CTFR 18/476,320 CTFR 97617 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. The amendments provided 04/22/2026 have been entered and considered. Claims 1, 20, and 21 have been amended. Claims 12 and 13 have been cancelled. Response to Amendment Specification objections In view of the amendments provided 04/22/2026 to the specification (being a change in title), the objections to the specification are hereby withdrawn. Drawing objections In view of the replacement drawing 04/22/2026, the objections to the Drawings are hereby withdrawn. Response to Arguments 101 rejection On pages 12-13 of the remarks (04/22/2026), applicant contends that the amended independent claim limitations overcome the 101 rejection due to not being practically performable in a human mind, as well as stating that they improve the performance of an object detection model (indicating an improvement to the computer technology). The examiner agrees that the amended claims overcome the 101 rejection, specifically with regard to not being practically performable in a human mind. Prior art rejections On pages 13-14 of the remarks (04/22/2026), applicant contends that the amended independent claim limitations are not taught by Sakamoto, and are thus not anticipated by it. Specifically suggesting, Sakamoto does not explicitly teach the overlap calculations required by the amendments, nor the training of the object detection model. They further state that Sakamoto does not perform the selection of a second ROI from among the first ROI candidates. The examiner agrees, specifically with regard to Sakamoto not teaching the overlap calculations and object detection model training performed in the independent claims. As such, the 102 rejection of the non-final rejection is hereby withdrawn. On page 15 of the remarks (04/22/2026), applicant contends that Mansoor, Piao, and Kitamura doe not teach or suggest “determine a second region of interest candidate from among the first region-of-interest candidates based on the estimated image region by calculating a degree of overlapping between each of the first region-of-interest candidates and the estimated image region” and “training the object detection model by machine learning using training data including the second region-of-interest candidate”. The examiner respectfully disagrees. Piao teaches the overlap operation of the independent claims in [0081], which discloses the overlap determination of each candidate region with respect to an estimation position 24 which is further tied to a person region. Thus teaching a determination of overlap that results in a further second region (being one that is determined to overlap with the estimation position of the person region). Additionally, Kitamura teaches a determination of overlap with an estimation region (see [0074] which describes what is understood as an estimation region, being a correct rectangle, and see [0096] which discusses overlap determination). Further, Kitamura teaches the training of an object detection model using the resultant region of interest from the overlap determination (see [0097]-[0098], which discloses training based on a region of interest that has undergone an overlap calculation, and is understood as a second ROI candidate). Claim Objections 07-29-01 AIA Claim s 1, 20, and 21 objected to because of the following informalities: Claim 1 line 8 “objection detection model” should read as “object detection model”. Claim 20 line 6 “objection detection model” should read as “object detection model” Claim 21 line 6 “objection detection model” should read as “object detection model” Appropriate correction is required. There was no mention of an objection detection model found in the original disclosure, and an object detection model is referenced back to in the same claim. Said object detection model is also mentioned in the original disclosure multiple times. Claim Rejections - 35 USC § 102 07-07-aia AIA 07-07 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 – 07-08-aia AIA (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. 07-12-aia AIA (a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention. 07-15 AIA Claim s 1, 9-10, 14-17, and 19-21 are rejected under 35 U.S.C. 102 ( a)(1)/(a)(2 ) as being anticipated by Song et al. (US publication 20190050981 A1; hereinafter “Song”) . In re to claim 1, Song teaches wherein: an information processing apparatus comprising: one or more processors (Fig. 4 shows a processing device that comprises a processor) ; and one or more storage devices in which an instruction executed by the one or more processors is stored, wherein the one or more processors are configured to (Fig. 4 shows a memory that contains the programs used to perform image processing) : acquire an image (image data; [0038] discloses the acquisition of image data that is processed by the system) , related information (label data; [0024] discloses labels associated with classifications. See also [0032] which discloses labels connected to ground truth data. Further, see [0048] which denotes labels associated with target objects) related to the image ([0038] discloses use of a memory in communication with the processor to obtain image data and perform the various processing programs to implement the system’s methodology) ; generate one or more first region-of-interest candidates included in the image using an objection detection model ([0032] discloses the generation of bounding box that is further compared to a corresponding anchor box. It is understood that the bounding box to be compared to the anchor box is the first region-of-interest candidate. Additionally, the objection detection model is understood as the learning network disclosed in [0032] line 1 (that performs the generation of the various regions and bounding boxes described in [0032])) ; estimate one or more image regions indicated by the related information from the image and from the related information ([0032] discloses that the system establishes anchor boxes. Further, per [0049], the system may generate these anchor boxes with a corresponding label from a ground truth box when training the system. It is understood that when the anchor box is attributed a label that