DETAILED ACTION
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on Mar 23 2026 has been entered.
Applicant’s response, filed Mar 23 2026, has been fully considered. Rejections and/or objections not reiterated from previous Office Actions are hereby withdrawn. The following rejections and/or objections are either reiterated or newly applied. They constitute the complete set presently being applied to the instant application.
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 .
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
Claim Status
Claims 1-5 and 7-20 are pending.
Claim 6 is canceled.
Claims 1-5 and 7-20 are rejected.
Priority
This application is a CON of 17/613,009, filed Nov 19 2021, which claims priority to PCT/JP2020/021017, filed May 27 2020, and foreign application No. JP2019-099716, filed May 28 2019. Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55 in parent application 17/613,009.
Claims 1-5 and 7-20 are accordingly afforded the effective filing date of May 28 2019.
Information Disclosure Statement
The information disclosure statement (IDS) filed on Mar 5 2026 is in compliance with the provisions of 37 CFR 1.97 and has therefore been considered. A signed copy of the IDS document is included with this Office Action.
Claim Rejections- 35 USC § 112
The outstanding rejections to the claims are withdrawn in view of the amendments submitted herein.
35 USC § 112(b)
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.
Claims 1-5, 7-18, and 20 are rejected under 35 U.S.C. 112(b) as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, regards as the invention. The rejection is maintained from the previous Office Action and updated based on claim amendment.
Claim 1 recites “A sorting system comprising: at least one hardware processor; and at least one non-transitory computer-readable storage medium storing processor-executable instructions that, when executed by the at least one hardware processor, cause the at least one hardware processor to perform controlling the sorting system to sort subsequent biological particles using a result of applying a trained machine learning model to subsequent optical data indicative of light emitted from the subsequent biological particles, the controlling comprising… transmitting one or more instructions, based on the result of applying the trained machine learning model, to the sorting system that cause the sorting system to perform a sorting step comprising: sorting, using the sorting system, the subsequent biological particles based on the one or more instructions”. Under the BRI, the sorting system of claim 1 is comprised on at least one hardware processor; and at least one non-transitory computer-readable storage medium storing processor-executable instructions and does not explicitly include any structure. Claim 1 requires that the sorting system is controlled to sort the subsequent biological particles. Claim 20 recites “At least one non-transitory computer-readable storage medium storing processor-executable instructions that, when executed by the at least one hardware processor, cause the at least one hardware processor to perform: controlling a sorting system to sort subsequent biological particles…” by performing the recited method steps and finally “transmitting one or more instructions, based on the result of applying the trained machine learning model, to the sorting system that cause the sorting system to perform a sorting step comprising: sorting, using the sorting system, the subsequent biological particles based on the one or more instructions”. The at least one non-transitory computer-readable storage medium of claim 20 includes only processor-executable instructions. The interpretation of the limitations directed to “sorting” requires the physical sorting of the subsequent biological particles, which requires specific physical equipment which is separate from the processor and computer-readable storage medium recited in the claims. The specification as published describes the sorting system as comprising an information processing apparatus and a sorting apparatus that acquires measurement data from a sample and that sorts particles to be sorted based on a determination made by the information processing apparatus (FIG. 1; [0040; 0056; 0088; 0092; 0102]. The specification as published discloses that the sorting apparatus includes a measurement unit and a sorting unit which may be a flow cell type or microchannel chip type [0044]. The specification as published also discusses that the sorting system may be formed of only the sorting apparatus, where the sorting apparatus includes the function of the information processing apparatus. However, the specification makes clear that the sorting of the biological particles is performed by a sorting apparatus and not by a processor or a computer-readable storage medium alone. The claims therefore are interpreted as reciting functional limitations because they recite the act of sorting biological particles rather than the structure required (see MPEP 2173.05(g)). The functional limitations fail to provide a clear-cut indication of the scope of the subject matter embraced by the claim, and are therefore considered to be indefinite. The rejection may be overcome by amending the claims to include a sorting apparatus or at least a sorting unit as described in the specification. Claims 2-5 and 7-18 are rejected based on their dependency from claim 1.
Response to Applicant Arguments
Applicant submits that the claims as amended, and as understood by one of ordinary skill upon reading the claims in light of the specification, are not indefinite.
It is respectfully submitted that this is not persuasive. As currently recited, the sorting system of claim 1 and the sorting system of claim 20 that is controlled by the instructions on the computer-readable storage medium merely encompass a general purpose computer with at least one processor and/or instructions. Claims 1 and 20 therefore do not include the structure required to perform the action of sorting the cells. The claims recite functional limitations which are indefinite, as described above. While claims must be given a broadest reasonable interpretation that is consistent with the specification as it would be interpreted by one of ordinary skill in the art (see MPEP 2173.01(I)), a claim is indefinite when the boundaries of the protected subject matter are not clearly delineated and the scope is unclear (see MPEP 2173.02(I)). Functional limitations render the boundaries of the claim scope indefinite when the claims merely recite a description of a problem to be solved or a function or result achieved by the invention (see MPEP 2173.05(g)), as is the case with the instant claims.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
A. Claims 1-4, 7-17, and 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Becht et al. (Bioinformatics, published 21 Jun 2018, 35(2):301-308; previously cited) in view of Lai et al. (US 2020/0105376; priority to Oct 1 2018; previously cited). Claim 17 is further evidenced by Radisic et al. (International Journal of Nanomedicine, 2006, 1(1):3-14; previously cited). Instantly claimed elements which are considered to be equivalent to the prior art teachings are described in bold for all claims. Any newly recited portions are necessitated by claim amendment.
The prior art to Becht discloses Hypergate, an algorithm which given a cell population of interest identifies a gating strategy optimized for high yield and purity (abstract). Becht, indicated by the open circles, teaches the instant features, indicated by the closed circles, as follows.
Claim 1 discloses a sorting system comprising at least one hardware processor, and at least one non-transitory computer-readable storage medium storing processor-executable instructions that, when executed by the at least one hardware processor, cause the at least one hardware processor to perform the method. Claim 19 discloses a method. Claim 20 discloses at least one non-transitory computer-readable storage medium storing processor-executable instructions that, when executed by the at least one hardware processor, cause the at least one hardware processor to perform the method. Becht teaches that Hypergate is an algorithm which is implemented computationally and available on CRAN (abstract), which is considered to read on the instant system and non-transitory computer-readable storage medium.
The method steps of claims 1 and 19-20 comprise:
controlling the sorting system to sort subsequent biological particles using a result of applying a trained machine learning model to subsequent optical data indicative of light emitted from the subsequent biological particles, the controlling comprising:
See the below steps as taught by Becht and Lai.
applying a nonlinear process to optical data indicative of light emitted from biological particles to generate compressed data;
Becht teaches the analysis of high-dimensional flow cytometry datasets generated from biological samples labeled with fluorochrome-conjugated antibodies (p. 302, col. 1, par. 4; p. 304, col. 1, par. 2-3), which is considered to read on optical data indicative of light emitted from biological particles as instantly claimed. Such an interpretation is supported by the instant specification at least at [0003-0006].
