DETAILED ACTION
This office action is in response to amendments filed on 08/05/2026.
Claims 1, 13, and 14 have been amended. Claims 1-20 are pending.
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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 09/08/2026 has been entered.
Response to Arguments
Rejections Under 35 USC § 112(b):
In light of applicant’s amendments to the claims (pg. 2-6), the rejections under 35 USC § 112(b) have been withdrawn.
Prior Art Rejections:
Applicant’s arguments regarding the prior art rejections (pg. 11-12) have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
The prior art rejections have been updated to include the amended limitations and to clarify the reasoning given for the limitations that were not amended.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-2, 8-9, 12-14, 17-18, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over
Dribus, U.S. Patent Application Publication US-20210073642-A1 (filed 07/14/2020) in view of
Middlemas et al. (hereinafter Middlemas) “Hyperuniformity Order Metric of Barlow Packings” (published 12/06/2018).
Regarding Claim 1,
Dribus teaches A method comprising:
generating a sparse topology for a feedforward neural network, where connectivity is based on a substantially hyperuniform topology, (Examiner notes that, according to specification paragraph 0034 of the instant application, “as used herein, ‘hyperuniform’ refers to the suppression of long-range density fluctuations, ‘nearly hyperuniform’ refers to the near suppression of long- range density fluctuations, and ‘substantially hyperuniform’ refers to both ‘hyperuniform’ and ‘nearly hyperuniform’ density fluctuations.” Dribus, 0041-0042: “The present invention provides novel architectures for artificial neural networks (ANNs), which simultaneously exhibit extreme sparseness, local structure similar to that of convolutional neural networks (CNNs), and connectivity similar to that of dense networks… The present invention also provides network architecture construction algorithms (NACAs) to construct such architectures.” 0201: “A specific example NACA is now provided for constructing families of 3-dimensional G-networks called Diamond4-nets, or D4-nets. Interesting subnetworks called Tetra-octa-nets, or TO-nets, are obtained by pruning away one of two topologically-connected components defined by the D4 kernel architecture…” 0224: “Networks constructed via this specific NACA are called square-layered w-weighted Diamond4-nets or Tetra-octa-nets, due to their layer and kernel shapes. The latter name is chosen because the pruned kernel architecture selects the nodes and most of the edges of the tetrahedral-octahedral honeycomb, a quasiregular tessellation of
R
3
.” A network architecture construction algorithm constructs a sparse topology for a feedforward neural network, where connectivity is based on the tetrahedral-octahedral honeycomb. The tetrahedral-octahedral honeycomb is a substantially hyperuniform topology, as shown by the teachings below:
Per Conway, the tetrahedral-octahedral honeycomb “consists of repeat tiling units made up of two regular tetrahedra and one regular octahedron, which is the Delaunay tessellation (17) of the face-centered cubic (fcc) lattice” (Conway et al., “New family of tilings of three-dimensional Euclidean space by tetrahedra and octahedra”, pg. 1, col. 2, para. 1).
Per Middlemas, the fcc lattice is hyperuniform: “One interesting example of a class of hyperuniform systems are… the Barlow packings… The most commonly known examples of Barlow packings are the fcc lattice and the hcp crystal” (Middlemas et al., “Hyperuniformity Order Metric of Barlow Packings”, pg. 1-2, section I).
training the feedforward neural network with the sparse topology using a set of training data; and (0004: “An ANN [artificial neural network]… is trained to minimize a loss function, which measures the error of actual outputs for training samples whose correct outputs are known.”)
performing a processing task using the trained feedforward neural network. (0003-0004: “Artificial neural networks (ANNs) are central to modern machine learning and have revolutionized such fields as image classification, speech recognition, vehicle navigation, and game strategy… It processes data by converting real-valued inputs to real-valued outputs via activation functions.”)