it is an estimated region indicated by related information (correspondent to the claims) ; determine a second region-of-interest candidate from among the first region-of- interest candidates based on the estimated image region by calculating a degree of overlapping between each of the first region-of-interest candidates and the estimated image region ([0032] lines 19-27 discloses the determination of positional and dimensional size offsets for bounding boxes in comparison to their corresponding anchor box. This is understood as the calculation of a degree of overlap by virtue of determining the level of offset being these areas. Further, the resultant bounding box with determined offsets is understood as the second region of interest candidate) ; and train the object detection model by machine learning using training data including the second region-of-interest candidate ([0044] discloses the various networks that comprise the object detection model (which includes that of the detection network). Further, [0050] discloses training using the second region of interest candidate, correspondent to the claims. See also [0062], which denotes training of the detection network using the mini-batches described in [0050]) . In re to claim 7 [dependent on claim 1], Song teaches wherein: the one or more first region-of-interest candidates include at least one of a bounding box, a heatmap, or a mask (anchor box; [0032] discloses the generation of boxes for each feature map grid cell. It is understood that the boxed area of the anchor box constitutes a bounding box) . In re to 8 [dependent on claim 1], Song teaches wherein: the one or more processors are configured to receive input of the image, the related information, and the one or more first region-of-interest candidates ([0038] discloses use of a memory in communication with the processor to obtain image data and perform the various processing programs to implement the system’s methodology. Additionally, as the processor is used to perform the processes of [0032], it is understood to use input of the image, related information, and first region of interest candidate (each correspondent to the claims, respectively)) . In re to claim 9 [dependent on claim 1], Song teaches wherein: the one or more processors are configured to: receive input of the image and the related information ([0038] discloses use of a memory in communication with the processor to obtain image data and perform the various processing programs to implement the system’s methodology) ; and acquire the one or more first region-of-interest candidates by generating the one or more first region-of-interest candidates based on the received image ([0032] discloses the generation of bounding box that is further compared to a corresponding anchor box. It is understood that the bounding box to be compared to the anchor box is the first region-of-interest candidate. Additionally, the network is an image processing algorithm that is processing image data (see Fig. 4). See also in [0032] that the operations take place with respect to detected objects) . In re to claim 10 [dependent on claim 9], Song teaches wherein: the one or more processors are configured to perform processing of disposing a plurality of bounding boxes as the first region-of-interest candidates at a constant interval on the image in a rule-based manner ([0032] discloses the generation of boxes for each feature map grid cell. it is understood that the boxed area of the anchor box constitutes a bounding box. Putting a bounding box at particular positions being understood as a “rule”. Additionally, as this is done according to a grid, it is understood to be at a constant interval) . In re to claim 14 [dependent on claim 1], Song teaches wherein: the one or more processors are configured to determine the first region-of-interest candidate included in the estimated image region among the one or more first region-of-interest candidates as the second region-of-interest candidate ([0032] lines 19-27 discloses the determination of positional and dimensional size offsets for bounding boxes in comparison to their corresponding anchor box. This is understood as the calculation of a degree of overlap by virtue of determining the level of offset being these areas. Further, the resultant bounding box with determined offsets is understood as the second region of interest candidate. Additionally, the network is an image processing algorithm that is processing image data (see Fig. 4), see also in [0032] that the operations take place with respect to detected objects) . In re to claim 15 [dependent on claim 1], Song teaches wherein: the one or more processors are configured to: acquire a plurality of the first region-of-interest candidates; and determine the second region-of-interest candidate from among the plurality of first region-of-interest candidates ([0032] lines 19-27 discloses the determination of positional and dimensional size offsets for bounding boxes in comparison to their corresponding anchor box. This is understood as the calculation of a degree of overlap by virtue of determining the level of offset being these areas. Further, the resultant bounding box with determined offsets is understood as the second region of interest candidate). In re to claim 16 [dependent on claim 1], Song teaches wherein: the one or more processors are configured to calculate a probability for the image region indicated by the related information in pixel units of the image ( [0024] discloses labels associated with classifications . Further, [0047] discloses that the classification network may perform calculations to determine loss with respect to a lost function regarding classification. As the label is a determined in respect of classification, it is understood that a probability for the image region indicated by the related information is performed. See also [0050] and formula 3, which indicate labels to have an associated probability related to their class. Further, the result of the loss determination being understood in units related to the pixel data of the image (thus being pixel units)). In re to claim 17 [dependent on claim 1], Song teaches wherein: the image region estimated