Becht teaches processing the data with dimensionality reduction techniques, including t-SNE, for visualization (p. 302, col. 1, par. 2; p. 303, col. 2, par. 3), which is considered to read on a non-linear process as instantly claimed. Such an interpretation is supported by instant claim 2 and the instant specification as published at least at [0051; 0061-0063], which provides examples of non-linear processing including dimensional compression, clustering and grouping, including the use of algorithms including t-SNE.
generating the trained machine learning model by training a machine learning model using training data comprising the optical data and information regarding sorting of the biological particles specified based on the compressed data;
Becht teaches that the inputs of Hypergate are the dataset, which is an expression matrix of events and parameters (i.e., optical data), and a cluster of interest which has been manually identified from clustered data (i.e., information regarding sorting of the biological particles specified based on the compressed data) (Fig. 1C; p. 302, col. 2, par. 3-4; p. 303, col. 2, par. 2-5). Becht teaches training Hypergate to obtain a gating strategy that identifies a malignant population of cells in a sample (i.e., training data) and applying these gating strategies to follow-up samples (i.e., subsequent samples as recited above and below) (p. 304, col. 1, par. 4 through col. 2, par. 1; p. 307, col. 1 through p. 308, col. 1, par. 3; Fig. 6).
applying, using the sorting system, the trained machine learning model to the subsequent optical data indicative of light emitted from the subsequent biological particles by inputting the subsequent optical data to the trained machine learning model; and transmitting one or more instructions, based on the result of applying the trained machine learning, model, to the sorting system that cause the sorting system to perform a sorting step comprising: sorting, using the sorting system, the subsequent biological particles based on the one or more instructions.
As Becht teaches that Hypergate optimizes sorting strategies (p. 307, col. 1, par. 4) and general methods for cell sorting (p. 307, col. 2, par. 2), it is considered that Becht fairly teaches a sorting system as instantly claimed.
Becht does not teach applying the trained learning model to subsequent optical data indicative of light emitted from subsequent biological particles by inputting the subsequent optical data to the trained learning model, transmitting instructions to sort the subsequent biological particles based on result of applying the trained learning model, or sorting the subsequent biological particles. Although Becht teaches training Hypergate, Becht does not explicitly teach that Hypergate is a machine learning model.
However, the prior art to Lai discloses methods and systems for a deep-learning platform for sorting cell populations (abstract). Lai teaches selecting a machine learning model which was trained based on similar phenotype information as indicated in the flow cytometry data, and applying the machine learning model to the flow cytometry data (abstract; claims 1, 8, and 15; [0042-0043]). Lai teaches that the trained machine learning models may analyze the input information and generate output classification information [0060]. Lai teaches instructing a flow cytometer to sort cell populations according to the assigned classifications (claim 19; [0043]).
Regarding claims 1 and 19-20, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine, in the course of routine experimentation and with a reasonable expectation of success, the methods of Becht and Lai because both references disclose methods for machine learning models for gating or classifying populations of cells in flow cytometry data. The motivation to apply the trained learning model taught by Becht to subsequent data would have been to select a particular machine learning model which is optimized, or otherwise advantageous, for the input information, while applying automatic gating tools to avoid user introduced variability, as taught by Lai [0005; 0043]. The motivation to train a machine learning model to perform the steps of Hypergate on the same data as examined by Hypergate in order to develop a gating strategy for a population of interest, as taught by Becht (p. 302, col. 1, par. 3) would have been to perform a substantially automated analysis of cell populations using supervised learning techniques, as taught by Lai [0006].
Regarding claim 2, Becht in view of Lai teaches the sorting system of claim 1 as described above. Claim 2 further adds that the nonlinear process is dimensional compression.
Becht teaches performing t-SNE on datasets for input to Hypergate (p. 303, col. 2, par. 3 and 5), which is considered to read on a nonlinear process of dimensional compression, as supported by the instant specification as published at [0074; 0084].
Regarding claim 3, Becht in view of Lai teaches the sorting system of claims 1 and 2 as described above. Claim 3 further adds that the dimensional compression compresses dimensions of the optical data into three dimensions or less.
Becht teaches visualizing the tSNE gated cells by tSNE1 and tSNE2 (Fig. 2A), which is considered to read on two dimensions.
Regarding claim 4, Becht in view of Lai teaches the sorting system of claim 1 as described above. Claim 4 further adds that the optical data is information obtained by performing fluorescent separation on the light emitted from the biological particles to obtain a level of expression of fluorescent dye of each color of a plurality of colors.
Becht teaches that the flow cytometry datasets have multiple markers (p. 302, col. 1, par. 4; p. 304, col. 1, par. 2). Becht teaches visualizing the cells by plotting two markers against one another (Fig. 2C; Fig. 6B-D), which is considered to read on fluorescent separation on the light emitted from the biological particles to obtain a level of expression of fluorescent dye of each color of a plurality of colors as instantly claimed.
Regarding claims 7, Becht in view of Lai teaches the sorting system of claim 1 as described above. Claim 7 further adds that training the machine learning model comprises performing supervised learning using the training data.
Becht teaches sampling negative and positive events from each dataset and for each cell population manually gated to make a training set to train six different classifiers (p. 303, col. 2, par. 9 through p. 304, col. 1, par. 1). Becht teaches a supervised version of Hypergate (p. 305, col. 1, par. 2).
Lai also teaches supervised learning techniques for machine learning [0006].
Regarding claims 8-9, Becht in view of Lai teaches the sorting system of claim 1 as described above. Claim 8 further adds that the information regarding sorting of the biological particles is information indicating whether to sort the biological particles. Claim 9 further adds that the information regarding sorting of the biological particles is information indicating to which collection unit the biological particles are to be sorted.
As Becht teaches that Hypergate optimizes sorting strategies (p. 307, col. 1, par. 4), it is considered that Becht fairly teaches whether to sort the biological particles, which also reads on which collection unit the biological particles are to be sorted to, i.e., whether they are collected or not.
Regarding claims 10-11, Becht in view of Lai teaches the sorting system of claim 1 as described above. Claim 10 further adds that generating the trained machine learning model from the optical data and information regarding sorting of the biological particles specified based on the compressed data further comprises determining whether a rate of correct answers of the generated trained machine learning model exceeds a threshold value. Claim 11 further adds that generating the trained machine learning model from the optical data and information regarding sorting of the biological particles specified based on the compressed data further comprises dividing the optical data into first and second portions, generating the trained machine learning model using the first portion of the optical data and calculating the rate of correct answers by applying the trained machine learning model to the second portion of the optical data.
