional: [0044]) and further includes: rescanning the tissue sample image to generate a rescanned tissue sample image (e.g., [0049], prompting a rescan; ‘028 provisional: [0048]) to be processed by the neural network model (Rescanned images are suitable for this intended use) responsive to a determination that a proportion of the tissue sample image within one or more detected artifact regions ([0049], “a number of pixels of the whole-slide image that include a level of image quality below a predefined amount …”) exceeds a threshold ([0049], “… exceeds a predefined limit (e.g., greater than 20% of the pixels of the whole-slide image)”) including excluding areas of the tissue sample within the one or more detected artifact regions (e.g., [0047]; ‘028 provisional: [0046]); and performing an analysis of the rescanned tissue sample image (e.g., [0034]-[0036], Fig. 1, scanned images are analyzed by image processing system 135; ‘028 provisional: [0033]-[0035]). Chen recognizes that artifacts may interfere with analysis and assessment of the image (e.g., [0003]; ‘028 provisional: [0002]). “For example, a region including an artifact may be incorrectly identified as being an area of interest (e.g., tumor cells) in the whole-slide image” ([0023]; ‘028 provisional: [0022]). In the context of Shaban’s image analysis and assessment, a large artifact area could obscure a significant portion of lymphocyte and/or tumor cells, thereby skewing the TILAb score (Pg. 6, The proposed TILAb score; Also see equations 1-3) and its prognostic capability. The rescanning corrective action of Chen can solve this problem by producing rescanned, corrected images without artifacts. Furthermore, automating the rescanning determination as taught by Chen is advantageous because it avoids an error-prone and time intensive manual review (e.g., [0023]). Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to modify the method of Shaban with the automated rescanning of Chen in order to improve the method with the reasonable expectation that this would result in a method that (a) avoided interference with or inaccuracy of slide image analysis and assessment, (b) obtained higher-quality images with fewer/no artifacts, and/or (c) avoided the need for an error-prone and time-intensive manual image quality review. This technique for improving the method of Shaban was within the ordinary ability of one of ordinary skill in the art based on the teachings of Chen. Therefore, it would have been obvious to one of ordinary skill in the art to combine the teachings of Shaban and Chen to obtain the invention as specified in claim 1. Regarding claim 2, Shaban in view of Chen teaches the computer-implemented method of claim 1, and Chen further teaches that rescanning the tissue sample image further comprises generating an artifact segmentation mask to exclude areas within a marked area provided by an operator (e.g., [0047]; ‘028 provisional: [0046]). Regarding claim 3, Shaban in view of Chen teaches the computer-implemented method of claim 1, and Chen further teaches that the rescanning the tissue sample image further includes obtaining a new tissue sample and scanning the new tissue sample (e.g., [0022], multiple tissue samples are obtained, each placed on its own slide; e.g., [0023], [0045], multiple slides may be processed as a set; e.g., [0049], corrective action may be to discard a whole-slide image; Discarding one image while processing a set of images for multiple slides, where each image corresponds to a different scan of a different tissue sample slice, includes obtaining a new tissue sample [i.e., the slice/slide for the next image] and scanning the new tissue sample [i.e., producing the whole-slide image of the next slice/slide]; ‘028 provisional: [0021]-[0022], [0044], [0048]). Regarding claim 4, Shaban in view of Chen teaches the computer-implemented method of claim 1, and Shaban further teaches that the analysis includes determining a tumor cell ratio (e.g., Supplementary Material: Section 1, equation 1, lymphocyte-to-tumour ratio) for a portion of the rescanned tissue sample image that does not include an artifact region (e.g., Supplementary Material: Section 1, equation 1, only portions classified as lymphocyte or tumour are considered in calculating the ratio; e.g., Pg. 4, Tissue region classifier, non-ROI portions are ignored; e.g., Fig. 4, only lymphocyte and tumour regions are considered – best seen in color). Regarding claim 6, Shaban in view of Chen teaches the computer-implemented method of claim 1, and Shaban further teaches that the neural network model identifies artifact regions and other outputs identifying objects (e.g., Fig. 2 and pg. 4, Tissue region classifier, neural network identifies artifact/non-ROI regions and produces other outputs identifying Tumour, Lymphocyte, and Stroma objects). Regarding claim 7, Examiner notes that the claim recites a system comprising a hardware processor; and a memory that stores a computer program, which, when executed by the hardware processor, causes the hardware processor to: perform a method that is substantially the same as the method of claim 1. Shaban in view of Chen teaches the method of claim 1 (see above). Shaban further teaches implementing its method using a system comprising a hardware processor that performs the method (Pg. 7, Implementations, processor and/or GPUs), but is silent regarding a memory that stores a computer program executed by the hardware processor. However, it has been taken as admitted prior art that it is old and well-known in the art of image analysis to implement an image processing method as a system comprising a hardware processor; and a memory that stores a computer program, which, when executed by the hardware processor, causes the hardware processor to perform the method. Such computer implementation advantageously allows the image processing method to be performed quickly and easily. Use of a memory storing a computer program is specifically advantageous because it allows the hardware processor to be instructed how to accomplish the desired image processing method and for those instructions to be retained so that the method can continue to be performed over time. Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to implement the image processing method of Shaban in view of Chen as a system comprising a hardware processor; and a memory that stores a computer program, which, when executed by the hardware processor, causes the hardware processor to perform the method with the reasonable expectation that this would result in a method that could be performed quickly and efficiently over time. Therefore, it would have been obvious to one of ordinary skill in the art to combine the teachings of Shaban and Chen to obtain the invention as specified in claim 7. Regarding claim 8, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 2. Shaban in view of Chen teaches the invention of claim 2 (see above). Accordingly, claim 8 is also rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Chen for substantially the same reasons as provided in the rejection of claim 2. Regarding claim 9, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 3. Shaban in view of Chen teaches the invention of claim 3 (see above). Accordingly, claim 9 is also rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Chen for substantially the same reasons as provided in the rejection of claim 3. Regarding claim 10, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 4. Shaban in view of Chen teaches the invention of claim 4 (see above). Accordingly, claim 10 is also rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Chen for substantially the same reasons as provided in the rejection of claim 4. Regarding claim 12, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 6. Shaban in view of Chen teaches the invention of claim 6 (see above). Accordingly, claim 12 is also rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Chen for substantially the same reasons as provided in the rejection of claim 6. Claim(s) 5 and 11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Chen as applied above, and further in view of ‘Szegedy’ (“Rethinking the Inception Architecture for Computer Vision,” 2016). Regarding claim 5, Shaban in view of Chen teaches the method of claim 1. Shaban teaches a neural network model (e.g., Fig. 2, Deep Neural Network) that produces output maps (e.g., Fig. 2, Prediction Map). Shaban tests using several different known types of convolutional neural network including Inception (Pg. 4, Tissue region classifier, 2nd paragraph; Pg. 7, Results, Tissue region classification), but does not explicitly describe details of their architectures. Instead, Shaban cites to other prior art references, such as the Szegedy reference for Inception (Pg. 4, Tissue region classifier, reference #32). Accordingly, Shaban itself does not explicitly teach that using an Inception convolutional neural network includes using a fully convolutional neural network model. However, Szegedy does teach that “Inception networks are fully convolutional” (Pg 2830, 1st full sentence). In view of Szegedy’s teachings, one of ordinary skill in the art would understand Shaban’s use of an Inception convolutional neural network to include use of a fully convolutional neural network as required by the claimed invention. Regarding claim 11, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 5. Shaban in view of Chen and Szegedy teaches the invention of claim 5 (see above). Accordingly, claim 11 is also rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Chen for substantially the same reasons as provided in the rejection of claim 5. Rejections giving the “loss function” phrase full weight – see § 112(b) rejection Claim(s) 1-4, 6-10, and 12 is/are rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of ‘Cokluk’ (“Kullback-Leibler divergence loss vs (weighted) cross entropy loss,” 2019) and Chen. Regarding claim 1, Shaban teaches a computer-implemented method for analyzing a tissue sample (e.g., Figure 2; Page 7, Implementations), comprising: analyzing a tissue sample image (e.g., Fig. 2, WSI and/or sub-images) using a neural network model (e.g., Fig. 2, Deep Neural Network; Pg. 4, Tissue region classifier) that includes a first output that identifies cells (Fig. 2; Pg. 4, Tissue region classifier, Lymphocyte cell identification output), a second output that identifies tumor cells (Fig. 2; Pg. 4, Tissue region classifier, Tumour cell identification output), and a third output that identifies an artifact region (Fig. 2; Pg. 4, Tissue region classifier, Non-ROI identification output, which indicates the presence of slide preparation and scanning artifacts, such as tissue folding and blurring), the neural network model optimized with a loss function including weights for the first output, the second output, and the third output (see Note Regarding Loss below); and excluding areas of the tissue sample image within one or more detected artifact regions (e.g., Pg. 4, Tissue region classifier, 1st par., “slide preparation and scanning artefacts … need to be ignored”; e.g., Pgs. 4-6, TIL quantification, only lymphocyte and tumor regions are considered for co-localization score analysis, with non-ROI/artifact areas being excluded from the co-localization score calculation). Note Regarding Loss. Shaban describes using a trained/optimized neural network (e.g., Pg. 4, Tissue region classifier). While neural network optimization is performed using a loss function, Shaban does not explicitly describe any specific loss function that it uses in the optimization. In particular, Shaban does not explicitly teach that the neural network model is optimized with a loss function including weights for the first output, the second output, and the third output. However, Cokluk does teach weighted cross entropy loss is a known type of loss function that may be used for multi-label classification tasks [i.e., the type of classification task performed by Shaban’s neural network] (e.g., Page 1). Specifically, weighted cross entropy is a “basic” cross entropy loss function (e.g., Pages 1-2) that is modified by applying weights to each target class (e.g., Page 2). Examiner notes that the target classes of Shaban’s neural network include first, second, and third outputs (Fig. 2; Pg. 4, Tissue region classifier, Lymphocyte cell identification output, Tumour cell output, and Non-ROI output, respectively), so modification of Shaban to use a weighted cross entropy loss function would result in a loss function including weights for the first output, the second output, and the third output as claimed. Cokluk teaches that the weights in weighted cross entropy may be used to reflect label’s importance levels (Pg. 2). Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to modify the method of Shaban with the weighted cross entropy loss function of Cokluk in order to improve the method with the reasonable expectation that this would result in a method that used a loss function that advantageously was suitable for its multi-label classification task and could reflect different labels’ importance levels. This technique for improving the method of Shaban was within the ordinary ability of one of ordinary skill in the art based on the teachings of Cokluk. Note Regarding Rescan. Shaban teaches ignoring artifact regions during analysis of its whole slide image (WSI) tissue sample regions (Pg. 4, Tissue region classifier) and the artifact regions in its example images are relatively minor and take up a small proportion of the image as a whole (e.g., Fig. 6, black non-ROI region is relatively small; Note that the Figure may best be seen in a color version of the reference). Shaban does not explicitly teach: rescanning the tissue sample image to generate a rescanned tissue sample image to be processed by the neural network model responsive to a determination that a proportion of the tissue sample image within one or more detected artifact regions exceeds a threshold; and performing an analysis of the rescanned tissue sample image. Cokluk also does not teach these features. However, Chen does teach a WSI inspection technique (e.g., Fig. 1) that, like Shaban, identifies artifact regions (e.g., [0045]; ‘028 provisional: [0044]) and further includes: rescanning the tissue sample image to generate a rescanned tissue sample image (e.g., [0049], prompting a rescan; ‘028 provisional: [0048]) to be processed by the neural network model (Rescanned images are suitable for this intended use) responsive to a determination that a proportion of the tissue sample image within one or more detected artifact regions ([0049], “a number of pixels of the whole-slide image that include a level of image quality below a predefined amount …”) exceeds a threshold ([0049], “… exceeds a predefined limit (e.g., greater than 20% of the pixels of the whole-slide image)”) including excluding areas of the tissue sample within the one or more detected artifact regions (e.g., [0047]; ‘028 provisional: [0046]); and performing an analysis of the rescanned tissue sample image (e.g., [0034]-[0036], Fig. 1, scanned images are analyzed by image processing system 135; ‘028 provisional: [0033]-[0035]). Chen recognizes that artifacts may interfere with analysis and assessment of the image (e.g., [0003]; ‘028 provisional: [0002]). “For example, a region including an artifact may be incorrectly identified as being an area of interest (e.g., tumor cells) in the whole-slide image” ([0023]; ‘028 provisional: [0022]). In the context of Shaban’s image analysis and assessment, a large artifact area could obscure a significant portion of lymphocyte and/or tumor cells, thereby skewing the TILAb score (Pg. 6, The proposed TILAb score; Also see equations 1-3) and its prognostic capability. The rescanning corrective action of Chen can solve this problem by producing rescanned, corrected images without artifacts. Furthermore, automating the rescanning determination as taught by Chen is advantageous because it avoids an error-prone and time intensive manual review (e.g., [0023]). Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to modify the method of Shaban in view of Cokluk as applied above with the automated rescanning of Chen in order to improve the method with the reasonable expectation that this would result in a method that (a) avoided interference with or inaccuracy of slide image analysis and assessment, (b) obtained higher-quality images with fewer/no artifacts, and/or (c) avoided the need for an error-prone and time-intensive manual image quality review. This technique for improving the method of Shaban in view of Cokluk was within the ordinary ability of one of ordinary skill in the art based on the teachings of Chen. Therefore, it would have been obvious to one of ordinary skill in the art to combine the teachings of Shaban, Cokluk and Chen to obtain the invention as specified in claim 1. Regarding claim 2, Shaban in view of Cokluk and Chen teaches the computer-implemented method of claim 1, and Chen further teaches that rescanning the tissue sample image further comprises generating an artifact segmentation mask to exclude areas within a marked area provided by an operator (e.g., [0047]; ‘028 provisional: [0046]). Regarding claim 3, Shaban in view of Cokluk and Chen teaches the computer-implemented method of claim 1, and Chen further teaches that the rescanning the tissue sample image further includes obtaining a new tissue sample and scanning the new tissue sample (e.g., [0022], multiple tissue samples are obtained, each placed on its own slide; e.g., [0023], [0045], multiple slides may be processed as a set; e.g., [0049], corrective action may be to discard a whole-slide image; Discarding one image while processing a set of images for multiple slides, where each image corresponds to a different scan of a different tissue sample slice, includes obtaining a new tissue sample [i.e., the slice/slide for the next image] and scanning the new tissue sample [i.e., producing the whole-slide image of the next slice/slide]; ‘028 provisional: [0021]-[0022], [0044], [0048]). Regarding claim 4, Shaban in view of Cokluk and Chen teaches the computer-implemented method of claim 1, and Shaban further teaches that the analysis includes determining a tumor cell ratio (e.g., Supplementary Material: Section 1, equation 1, lymphocyte-to-tumour ratio) for a portion of the rescanned tissue sample image that does not include an artifact region (e.g., Supplementary Material: Section 1, equation 1, only portions classified as lymphocyte or tumour are considered in calculating the ratio; e.g., Pg. 4, Tissue region classifier, non-ROI portions are ignored; e.g., Fig. 4, only lymphocyte and tumour regions are considered – best seen in color). Regarding claim 6, Shaban in view of Cokluk and Chen teaches the