by the one or more processors includes at least one of a bounding box, a heatmap, or a mask ([0032] discloses the generation of bounding box that is further compared to a corresponding anchor box). In re to claim 19 [dependent on claim 1], Song teaches wherein: the one or more processors are configured to: calculate an evaluation value of the one or more first region-of-interest candidates from the estimated image region ([0032] discloses the calculation of offset for size with respect to the first region-of-interest candidate (correspondent to the claims). It is understood that the offset determination for size is an evaluation value) ; and determine the second region-of-interest candidate based on the evaluation value ([0032] lines 19-27 discloses the determination of positional and dimensional size offsets for bounding boxes in comparison to their corresponding anchor box. This is understood as the calculation of a degree of overlap by virtue of determining the level of offset (determined by the offsets per [0032]) being these areas. Further, the resultant bounding box with determined offsets is understood as the second region of interest candidate) . As to claim 20, it is the method executed by the apparatus of claim 1. As such, it recites similar limitations to claim 1 and is rejected for the same reasons as provided above. As to claim 21, it is the non-transitory computer readable tangible recording medium that performs the processes of the apparatus of claim 1. As such, it recites similar limitations to claim 1 and is rejected for the same reasons as provided above . Claim Rejections - 35 USC § 103 07-20-aia AIA 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. 07-21-aia AIA Claim s 2-6 are rejected under 35 U.S.C. 103 as being unpatentable over Song in view of Cherubini et al. (US publication 20230050833 A1; hereinafter “Cherubini”) . In re to claim 2 [dependent on claim 1], Song does not explicitly teach wherein: the related information includes a text related to a content of the image However, in the same field of endeavor, Cherubini teaches wherein: the related information includes a text related to a content of the image ([0065] discloses that the associated label information to that of classifications done by the system may be represented by textual overlays) . Cherubini, like Song, discloses a medical image processing system that performs classification operations on input image data. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Song to utilize text, as taught by Cherubini . The motivation for the proposed modification would have been to provide the user information that is easily understood by a human, increasing their understanding of the processing operations of the system. In re to claim 3 [dependent on claim 1], Song does not explicitly teach wherein: the related information includes a text described with respect to a region of interest included in the image () . However, in the same field of endeavor, Cherubini teaches wherein: the related information includes a text described with respect to a region of interest included in the image ([0065] discloses that the associated label information to that of classifications done by the system may be represented by textual overlays. Further, [0064] discloses that overlays may include borders indicating objects of a region (understood as a region of interest)) . Cherubini, like Song, discloses a medical image processing system that performs classification operations on input image data. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Song to utilize text, as taught by Cherubini . The motivation for the proposed modification would have been to provide the user information that is easily understood by a human, increasing their understanding of the processing operations of the system. In re to claim 4 [dependent on claim 1], Song teaches wherein: the related information includes information about a label including at least one of a size, a position, or a property of a region of interest included in the image ([0048] discloses labels as indicating a property of a region of interest, such as what an area contains) . Song does not explicitly teach wherein: label data is a structured text However, in the same field of endeavor, Cherubini teaches wherein: label data is a structured text ([0065] discloses that the associated label information to that of classifications done by the system may be represented by textual overlays that represent a structured word). Cherubini, like Song, discloses a medical image processing system that performs classification operations on input image data. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Song to utilize text, as taught by Cherubini . The motivation for the proposed modification would have been to provide the user information that is easily understood by a human, increasing their understanding of the processing operations of the system. In re to claim 5 [dependent on claim 2], Song teaches wherein: the one or more processors are configured to estimate at least one of a position, a size, or a property indicated by the text ([0030] discloses labels attributed to bounding boxes related to the classification result. Additionally, [0048] discloses labels as indicating a property of a region of interest, such as what an area contains) . In re to claim 6 [dependent on claim 2], Song, in view of Cherubini, teaches wherein: the image is a medical image (Fig. 1 shows an exemplary image that is a volumetric chest CT image. See also Fig. 4, which shows the image data processed by the system is medical image data) the one or more processors are configured to: recognize an organ included in the image ([0022] discloses that the target processed may be an organ, and provides an example that the recognition of nodules may include a recognition of a lung nodule (thus being understood as recognition of an organ included in the image)) ; and Estimate the region from the text (Cherubini [0065] discloses that the associated label information to that of classifications done by the system may be represented by textual overlays that represent a structured word) and from a recognition result of the organ ([0032] discloses that the system establishes anchor boxes. Further, per [0049], the system may generate these anchor boxes with a corresponding label from a ground truth box when training the system. It is understood that when the anchor box is attributed a label that it is an estimated region indicated by related information (correspondent to the claims). As the labels may indicate presence of a nodule (see [0048]), it is understood to be an estimation of a recognition result (being the recognition of the presence of a nodule)) . The reason for combination is the same as provided above . 