Dribus does not appear to explicitly disclose wherein the substantially hyperuniform topology maps to a bond connectivity of a nearly hyperuniform material;
However, Middlemas teaches wherein the substantially hyperuniform topology maps to a bond connectivity of a nearly hyperuniform material; (Pg. 1, section I: “Hyperuniformity arises in a variety of systems across multiple disciplines, including in the early density fluctuations of the universe [4, 5, 13, 14], classical disordered ground states [12, 15–24], maximally random jammed packings [25–30], models of plasmas [7, 31–33], patterning of avian photoreceptor cells [34], quasicrystals [8, 35, 36], and the spatial distribution of prime numbers [37, 38].” The substantially hyperuniform topology maps to quasicrystals (i.e. bond connectivity of a nearly hyperuniform material).)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Dribus and Middlemas. Dribus teaches a network architecture construction algorithm for constructing a sparse neural network topology, where connectivity is based on the hyperuniform tetrahedral-octahedral honeycomb. Middlemas teaches that hyperuniformity arises in a variety of systems, including the particle distribution and bond connectivity of quasicrystals. One of ordinary skill would have motivation to combine Dribus and Middlemas because Dribus draws on physical/materials science analogies (e.g. “a process resembling crystallization of a supersaturated solution,” Dribus, 0369), and Middlemas provides a real-world structural analogue that grounds and characterizes Dribus’s hyperuniform neural network topology as corresponding to a known, physically realized system.
Regarding Claim 2, Dribus and Middlemas teach The method of claim 1, as shown above.
Dribus also teaches wherein the processing task comprises one of classification, regression, and correlations at different scales. (0003: “Artificial neural networks (ANNs) are central to modern machine learning and have revolutionized such fields as image classification, speech recognition, vehicle navigation, and game strategy.”)
Regarding Claim 8, Dribus and Middlemas teach The method of claim 1, as shown above.
Dribus also teaches wherein the sparse topology comprises a network of ring configurations of nodes in two dimensions or surface configurations of nodes in greater than two dimensions, the ring configurations and surface configurations having variable connectivity between the corresponding nodes, and wherein a copy of information flows from a first node in a first direction through a first portion of a given ring configuration or a given surface configuration of the trained feedforward neural network and in a second direction through a second portion of the given ring or the given surface to meet at a second node of the given ring or the given surface. (0022: “FIG. 4A shows six ‘toy’ networks, along with triples of numbers encoding their density, locality, and input-output connectivity, in accordance with several different embodiments of the present invention.” The figure is copied below for reference. Network 3, for example, is a sparse topology comprising a network of ring configurations in two dimensions. An example of such a ring configuration is highlighted in green. The nodes of the highlighted ring configuration have variable connectivity (e.g., the bottom node has a degree of 4, while the top node has a degree of 6). A copy of information flows from the leftmost node through both the top and bottom portions of the ring configuration to meet at the rightmost node.)
PNG
media_image1.png
432
680
media_image1.png
Greyscale
Regarding Claim 9, Dribus and Middlemas teach The method of claim 1, as shown above.
Dribus also teaches wherein at least one ring configuration of nodes of the trained feedforward neural network has a different number of nodes than another configuration ring of nodes of the trained feedforward neural network. (0022: “FIG. 4A shows six ‘toy’ networks, along with triples of numbers encoding their density, locality, and input-output connectivity, in accordance with several different embodiments of the present invention.” The figure is copied below for reference. Network 3, for example, is a sparse topology comprising a network of ring configurations. Two examples of such ring configurations are highlighted. The ring configuration highlighted in blue has 6 nodes, while the ring configuration highlighted in green has 4 nodes.)
PNG
media_image2.png
432
680
media_image2.png
Greyscale
Regarding Claim 12, Dribus and Middlemas teach The method of claim 1, as shown above.
Dribus also teaches wherein, in the operation of generating the sparse topology for the feedforward neural network, the sparse topology comprises a network of surface configurations of nodes in greater than two dimensions. (0201: “A specific example NACA is now provided for constructing families of 3-dimensional G-networks called Diamond4-nets, or D4-nets. Interesting subnetworks called Tetra-octa-nets, or TO-nets, are obtained by pruning away one of two topologically-connected components defined by the D4 kernel architecture…” 0224: “Networks constructed via this specific NACA are called square-layered w-weighted Diamond4-nets or Tetra-octa-nets, due to their layer and kernel shapes. The latter name is chosen because the pruned kernel architecture selects the nodes and most of the edges of the tetrahedral-octahedral honeycomb, a quasiregular tessellation of
R
3
.” The sparse topology for TO-nets comprises a 3-dimensional honeycomb network of tetrahedra and octahedra (i.e. surface configurations).)
Claim 13 is a product claim containing substantially the same elements as method claim 1. Dribus and Middlemas teach the elements of claim 1, as shown above.