computer-implemented method of claim 1, and Shaban further teaches that the neural network model identifies artifact regions and other outputs identifying objects (e.g., Fig. 2 and pg. 4, Tissue region classifier, neural network identifies artifact/non-ROI regions and produces other outputs identifying Tumour, Lymphocyte, and Stroma objects). Regarding claim 7, Examiner notes that the claim recites a system comprising a hardware processor; and a memory that stores a computer program, which, when executed by the hardware processor, causes the hardware processor to: perform a method that is substantially the same as the method of claim 1. Shaban in view of Cokluk and Chen teaches the method of claim 1 (see above). Shaban further teaches implementing its method using a system comprising a hardware processor that performs the method (Pg. 7, Implementations, processor and/or GPUs), but is silent regarding a memory that stores a computer program executed by the hardware processor. However, it has been taken as admitted prior art that it is old and well-known in the art of image analysis to implement an image processing method as a system comprising a hardware processor; and a memory that stores a computer program, which, when executed by the hardware processor, causes the hardware processor to perform the method. Such computer implementation advantageously allows the image processing method to be performed quickly and easily. Use of a memory storing a computer program is specifically advantageous because it allows the hardware processor to be instructed how to accomplish the desired image processing method and for those instructions to be retained so that the method can continue to be performed over time. Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to implement the image processing method of Shaban in view of Cokluk and Chen as a system comprising a hardware processor; and a memory that stores a computer program, which, when executed by the hardware processor, causes the hardware processor to perform the method with the reasonable expectation that this would result in a method that could be performed quickly and efficiently over time. Therefore, it would have been obvious to one of ordinary skill in the art to combine the teachings of Shaban, Cokluk and Chen to obtain the invention as specified in claim 7. Regarding claim 8, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 2. Shaban in view of Cokluk and Chen teaches the invention of claim 2 (see above). Accordingly, claim 8 is also rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Cokluk Chen for substantially the same reasons as provided in the rejection of claim 2. Regarding claim 9, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 3. Shaban in view of Cokluk and Chen teaches the invention of claim 3 (see above). Accordingly, claim 9 is also rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Cokluk and Chen for substantially the same reasons as provided in the rejection of claim 3. Regarding claim 10, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 4. Shaban in view of Cokluk and Chen teaches the invention of claim 4 (see above). Accordingly, claim 10 is also rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Cokluk and Chen for substantially the same reasons as provided in the rejection of claim 4. Regarding claim 12, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 6. Shaban in view of Cokluk and Chen teaches the invention of claim 6 (see above). Accordingly, claim 12 is also rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Cokluk and Chen for substantially the same reasons as provided in the rejection of claim 6. Claim(s) 5 and 11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Cokluk and Chen as applied above, and further in view of ‘Szegedy’ (“Rethinking the Inception Architecture for Computer Vision,” 2016). Regarding claim 5, Shaban in view of Chen teaches the method of claim 1. Shaban teaches a neural network model (e.g., Fig. 2, Deep Neural Network) that produces output maps (e.g., Fig. 2, Prediction Map). Shaban tests using several different known types of convolutional neural network including Inception (Pg. 4, Tissue region classifier, 2nd paragraph; Pg. 7, Results, Tissue region classification), but does not explicitly describe details of their architectures. Instead, Shaban cites to other prior art references, such as the Szegedy reference for Inception (Pg. 4, Tissue region classifier, reference #32). Accordingly, Shaban itself does not explicitly teach that using an Inception convolutional neural network includes using a fully convolutional neural network model. However, Szegedy does teach that “Inception networks are fully convolutional” (Pg 2830, 1st full sentence). In view of Szegedy’s teachings, one of ordinary skill in the art would understand Shaban’s use of an Inception convolutional neural network to include use of a fully convolutional neural network as required by the claimed invention. Regarding claim 11, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 5. Shaban in view of Cokluk, Chen and Szegedy teaches the invention of claim 5 (see above). Accordingly, claim 11 is also rejected under 35 U.S.C. 103 as being unpatentable over Shaban in view of Cokluk, Chen and Szegedy for substantially the same reasons as provided in the rejection of claim 5. 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 nonprovisional extension fee (37 CFR 1.17(a)) 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 mailing date of this final action. Any inquiry concerning this communication or earlier communications from the examiner should be directed to GEOFFREY E SUMMERS whose telephone number is (571)272-9915. The examiner can normally be reached Monday-Friday, 7:00 AM to 3:30 PM ET. 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, Chan Park can be reached at (571) 272-7409. 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. /GEOFFREY E SUMMERS/Examiner, Art Unit 2669 1 Chen claims priority to, and incorporates by reference ([0001]), U.S. Provisional Patent Application No. 63/170,028 filed April 2, 2021 (the ‘028 provisional). Corresponding citations to the ‘028 provisional are provided in the prior art rejections in an effort to promote compact prosecution.
Read full office action