07-21-aia AIA Claim 11 is rejected under 35 U.S.C. 103 as being unpatentable over Song in view of Mansoor et al. (non-patent literature titled “REGION PROPOSAL NETWORKS WITH CONTEXTUAL SELECTIVE ATTENTION FOR REAL-TIME ORGAN DETECTION”; hereinafter “Mansoor”) . In re to claim 11 [dependent on claim 9], Song teaches wherein: the one or more processors are configured to generate the one or more first region-of-interest candidates from the image using the machine learning model to receive input of the image and estimate the one or more first region-of-interest candidates from the image ([0032] discloses the generation of boxes for each feature map grid cell by a 3D learning network) . Song does not explicitly teach wherein: using a machine learning model that is trained to receive input of the image and estimate the one or more first region-of-interest candidates from the image However, in a similar field of endeavor, Mansoor teaches wherein: using a machine learning model that is trained to receive input of the image and estimate the one or more first region-of-interest candidates from the image (section 2 discloses the utilized networks leveraged by the system to perform its methodology. Additionally, as the system comprises a neural network architecture, it is understood to be a machine learning model. Further, sections 2.2.2 and 2.2.3 disclose training being performed for the networks that comprise the system (which per Fig. 1 and section 2 is used to process input image data for the generation of first region-of-interest candidates, correspondent to the claims)) . Mansoor, like Song, is a medical focused image processing system that performs actions with respect to anatomical structures in image data. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Song to use a trained learning model, as taught by Mansoor. The motivation for the proposed modification would have been to make it so that the system is able to acquire input of image data by itself, reducing potential input of directions by a user . 07-21-aia AIA Claim 18 is rejected under 35 U.S.C. 103 as being unpatentable over Song in further view of Piao (US publication 20210073564 A1; hereinafter “Piao”) and Kitamura et al. (US publication 20220004797 A1; hereinafter “Kitamura”) In re to claim 18 [dependent on claim 1], Song, does not explicitly teach wherein: the one or more processors are configured to: calculate a confidence degree of the first region-of-interest candidate; and However, in a similar field of endeavor, Piao teaches wherein: the one or more processors are configured to: calculate a confidence degree of the first region-of-interest candidate ([0027] discloses the determination of a probability that the target object is represented within a region (understood as a confidence degree)) . Piao, like Song, is an image processing system that performs segmentation operations on the input image data. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Song, in view of Mansoor, to calculate a confidence degree, as taught by Piao. The motivation for the proposed modification would have been to decrease the likelihood that an area does not include objects of interest by requiring that they meet a threshold value (as is performed in Piao [0027]). Song in view of Piao, does not explicitly teach wherein: delete the first region-of-interest candidate not corresponding to the estimated image region among the one or more first region-of-interest candidates However, in a similar field of endeavor, Kitamura teaches wherein: to delete the first region-of-interest candidate not corresponding to the estimated image region among the one or more first region-of-interest candidates ([0141] and [142] disclose the deletion of candidate regions with respect to an area that constitutes a feature map (understood to be an estimated region, being a determined region output from another network, per [0062])) . Kitamura, like Song, discloses a medically focused image processing system that performs region based analysis of medical image data. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Song, in view Piao, to include deletion of regions, as taught by Kitamura. The motivation for the proposed modification would have been to reduce the number of region proposals that the system has to process while still considering a plurality of regions. Conclusion 07-39 AIA THIS ACTION IS MADE FINAL. 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 KEVIN M COOMBER whose telephone number is (571)270-0950. The examiner can normally be reached Monday - Friday 8:00am-5:00pm. 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, Gregory Morse can be reached at (571) 272-3838. 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. /KEVIN M COOMBER/Examiner, Art Unit 2663 /GREGORY A MORSE/Supervisory Patent Examiner, Art Unit 2698 Application/Control Number: 18/476,320 Page 2 Art Unit: 2663 Application/Control Number: 18/476,320 Page 3 Art Unit: 2663 Application/Control Number: 18/476,320 Page 4 Art Unit: 2663 Application/Control Number: 18/476,320 Page 5 Art Unit: 2663 Application/Control Number: 18/476,320 Page 6 Art Unit: 2663 Application/Control Number: 18/476,320 Page 7 Art Unit: 2663 Application/Control Number: 18/476,320 Page 8 Art Unit: 2663 Application/Control Number: 18/476,320 Page 9 Art Unit: 2663 Application/Control Number: 18/476,320 Page 10 Art Unit: 2663 Application/Control Number: 18/476,320 Page 11 Art Unit: 2663 Application/Control Number: 18/476,320 Page 12 Art Unit: 2663 Application/Control Number: 18/476,320 Page 14 Art Unit: 2663 Application/Control Number: 18/476,320 Page 15 Art Unit: 2663
Prosecution Insights