Dribus also teaches A non-transitory computer readable medium comprising computer executable instructions which when executed by a computer cause the computer to perform the method (Examiner notes this limitation is interpreted as implementation of the disclosed method in a computing environment. 0226: “FIG. 7 shows a toy case involving D4-nets of size 20×20×10, computed via a Python script…” Computation using python necessitates implementation in a computing environment.)
Claims 14, 17-18, and 20 are system claims containing substantially the same elements as method claims 1, 8-9, and 12, respectively. Dribus and Middlemas teach the elements of claims 1, 8-9, and 12, as shown above.
Dribus also teaches An apparatus comprising: a memory; and at least one processor, coupled to said memory, and operative to perform operations comprising: (Examiner notes this limitation is interpreted as implementation of the disclosed method in a computing environment. 0226: “FIG. 7 shows a toy case involving D4-nets of size 20×20×10, computed via a Python script…” Computation using python necessitates implementation in a computing environment.)
Claim 3 is rejected under 35 U.S.C. 103 as being unpatentable over Dribus in view of Middlemas, and further in view of
Kathiravelu et al. (hereinafter Kathiravelu), “A DICOM Framework for Machine Learning Pipelines against Real-Time Radiology Images”.
Regarding Claim 3, Dribus and Middlemas teach The method of claim 1, as shown above.
Dribus and Middlemas do not appear to explicitly disclose the remaining features of claim 3.
However, Kathiravelu teaches wherein the processing task comprises connecting to a medical imaging device via a network module, obtaining a medical image, processing the medical image using the trained feedforward neural network, and treating a patient based on results of the processing of the medical image. (Pg. 1-2, section 1: “Radiology departments consist of several clinical systems such as PACS [17] and Vendor-Neutral Archives (VNAs) [26] that receive images real-time from various scanners… ML pipelines have been proposed to reliably prognosis and predict cancer… This paper presents Niffler, an ML framework that retrieves images from the PACS using DICOM network listeners, and extracts and processes metadata from the acquired images at the research clusters. It then executes ML pipelines and real-time analytics pipelines on radiology images and their textual metadata.” Scanned images (i.e. medical images) are obtained by connecting to scanners (i.e. medical imaging devices) through the PACS system and DICOM protocol (i.e. network module) and processed by ML pipelines (i.e. trained feedforward neural network) to perform prognosis (i.e. treat a patient).)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Dribus, Middlemas, and Kathiravelu. Dribus teaches a network architecture construction algorithm for constructing a sparse neural network topology, where connectivity is based on the hyperuniform tetrahedral-octahedral honeycomb. Middlemas teaches that hyperuniformity arises in a variety of systems, including the particle distribution and bond connectivity of quasicrystals. Kathiravelu teaches a system for real-time execution of machine learning models on medical imaging data. One of ordinary skill would have motivation to combine Dribus, Middlemas, and Kathiravelu because Dribus provides sparse and efficient neural networks for “such fields as image classification” (Dribus, 0003), and Kathiravelu’s Niffler framework enables efficient real-time execution of ML models, including neural networks, on medical imaging data: “we presented Niffler, a framework that supports the seamless transfer of data from the PACS to the research clusters and enables efficient execution of ML pipelines on the images, reports, and the extracted textual metadata. Niffler facilitates the execution of ML models with a minimal tuning of infrastructure” (Kathiravelu, Pg. 1, section 1).
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Dribus in view of Middlemas, and further in view of
Berdouz Qrichi Aniba (hereinafter Berdouz), U.S. Patent Application Publication US 20240056471 A1.
Regarding Claim 4, Dribus and Middlemas teach The method of claim 1, as shown above.
Dribus and Middlemas do not appear to explicitly disclose the remaining features of claim 4.