Prosecution Timeline

Jul 23, 2024
Application Filed
May 18, 2026
Non-Final Rejection mailed — §103, §112
Jul 23, 2026
Interview Requested
Jul 30, 2026
Examiner Interview Summary
Jul 30, 2026
Applicant Interview (Telephonic)
Aug 14, 2026
Response Filed
Oct 15, 2026
Final Rejection mailed — §103, §112 (current)

Precedent Cases

Applications granted by this same examiner with similar technology

Patent 12755390
ENHANCED MULTI-VIEW BACKGROUND MATTING FOR VIDEO CONFERENCING
3y 9m to grant Granted Oct 06, 2026
Patent 12725265
SYSTEM AND METHOD FOR DIFFERENTIATING A TISSUE OF INTEREST FROM ANOTHER PART OF A MEDICAL SCANNER IMAGE
3y 11m to grant Granted Sep 01, 2026
Patent 12718319
DEEP LEARNING BASED COPYING AND PASTING OF TRANSPARENT OBJECTS
3y 5m to grant Granted Aug 25, 2026
Patent 12718391
POLARIZED SEMI-GLOBAL MATCHING
2y 8m to grant Granted Aug 25, 2026
Patent 12718512
ROTATED OBJECT DETECTION
2y 7m to grant Granted Aug 25, 2026
Study what changed to get past this examiner. Based on 5 most recent grants.

Strategy Recommendation AI-generated — please review before filing

Get a prosecution strategy drawn from examiner precedents, rejection analysis, and claim mapping.
Typically takes 5-10 seconds — AI-generated, attorney review required before filing

Prosecution Projections

3-4
Expected OA Rounds
72%
Grant Probability
99%
With Interview (+35.2%)
2y 5m (~1m remaining)
Median Time to Grant
Moderate
PTA Risk
Based on 367 resolved cases by this examiner. Grant probability derived from career allowance rate.

Sign in with your work email

Enter your email to receive a magic link. No password needed.

Personal email addresses (Gmail, Yahoo, etc.) are not accepted.

Free tier: 3 strategy analyses per month