Becht teaches that Hypergate terminates when no move increases the Fβ score or the last channel added did not contribute more than a user-specified threshold (p. 303, col. 1, par. 5) As Becht teaches that the Fβ score reflects the number of true positive, false negative, and false positive events, it is considered that Becht fairly teaches a rate of correct answers exceeding a threshold as instantly claimed. Becht further teaches training six classifiers on a training set of data (i.e., first portion) and then predicting the class of cells in a left-out, or test, dataset (i.e., second portion) to compute the corresponding F1-scores and accuracies (i.e., the rate of correct answers) (p. 304, col. 1, par. 1).
Regarding claim 12, Becht in view of Lai teaches the sorting system of claims 1 and 6 as described above. Claim 12 further adds that generating the trained machine learning model from the optical data and information regarding sorting of the biological particles specified based on the compressed data further comprises generating a notification indicating completion of the trained machine learning when a rate of correct answers of the trained machine learning model exceeds a threshold.
Becht teaches that Hypergate terminates when no move increases the Fβ score or the last channel added did not contribute more than a user-specified threshold (p. 303, col. 1, par. 5) As Becht teaches that the Fβ score reflects the number of true positive, false negative, and false positive events, it is considered that Becht fairly teaches a rate of correct answers exceeding a threshold as instantly claimed. Becht further teaches training six classifiers on a training set of data (i.e., first portion) and then predicting the class of cells in a left-out, or test, dataset (i.e., second portion) to compute the corresponding F1-scores and accuracies (i.e., the rate of correct answers) (p. 304, col. 1, par. 1). Becht teaches biplots comparing the F1 scores and accuracy in binary classification of the classifiers (Fig. 5), which is considered to teach generating a notification.
Regarding claim 13-14, Becht in view of Lai teaches the sorting system of claim 1 as described above. Claim 13 further adds outputting the compressed data. Claim 14 further adds that outputting the compressed data further comprises mapping the compressed data to an area of three dimensions or less.
Becht also teaches visualizing the tSNE gated cells by tSNE1 and tSNE2 (Fig. 2), which is considered to read on outputting by mapping the compressed data to two dimensions.
Regarding claim 15, Becht in view of Lai teaches the sorting system of claim 1 as described above. Claim 15 further adds applying the non-linear process to a second set of optical data to generate a second set of compressed data, the second set of optical data including the optical data indicative of light emitted from the biological particles used for generating the trained machine learning model and the subsequent optical data indicative of light emitted from the subsequent biological particles, and outputting the second set of compressed data.
Becht is considered to teach applying a non-linear process to the optical data to generate a compressed dataset (p. 302, col. 1, par. 4; p. 304, col. 1, par. 2-3) and subsequent optical data (p. 304, col. 1, par. 4 through col. 2, par. 1; p. 307, col. 1 through p. 308, col. 1, par. 3; Fig. 6), as described above. Becht teaches applying the non-linear process to subsequent data at least in Fig 4B and Fig. 6C-D.
Regarding claim 16, Becht in view of Lai teaches the sorting system of claim 1 as described above. Claim 16 further adds that the biological particles are cells.
Becht teaches analyzing data generated from cells (abstract; p. 303, col. 1, par. 4).
Regarding claim 17, Becht in view of Lai teaches the sorting system of claim 1 as described above. Claim 17 further adds that the sorting system comprises a light source configured to irradiate laser light to the biological particles; and a photodetector configured to obtain the light emitted from the biological particles, the light being emitted in response to irradiating the biological particles with the laser light.
Becht teaches the use of flow cytometry to collect fluorescent data from single cells, as described above (abstract; p. 302, col. 1, par. 4; p. 304, col. 1, par. 2-4). It is considered that a light source to irradiate laser light to the biological particles and a photodetector configured to obtain the light emitted from the biological particles are inherent features of a flow cytometer which analyzes fluorescently labeled cells, as evidenced by Radisic (p .5, col. 2, par. 3 through p. 7, col. 1, par. 1).
B. Claims 5 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Becht in view of Lai, as applied to claims 1 and 4 above, and as evidenced by Radisic, as applied to claim 17 above, and in further view of Novo et al. (Cytometry Part A 2013, 83(5):508-520; previously cited). Instantly claimed elements which are considered to be equivalent to the prior art teachings are described in bold for all claims. Any newly recited portions are necessitated by claim amendment.
Regarding claim 5, Becht in view of Lai teaches the sorting system of claims 1 and 4 as described above. Claim 5 further adds that the fluorescent separation is performed by a least-squares method, which Becht does not teach.
However, Novo discloses a study of methods to unmix (i.e., separate) fluorescence data from flow cytometry (abstract). Novo teaches a weighted least-squares solution for unmixing (abstract; p. 509, col. 2, par. 2; p. 512, col. 2, par. 2 through p. 513, col. 1).
Regarding claim 18, Becht in view of Lai teaches the sorting system of claims 1 and 17 as described above. Claim 18 further adds that the photodetector is a photodetector array in which a plurality of photoelectric conversion elements are arranged in an array, and the photodetector array is configured to detect the light emitted from the biological particles by spectrally separating fluorescence from the biological particles, which Becht does not teach.
However, Novo teaches that multispectral and hyperspectral flow cytometers may have multispectral detector arrays and are commercially available (abstract; p. 509, col. 2, par. 5).
Regarding claims 5 and 18, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine, in the course of routine experimentation and with a reasonable expectation of success, the methods of Becht in view of Lai, as evidenced by Radisic, with the method of Novo because each reference discloses methods for analyzing fluorescence flow cytometry data from cells. The motivation would have been to use a flow cytometer with a multispectral detector array and to use a technique to unmix the resulting fluorescence signals using a method which avoids negative values in biomarker detection, as taught by Novo (abstract).
Response to Applicant Arguments
At p. 7-8, Applicant submits that the remarks submitted after final on Jan 23 2026 were identified as not being commensurate with the scope of the claims. Applicant submits that the claims now recite “a trained machine learning model” and that the remarks are commensurate with the scope of the current claims, providing the example that Becht and Lai fail to disclose the recited method.
It is respectfully submitted that this is not persuasive. The remarks submitted on Jan 23 2026 were not entered and are not considered herein. If Applicant would like to have those remarks considered, it suggested that they resubmit the remarks with the next amendment. However, it is noted that Becht in view of Lai are considered to teach “a trained machine learning model” as currently recited, as described in the above rejection. Becht clearly teaches training a model with data that reads on those recited in the claims, and Lai clearly teaches deep learning/machine learning models and their training. It is considered that one of ordinary skill in the art would understand that the model of Becht could be modified by substituting a machine learning model for the model used in Hypergate to be trained on and applied to the same types of data therein, which would provide the advantages of machine learning models as set forth by Lai for automatic sorting of cells. Therefore, Applicant’s arguments as submitted, but not entered on Jan 23 2026, are not convincing.
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
A. Instant claims 1-5 and 7-20 are rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1-2, 6-7, 10-11 of U.S. Patent No. 11,137,338 (reference patent) in view of Becht et al. (Bioinformatics, published 21 Jun 2018, 35(2):301-308; previously cited) and Lai et al. (US 2020/0105376; priority to Oct 1 2018; previously cited). Any newly cited portions are necessitated by claim amendment.
The reference patent discloses an information processing apparatus, a particle sorting system, and a non-transitory computer readable medium for particle sorting based on a machine learning algorithm trained to learn a characteristic of the detection data, indicating amounts of luminescence of fluorescence, determined as the process target.
Regarding instant claims 1 and 19-20, reference claims 1, 7, and 10-11 disclose the limitations of instant claims 1 and 19-20 except for applying a nonlinear process to optical data indicative of light emitted from biological particles to generate compressed data, applying the trained learning model to subsequent optical data indicative of light emitted from subsequent biological particles by inputting the subsequent optical data to the trained learning model, and sorting the subsequent biological particles based on result of applying the trained learning model.
However, the prior art to Becht discloses Hypergate, an algorithm which given a cell population of interest identifies a gating strategy optimized for high yield and purity (abstract). Becht teaches the analysis of high-dimensional flow cytometry datasets generated from biological samples labeled with fluorochrome-conjugated antibodies (p. 302, col. 1, par. 4; p. 304, col. 1, par. 2-3), which is considered to read on optical data indicative of light emitted from biological particles as instantly claimed. Such an interpretation is supported by the instant specification as published at least at [0003-0006]. Becht teaches processing the data with dimensionality reduction techniques, including t-SNE, for visualization (p. 302, col. 1, par. 2; p. 303, col. 2, par. 3), which is considered to read on non-linear data compression as instantly claimed. Such an interpretation is supported by instant claim 2 and the instant specification as published at least at [0051; 0061-0063], which provides examples of non-linear processing including dimensional compression, clustering and grouping, including the use of algorithms including t-SNE. Becht does not teach applying the trained learning model to subsequent optical data indicative of light emitted from subsequent biological particles by inputting the subsequent optical data to the trained learning model, and sorting the subsequent biological particles based on result of applying the trained learning model.
However, the prior art to Lai discloses methods and systems for a deep-learning platform for sorting cell populations (abstract). Lai teaches selecting a machine learning model which was trained based on similar phenotype information as indicated in the flow cytometry data, and applying the machine learning model to the flow cytometry data (abstract; claims 1, 8, and 15; [0042-0043]. Lai teaches that the trained machine learning models may analyze the input information and generate output classification information [0060]. Lai teaches instructing a flow cytometer to sort cell populations according to the assigned classifications (claim 19; [0043]).
Regarding instant claim 2-3, the reference patent does not disclose the limitations of instant claim 2-3.
However, Becht teaches performing t-SNE on datasets for input to Hypergate (p. 303, col. 2, par. 3 and 5), which is considered to read on a nonlinear process of dimensional compression, as supported by the instant specification as published at [0074; 0084]. Becht teaches visualizing the tSNE gated cells by tSNE1 and tSNE2 (Fig. 2A), which is considered to read on two dimensions.
Regarding instant claims 4-5, reference claim 2 discloses the limitations of instant claims 4-5.
Regarding instant claims 7, reference claim 1 discloses the limitations of instant claim 7. Using the teaching data associated with the detection data of one or more groups of particles in the machine learning algorithm in the reference application is considered to read on supervised training as recited in instant claim 7.
Regarding instant claims 8-9, reference claim 1 disclose the limitations of instant claims 8-9.
Regarding instant claims 10-11, the reference patent does not disclose the limitations of instant claims 10-11.
However, Becht teaches that Hypergate terminates when no move increases the Fβ score or the last channel added did not contribute more than a user-specified threshold (p. 303, col. 1, par. 5) As Becht teaches that the Fβ score reflects the number of true positive, false negative, and false positive events, it is considered that Becht fairly teaches a rate of correct answers exceeding a threshold as instantly claimed. Becht further teaches training six classifiers on a training set of data (i.e., first portion) and then predicting the class of cells in a left-out, or test, dataset (i.e., second portion) to compute the corresponding F1-scores and accuracies (i.e., the rate of correct answers) (p. 304, col. 1, par. 1).
Regarding instant claim 12, the reference patent does not disclose the limitations of instant claim 12.
However, Becht teaches that Hypergate terminates when no move increases the Fβ score or the last channel added did not contribute more than a user-specified threshold (p. 303, col. 1, par. 5) As Becht teaches that the Fβ score reflects the number of true positive, false negative, and false positive events, it is considered that Becht fairly teaches a rate of correct answers exceeding a threshold as instantly claimed. Becht further teaches training six classifiers on a training set of data (i.e., first portion) and then predicting the class of cells in a left-out, or test, dataset (i.e., second portion) to compute the corresponding F1-scores and accuracies (i.e., the rate of correct answers) (p. 304, col. 1, par. 1). Becht teaches biplots comparing the F1 scores and accuracy in binary classification of the classifiers (Fig. 5), which is considered to teach generating a notification.
Regarding instant claims 13-14, the reference patent does not disclose the limitations of instant claims 13-14.
However, Becht also teaches visualizing the tSNE gated cells by tSNE1 and tSNE2 (Fig. 2), which is considered to read on outputting by mapping the compressed data to two dimensions.
Regarding instant claim 15, the reference patent does not disclose the limitations of instant claim 15.
However, Becht is considered to teach applying a non-linear process to the optical data to generate a compressed dataset (p. 302, col. 1, par. 4; p. 304, col. 1, par. 2-3) and subsequent optical data (p. 304, col. 1, par. 4 through col. 2, par. 1; p. 307, col. 1 through p. 308, col. 1, par. 3; Fig. 6). Becht teaches applying the non-linear process to subsequent data at least in Fig 4B and Fig. 6C-D.
Regarding instant claim 16, reference claim 6 disclose the limitations of instant claims 16.
Regarding instant claims 17-18, reference claim 7 disclose the limitations of instant claims 17-18 except for the photodetector array of claim 18. However, the reference specification discloses that the fluorescence detection unit can be one of several types of arrays (FIG. 2; col 8, lines 18-45). Therefore, the detector of reference claim 15 is considered to read on the photodetector array of claim 18.
Regarding instant claims 1-5 and 7-20, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine, in the course of routine experimentation and with a reasonable expectation of success, the reference patent with Becht and Lai because each reference discloses methods for analyzing fluorescence data from flow cytometry for cell sorting applications. The motivation to perform dimensionality reduction techniques on the optical data would have been to analyze high-dimensional data in order to gate cells, where classical gating techniques are not appropriate, as taught by Becht (p. 302, col. 1, par. 2-3). The motivation to apply the trained machine learning model taught by the reference application to subsequent data would have been to select a particular machine learning model which is optimized, or otherwise advantageous, for the input information, while applying automatic gating tools to avoid user introduced variability, as taught by Lai [0005; 0043].
B. Instant claims 1-5 and 7-20 are provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1, 3, 9-10, 14, and 23-24 of copending Application No. 17/468,368 (reference patent; now U.S. Patent No. 12,287,275) in view of Becht et al. (Bioinformatics, published 21 Jun 2018, 35(2):301-308; previously cited) and Lai et al. (US 2020/0105376; priority to Oct 1 2018; previously cited). Any newly recited portions are necessitated by claim amendment. Instant claim 18 is further rejected by Novo et al. (Cytometry Part A 2013, 83(5):508-520; previously cited). Any newly cited portions are necessitated by claim amendment.
The reference patent discloses an information processing method, an information processing apparatus, and a non-transitory computer readable medium for generating teaching data by associating fluorophore detection data with one or more groups of particles determined to be the process target and generating dictionary data by applying a machine learning algorithm to the teaching data, wherein the dictionary data is used to control a sorting mechanism.
Regarding instant claims 1 and 19-20, reference claims 1, 9, and 23-24 disclose the limitations of instant claims 1 and 19-20 except for applying a nonlinear process to optical data indicative of light emitted from biological particles to generate compressed data, applying the trained learning model to subsequent optical data indicative of light emitted from subsequent biological particles by inputting the subsequent optical data to the trained learning model, and sorting the subsequent biological particles based on result of applying the trained learning model.
However, the prior art to Becht discloses Hypergate, an algorithm which given a cell population of interest identifies a gating strategy optimized for high yield and purity (abstract). Becht teaches the analysis of high-dimensional flow cytometry datasets generated from biological samples labeled with fluorochrome-conjugated antibodies (p. 302, col. 1, par. 4; p. 304, col. 1, par. 2-3), which is considered to read on optical data indicative of light emitted from biological particles as instantly claimed. Such an interpretation is supported by the instant specification as published at least at [0003-0006]. Becht teaches processing the data with dimensionality reduction techniques, including t-SNE, for visualization (p. 302, col. 1, par. 2; p. 303, col. 2, par. 3), which is considered to read on non-linear data compression as instantly claimed. Such an interpretation is supported by instant claim 2 and the instant specification as published at least at [0051; 0061-0063], which provides examples of non-linear processing including dimensional compression, clustering and grouping, including the use of algorithms including t-SNE.
Becht does not teach applying the trained learning model to subsequent optical data indicative of light emitted from subsequent biological particles by inputting the subsequent optical data to the trained learning model, and sorting the subsequent biological particles based on result of applying the trained learning model.
However, the prior art to Lai discloses methods and systems for a deep-learning platform for sorting cell populations (abstract). Lai teaches selecting a machine learning model which was trained based on similar phenotype information as indicated in the flow cytometry data, and applying the machine learning model to the flow cytometry data (abstract; claims 1, 8, and 15; [0042-0043]. Lai teaches that the trained machine learning models may analyze the input information and generate output classification information [0060]. Lai teaches instructing a flow cytometer to sort cell populations according to the assigned classifications (claim 19; [0043]).
Regarding instant claim 2-3, the reference patent does not disclose the limitations of instant claim 2-3.
However, Becht teaches performing t-SNE on datasets for input to Hypergate (p. 303, col. 2, par. 3 and 5), which is considered to read on a nonlinear process of dimensional compression, as supported by the instant specification as published at [0074; 0084]. Becht teaches visualizing the tSNE gated cells by tSNE1 and tSNE2 (Fig. 2A), which is considered to read on two dimensions.
Regarding instant claims 4-5, reference claim 3 discloses the limitations of instant claims 4-5.
Regarding instant claims 7, reference claims 1 and 23-24 discloses the limitations of instant claim 7. Using the teaching data associated with the detection data of one or more groups of particles in the machine learning algorithm in the reference application is considered to read on supervised training as recited in instant claim 7.
Regarding instant claims 8-9, reference claim 1 and 23-24 disclose the limitations of instant claims 8-9.
Regarding instant claims 10-11, the reference patent does not disclose the limitations of instant claims 10-11.
However, Becht teaches that Hypergate terminates when no move increases the Fβ score or the last channel added did not contribute more than a user-specified threshold (p. 303, col. 1, par. 5) As Becht teaches that the Fβ score reflects the number of true positive, false negative, and false positive events, it is considered that Becht fairly teaches a rate of correct answers exceeding a threshold as instantly claimed. Becht further teaches training six classifiers on a training set of data (i.e., first portion) and then predicting the class of cells in a left-out, or test, dataset (i.e., second portion) to compute the corresponding F1-scores and accuracies (i.e., the rate of correct answers) (p. 304, col. 1, par. 1).
Regarding instant claim 12, the reference patent does not disclose the limitations of instant claim 12.
However, Becht teaches that Hypergate terminates when no move increases the Fβ score or the last channel added did not contribute more than a user-specified threshold (p. 303, col. 1, par. 5) As Becht teaches that the Fβ score reflects the number of true positive, false negative, and false positive events, it is considered that Becht fairly teaches a rate of correct answers exceeding a threshold as instantly claimed. Becht further teaches training six classifiers on a training set of data (i.e., first portion) and then predicting the class of cells in a left-out, or test, dataset (i.e., second portion) to compute the corresponding F1-scores and accuracies (i.e., the rate of correct answers) (p. 304, col. 1, par. 1). Becht teaches biplots comparing the F1 scores and accuracy in binary classification of the classifiers (Fig. 5), which is considered to teach generating a notification.
Regarding instant claims 13-14, the reference patent does not disclose the limitations of instant claims 13-14.
However, Becht also teaches visualizing the tSNE gated cells by tSNE1 and tSNE2 (Fig. 2), which is considered to read on outputting by mapping the compressed data to two dimensions.
Regarding instant claim 15, the reference patent does not disclose the limitations of instant claim 15.
However, Becht is considered to teach applying a non-linear process to the optical data to generate a compressed dataset (p. 302, col. 1, par. 4; p. 304, col. 1, par. 2-3) and subsequent optical data (p. 304, col. 1, par. 4 through col. 2, par. 1; p. 307, col. 1 through p. 308, col. 1, par. 3; Fig. 6). Becht teaches applying the non-linear process to subsequent data at least in Fig 4B and Fig. 6C-D.
Regarding instant claim 16, reference claims 10 and 14 disclose the limitations of instant claims 16.
Regarding instant claims 17-18, the reference patent does not disclose the limitations of instant claims 17-18.
However, Becht teaches the use of flow cytometry to collect fluorescent data from single cells, as described above (abstract; p. 302, col. 1, par. 4; p. 304, col. 1, par. 2-4). It is considered that a light source to irradiate laser light to the biological particles and a photodetector configured to obtain the light emitted from the biological particles are inherent features of a flow cytometer which analyzes fluorescently labeled cells, as evidenced by Radisic (p .5, col. 2, par. 3 through p. 7, col. 1, par. 1).
However, Novo teaches that multispectral and hyperspectral flow cytometers may have multispectral detector arrays and are commercially available (abstract; p. 509, col. 2, par. 5).
Regarding instant claims 1-5 and 7-20, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine, in the course of routine experimentation and with a reasonable expectation of success, the reference patent with Becht, Lai, and Novo because each reference discloses methods for analyzing fluorescence data from flow cytometry for cell sorting applications. The motivation to perform dimensionality reduction techniques on the optical data would have been to analyze high-dimensional data in order to gate cells, where classical gating techniques are not appropriate, as taught by Becht (p. 302, col. 1, par. 2-3). The motivation to apply the trained machine learning model taught by the reference application to subsequent data would have been to select a particular machine learning model which is optimized, or otherwise advantageous, for the input information, while applying automatic gating tools to avoid user introduced variability, as taught by Lai [0005; 0043]. The motivation to use the detector arrays of Novo would have been that they are commercially available, as taught by Novo (abstract; p. 509, col. 2, par. 5).
C. Instant claims 1-2, 7-9, 12-13, 15-20 are provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1-2, 8-9, 11, 13-16, 17-18, 24-25, and 30-32 of copending Application No. 17/613,009 (reference application). Instant claims 3, 10-11, 14 are unpatentable over the reference application, as applied to claim 1, and in further view of Becht et al. (Bioinformatics, published 21 Jun 2018, 35(2):301-308; previously cited). Instant claim 4-5 are unpatentable over the reference application, as applied to claims 1 and 4, and in further view of Novo et al. (Cytometry Part A 2013, 83(5):508-520; previously cited). The rejection is newly stated and is necessitated by claim amendment.
The reference application discloses an information processing system, an information processing method, and a non-transitory computer-readable storage medium for applying a data compression process to data indicating light emitted from biological particles to output groups of biological particles, use the data corresponding to the groups in training a machine learning model, and applying the trained machine learning model to subsequent data to provide information about sorting subsequent biological particles.
Regarding instant claims 1-2 and 19-20, reference claims 1-2, 11, 13-15, 18, 24, and 30-32 disclose the limitations of instant claims 1 and 19-20, as supported by the instant specification which provides an example of non-linear processing of clustering at [0051].
Regarding instant claim 3, the reference application does not disclose the limitation of instant claim 3.
However, the prior art to Becht discloses Hypergate, an algorithm which given a cell population of interest identifies a gating strategy optimized for high yield and purity (abstract). Becht teaches processing the data with dimensionality reduction techniques, including t-SNE, for visualization (p. 302, col. 1, par. 2; p. 303, col. 2, par. 3), which is considered to read on non-linear data compression as instantly claimed. Such an interpretation is supported by instant claim 2 and the instant specification as published at least at [0051; 0061-0063], which provides examples of non-linear processing including dimensional compression, clustering and grouping, including the use of algorithms including t-SNE. Becht teaches visualizing the tSNE gated cells by tSNE1 and tSNE2 (Fig. 2A), which is considered to read on two dimensions.
Regarding instant claim 4-5, reference claims 8, 17, and 25 disclose the limitations of instant claim 4-5 except for performing fluorescent separation in claim 4 and the least-squares method of claim 5.
However, the prior art to Novo discloses a study of methods to unmix (i.e., separate) fluorescence data from flow cytometry (abstract). Novo teaches a weighted least-squares solution for unmixing (abstract; p. 509, col. 2, par. 2; p. 512, col. 2, par. 2 through p. 513, col. 1).
Regarding claims 4-5, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine, in the course of routine experimentation and with a reasonable expectation of success, the reference application with the method of Novo because each reference discloses methods for analyzing fluorescence flow cytometry data from cells. The motivation would have been to use flow cytometer with a multispectral detector array and to use a technique to unmix the resulting fluorescence signals using a method which avoids negative values in biomarker detection, as taught by Novo (abstract).
Regarding instant claim 7, reference claims 1-2, 11, 13-15, 18, 24, and 30-32 disclose the limitations of instant claim 7. Using the data corresponding to the one or more groups of the biological particles in training the statistical model in the reference application is considered to read on supervised training as recited in instant claim 7.
Regarding instant claim 8-9, reference claims 1, 13-15, 24, and 30 disclose the limitations of instant claims 8-9.
Regarding instant claim 10-11, the reference application does not disclose the limitations of claims 10-11.
However, Becht teaches that Hypergate terminates when no move increases the Fβ score or the last channel added did not contribute more than a user-specified threshold (p. 303, col. 1, par. 5) As Becht teaches that the Fβ score reflects the number of true positive, false negative, and false positive events, it is considered that Becht fairly teaches a rate of correct answers exceeding a threshold as instantly claimed. Becht further teaches training six classifiers on a training set of data (i.e., first portion) and then predicting the class of cells in a left-out, or test, dataset (i.e., second portion) to compute the corresponding F1-scores and accuracies (i.e., the rate of correct answers) (p. 304, col. 1, par. 1).
Regarding instant claim 12, reference claims 1, 13-15, 24, and 30 disclose the limitations of claim 12.
Regarding instant claim 13, reference claims 1, 13-15, 24, and 30 disclose the limitations of claim 13.
Regarding instant claim 14, the reference application does not disclose the limitations of claim 14.
However, Becht teaches visualizing the tSNE gated cells by tSNE1 and tSNE2 (Fig. 2), which is considered to read on outputting by mapping the compressed data to two dimensions.
Regarding instant claims 3, 10-11, and 14, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine, in the course of routine experimentation and with a reasonable expectation of success, the reference patent and Becht because each reference discloses methods for analyzing fluorescence data from flow cytometry for cell sorting applications. The motivation to perform dimensionality reduction techniques on the optical data would have been to analyze high-dimensional data in order to gate cells, where classical gating techniques are not appropriate, as taught by Becht (p. 302, col. 1, par. 2-3).
Regarding instant claim 15, the reference application does not disclose the limitations of claim 15.
However, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify, in the course of routine experimentation and with a reasonable expectation of success, the features of the reference application to apply a non-linear process to the subsequent optical data to generate a second set of compressed data because the reference application already teaches each of the recited processes and datasets. Therefore, it would have been obvious to one of ordinary skill in the art to apply those same processes to the subsequent optical data to predictably result in compressed data of the subsequent optical data.
Regarding instant claim 16, reference claims 9 and 26 disclose the limitations of claim 16.
Regarding instant claim 17-18, reference claim 15 disclose the limitations of claims 17-18.
This is a provisional nonstatutory double patenting rejection.
D. Instant claims 1-3 and 7-20 are provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1-4, 9-11, and 12-15 of copending Application No. 18/570,354 (reference application). Instant claims 4-5 are unpatentable over the reference application, as applied to claim 1, and in further view of Novo et al. (Cytometry Part A 2013, 83(5):508-520; previously cited). Any newly recited portions are necessitated by claim amendment.
The reference patent discloses a particle analysis system, an information processing apparatus, and a sorting apparatus for a dimensionally compressing multidimensional data regarding light output and a machine learning model for determining a relationship between the dimensionally compressed data and the multidimensional data.
Regarding instant claims 1 and 19-20, reference claims 1-2, 9-10, 12, and 14-15 disclose the limitations of claims 1 and 19-20.
Regarding instant claim 2, reference claims 1-2 disclose the limitations of instant claim 2.
Regarding instant claim 3, reference claims 3-4 disclose the limitations of instant claim 3.
Regarding instant claims 4-5, the reference application does not disclose the limitations of instant claims 4-5.
However, the prior art to Novo discloses a study of methods to unmix (i.e., separate) fluorescence data from flow cytometry (abstract). Novo teaches a weighted least-squares solution for unmixing (abstract; p. 509, col. 2, par. 2; p. 512, col. 2, par. 2 through p. 513, col. 1).
Regarding claims 4-5, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine, in the course of routine experimentation and with a reasonable expectation of success, the reference application with the method of Novo because each reference discloses methods for analyzing fluorescence flow cytometry data from cells. The motivation would have been to use flow cytometer with a multispectral detector array and to use a technique to unmix the resulting fluorescence signals using a method which avoids negative values in biomarker detection, as taught by Novo (abstract).
Regarding instant claims 7, reference claims 1, 10, and 14-15 disclose the limitations of instant claim 7. Using the data corresponding to the relationship between the dimensionally compressed data as a teacher in the reference application is considered to read on supervised training as recited in instant claim 7.
Regarding instant claim 8-9, reference claims 10-11 disclose the limitations of instant claims 8-9.
Regarding instant claim 10-11, the reference application does not disclose the limitations of claims 10-11.
However, Becht teaches that Hypergate terminates when no move increases the Fβ score or the last channel added did not contribute more than a user-specified threshold (p. 303, col. 1, par. 5) As Becht teaches that the Fβ score reflects the number of true positive, false negative, and false positive events, it is considered that Becht fairly teaches a rate of correct answers exceeding a threshold as instantly claimed. Becht further teaches training six classifiers on a training set of data (i.e., first portion) and then predicting the class of cells in a left-out, or test, dataset (i.e., second portion) to compute the corresponding F1-scores and accuracies (i.e., the rate of correct answers) (p. 304, col. 1, par. 1).
Regarding instant claims 10-11, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine, in the course of routine experimentation and with a reasonable expectation of success, the reference patent and Becht because each reference discloses methods for analyzing fluorescence data from flow cytometry for cell sorting applications. The motivation to perform dimensionality reduction techniques on the optical data would have been to analyze high-dimensional data in order to gate cells, where classical gating techniques are not appropriate, as taught by Becht (p. 302, col. 1, par. 2-3).
Regarding instant claim 12, reference claim 13 discloses the limitations of claim 12.
Regarding instant claim 13-14, reference claims 1, 3, and 13 disclose the limitations of claims 13-14.
Regarding instant claim 15, the reference application does not disclose the limitations of claim 15.
However, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify, in the course of routine experimentation and with a reasonable expectation of success, the features of the reference application to apply a non-linear process to the subsequent optical data to generate a second set of compressed data because the reference application already teaches each of the recited processes and datasets. Therefore, it would have been obvious to one of ordinary skill in the art to apply those same processes to the subsequent optical data to predictably result in compressed data of the subsequent optical data.
Regarding instant claim 16, reference claims 1, 9-11, and 14-15 disclose the limitations of claim 16 but do not explicitly disclose a cell. However, the reference specification discloses an example of biological particles being cells [0017]. Therefore, the biologically derived particle recited in the claims of the reference application read on a cells as instantly claimed.
Regarding instant claim 17-18, reference claim 15 disclose the limitations of claim 17-18 except for the photodetector array of claim 18. However, the reference specification discloses an example of a light receiving element array of a plurality of light receiving elements [0025]. Therefore, the light receiving unit of reference claim 15 reads on the photodetector array of claim 18.
This is a provisional nonstatutory double patenting rejection.
E. Instant claims 1-5 and 7-20 are provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1, 3, 9-10, 14-15, 20, and 22-24 of copending Application No. 19/004754 in view of Becht et al. (Bioinformatics, published 21 Jun 2018, 35(2):301-308; previously cited) and Lai et al. (US 2020/0105376; priority to Oct 1 2018; previously cited). Instant claim 18 is further rejected by Novo et al. (Cytometry Part A 2013, 83(5):508-520; previously cited). Any newly cited portions are necessitated by claim amendment.
The reference application discloses an information processing method, an information processing apparatus, and a non-transitory computer readable medium for generating teaching data by associating fluorophore detection data with one or more groups of particles determined to be the process target and generating dictionary data by applying a machine learning algorithm to the teaching data, wherein the dictionary data is used to control a sorting mechanism.
Regarding instant claims 1 and 19-20, reference claims 21, 25, 29-30, 34, 37-38, and 42 disclose the limitations of instant claims 1 and 19-20 except for applying a nonlinear process to optical data indicative of light emitted from biological particles to generate compressed data, applying the trained learning model to subsequent optical data indicative of light emitted from subsequent biological particles by inputting the subsequent optical data to the trained learning model, and sorting the subsequent biological particles based on result of applying the trained learning model.
However, the prior art to Becht discloses Hypergate, an algorithm which given a cell population of interest identifies a gating strategy optimized for high yield and purity (abstract). Becht teaches the analysis of high-dimensional flow cytometry datasets generated from biological samples labeled with fluorochrome-conjugated antibodies (p. 302, col. 1, par. 4; p. 304, col. 1, par. 2-3), which is considered to read on optical data indicative of light emitted from biological particles as instantly claimed. Such an interpretation is supported by the instant specification as published at least at [0003-0006]. Becht teaches processing the data with dimensionality reduction techniques, including t-SNE, for visualization (p. 302, col. 1, par. 2; p. 303, col. 2, par. 3), which is considered to read on non-linear data compression as instantly claimed. Such an interpretation is supported by instant claim 2 and the instant specification as published at least at [0051; 0061-0063], which provides examples of non-linear processing including dimensional compression, clustering and grouping, including the use of algorithms including t-SNE.
Becht does not teach applying the trained learning model to subsequent optical data indicative of light emitted from subsequent biological particles by inputting the subsequent optical data to the trained learning model, and sorting the subsequent biological particles based on result of applying the trained learning model.
However, the prior art to Lai discloses methods and systems for a deep-learning platform for sorting cell populations (abstract). Lai teaches selecting a machine learning model which was trained based on similar phenotype information as indicated in the flow cytometry data, and applying the machine learning model to the flow cytometry data (abstract; claims 1, 8, and 15; [0042-0043]. Lai teaches that the trained machine learning models may analyze the input information and generate output classification information [0060]. Lai teaches instructing a flow cytometer to sort cell populations according to the assigned classifications (claim 19; [0043]).
Regarding instant claim 2-3, the reference patent does not disclose the limitations of instant claim 2-3.
However, Becht teaches performing t-SNE on datasets for input to Hypergate (p. 303, col. 2, par. 3 and 5), which is considered to read on a nonlinear process of dimensional compression, as supported by the instant specification as published at [0074; 0084]. Becht teaches visualizing the tSNE gated cells by tSNE1 and tSNE2 (Fig. 2A), which is considered to read on two dimensions.
Regarding instant claims 4-5, reference claim 32 discloses the limitations of instant claims 4-5.
Regarding instant claims 7, reference claims 21 and 38 disclose the limitations of instant claim 7. Using the teaching data associated with the detection data of one or more groups of particles in the machine learning algorithm in the reference application is considered to read on supervised training as recited in instant claim 7.
Regarding instant claims 8-9, reference claim 21, 29, and 37-38 disclose the limitations of instant claims 8-9.
Regarding instant claims 10-11, the reference patent does not disclose the limitations of instant claims 10-11.
However, Becht teaches that Hypergate terminates when no move increases the Fβ score or the last channel added did not contribute more than a user-specified threshold (p. 303, col. 1, par. 5) As Becht teaches that the Fβ score reflects the number of true positive, false negative, and false positive events, it is considered that Becht fairly teaches a rate of correct answers exceeding a threshold as instantly claimed. Becht further teaches training six classifiers on a training set of data (i.e., first portion) and then predicting the class of cells in a left-out, or test, dataset (i.e., second portion) to compute the corresponding F1-scores and accuracies (i.e., the rate of correct answers) (p. 304, col. 1, par. 1).
Regarding instant claim 12, the reference patent does not disclose the limitations of instant claim 12.
However, Becht teaches that Hypergate terminates when no move increases the Fβ score or the last channel added did not contribute more than a user-specified threshold (p. 303, col. 1, par. 5) As Becht teaches that the Fβ score reflects the number of true positive, false negative, and false positive events, it is considered that Becht fairly teaches a rate of correct answers exceeding a threshold as instantly claimed. Becht further teaches training six classifiers on a training set of data (i.e., first portion) and then predicting the class of cells in a left-out, or test, dataset (i.e., second portion) to compute the corresponding F1-scores and accuracies (i.e., the rate of correct answers) (p. 304, col. 1, par. 1). Becht teaches biplots comparing the F1 scores and accuracy in binary classification of the classifiers (Fig. 5), which is considered to teach generating a notification.
Regarding instant claims 13-14, the reference patent does not disclose the limitations of instant claims 13-14.
However, Becht also teaches visualizing the tSNE gated cells by tSNE1 and tSNE2 (Fig. 2), which is considered to read on outputting by mapping the compressed data to two dimensions.
Regarding instant claim 15, the reference patent does not disclose the limitations of instant claim 15.
However, Becht is considered to teach applying a non-linear process to the optical data to generate a compressed dataset (p. 302, col. 1, par. 4; p. 304, col. 1, par. 2-3) and subsequent optical data (p. 304, col. 1, par. 4 through col. 2, par. 1; p. 307, col. 1 through p. 308, col. 1, par. 3; Fig. 6). Becht teaches applying the non-linear process to subsequent data at least in Fig 4B and Fig. 6C-D.
Regarding instant claim 16, reference claim 30 discloses the limitations of instant claims 16.
Regarding instant claims 17-18, the reference patent does not disclose the limitations of instant claims 17-18.
However, Becht teaches the use of flow cytometry to collect fluorescent data from single cells, as described above (abstract; p. 302, col. 1, par. 4; p. 304, col. 1, par. 2-4). It is considered that a light source to irradiate laser light to the biological particles and a photodetector configured to obtain the light emitted from the biological particles are inherent features of a flow cytometer which analyzes fluorescently labeled cells, as evidenced by Radisic (p .5, col. 2, par. 3 through p. 7, col. 1, par. 1).
However, Novo teaches that multispectral and hyperspectral flow cytometers may have multispectral detector arrays and are commercially available (abstract; p. 509, col. 2, par. 5).
Regarding instant claims 1-5 and 7-20, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine, in the course of routine experimentation and with a reasonable expectation of success, the reference patent with Becht, Lai, and Novo because each reference discloses methods for analyzing fluorescence data from flow cytometry for cell sorting applications. The motivation to perform dimensionality reduction techniques on the optical data would have been to analyze high-dimensional data in order to gate cells, where classical gating techniques are not appropriate, as taught by Becht (p. 302, col. 1, par. 2-3). The motivation to apply the trained machine learning model taught by the reference application to subsequent data would have been to select a particular machine learning model which is optimized, or otherwise advantageous, for the input information, while applying automatic gating tools to avoid user introduced variability, as taught by Lai [0005; 0043]. The motivation to use the detector arrays of Novo would have been that they are commercially available, as taught by Novo (abstract; p. 509, col. 2, par. 5).
This is a provisional nonstatutory double patenting rejection.
Response to Applicant Arguments
At p. 8-9, Applicant requests reconsideration in view of the distinctions over Becht and Lai.
It is respectfully submitted that this is not persuasive. The remarks submitted on Jan 23 2026 were not entered and are not considered herein. If Applicant would like to have those remarks considered, it suggested that they resubmit the remarks with the next amendment. However, it is noted that Becht in view of Lai are considered to teach “a trained machine learning model” as currently recited, as described in the above rejection. Becht clearly teaches training a model with data that reads on those recited in the claims, and Lai clearly teaches deep learning/machine learning models and their training. It is considered that one of ordinary skill in the art would understand that the model of Becht could be modified by substituting a machine learning model for the model used in Hypergate to be trained on and applied to the same types of data therein, which would provide the advantages of machine learning models as set forth by Lai for automatic sorting of cells.
Conclusion
No claims are allowed.
Inquiries
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JANNA NICOLE SCHULTZHAUS whose telephone number is (571)272-0812. The examiner can normally be reached on Monday - Friday 8-4.
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, Olivia Wise can be reached on (571)272-2249. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see https://ppair-my.uspto.gov/pair/PrivatePair. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/JANNA NICOLE SCHULTZHAUS/Examiner, Art Unit 1685