However, Berdouz teaches wherein the processing task comprises obtaining financial information via a network module, processing the financial information using the trained feedforward neural network, and detecting and mitigating financial fraud based on results of the processing of the financial information. ([0002]: “The invention belongs to the field of automated fraud prediction and detection, and finds various applications in systems wherein frauds can occur, for example financial systems…” [0008]: “To this end, the invention proposes, according to one aspect, a method of automated detection of the risk of fraud in a monitored system, from data streams generated by said monitored system…” [0036]: “The monitored system generates data streams 41 to 4k, which are recorded in one or a plurality of electronic memory units 6, for example in the form of a database or any other data storage structure. For example, the unit 6 consists of a plurality of interconnected memories.” [0047]: “The data streams 4l, stored over associated periods of time Tl are accessible to an automated fraud detection system 10 according to the invention.” [0095]: “The initialization step 50 is followed by substantially concomitant steps 52 and 54, which are a step 52 of application of a first parameterized process for the estimation of the risk of fraud, thereafter called ‘process 1’…” [0118]: “For example, in one embodiment, the ‘process 1’ implements a Convolutional Neural Network (CNN)…” [0127]: “In the case where the operator identified by Id_M is presumed to be a fraudster, the step 64 is followed by a step 68 for raising an alert, e.g. by sending a notification, or any other means, to the managing authorities of the monitored system.” Data streams generated by the monitored system (i.e. financial information) are obtained from the memory units (i.e. via a network module) and processed by ‘process 1’ implementing a CNN (i.e. trained feedforward neural network) to detect fraud and raise an alert (i.e. detect and mitigate the fraud).)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Dribus, Middlemas, and Berdouz. Dribus teaches a network architecture construction algorithm for constructing a sparse neural network topology, where connectivity is based on the hyperuniform tetrahedral-octahedral honeycomb. Middlemas teaches that hyperuniformity arises in a variety of systems, including the particle distribution and bond connectivity of quasicrystals. Berdouz teaches financial fraud detection and mitigation using ML models including neural networks. One of ordinary skill would have motivation to combine Dribus, Middlemas, and Berdouz because Dribus provides sparse and efficient neural networks, and Berdouz’s framework uses neural networks to perform financial fraud detection more reliably than other methods: “However, it has been found that most prior art fraud detection methods provide too high false-positive rates, which have a negative impact on the system in terms of performance and customer experience and do not guarantee that all false-negatives have been detected. One of the goals of the invention is to remedy such drawback by proposing a more reliable method” (Berdouz, 0006-0007).
Claim 5-7, 11, and 15-16 is rejected under 35 U.S.C. 103 as being unpatentable over Dribus in view of Middlemas, and further in view of
Aggarwal, “An Introduction to Neural Networks” (published 08/26/2018).
Regarding Claim 5, Dribus and Middlemas teach The method of claim 1, as shown above.
Dribus and Middlemas do not appear to explicitly disclose the remaining features of claim 5.
However, Aggarwal teaches wherein the performing of the processing task comprises:
feeding input elements of an input vector into a first set of hidden nodes of the trained feedforward neural network; (Pg. 14, figure 1.9 shows an example of a neural network with a softmax layer for categorical classification. The figure is copied below for reference. Input elements
x
1
,
…
,
x
5
are fed into the nodes of the hidden layer.)
PNG
media_image3.png
180
368
media_image3.png
Greyscale
applying a linear transformation by each of the first set of hidden nodes to a corresponding input element or corresponding input vector; (Pg. 5, section 1.2.1: “The input layer contains
d
nodes that transmit the
d
features
X
=
[
x
1
.
.
.
x
d
]
with edges of weight
W
=
[
w
1
.
.
.
w
d
]
… The linear function
W
-
·
X
-
=
∑
i
=
1
d
w
i
x
i
is computed…”)
applying an activation function to a result of the linear transformation; and (See figure 1.9 copied above. The softmax activation function is applied to the result of the hidden layer’s linear transformation.)
obtaining an output value. (See figure 1.9 copied above. The output values are obtained from the result of the softmax activation function.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Dribus, Middlemas, and Aggarwal. Dribus teaches a network architecture construction algorithm for constructing a sparse neural network topology, where connectivity is based on the hyperuniform tetrahedral-octahedral honeycomb. Middlemas teaches that hyperuniformity arises in a variety of systems, including the particle distribution and bond connectivity of quasicrystals. Aggarwal teaches an introduction to the basic architecture of neural networks. One of ordinary skill would have motivation to combine Dribus, Middlemas, and Aggarwal because Dribus is concerned with constructing artificial neural networks, and Aggarwal teaches standard practices for their implementation.
Regarding Claim 6, Dribus and Middlemas teach The method of claim 1, as shown above.
Dribus and Middlemas do not appear to explicitly disclose the remaining features of claim 6.
However, Aggarwal teaches wherein a layer of the sparse topology comprises a set of nodes connected in one direction and an output value of a node of a given layer is an input value of one or more subsequent nodes of the given layer or another layer. (Pg. 14, figure 1.9 shows an example of a neural network with a softmax layer for categorical classification. The figure is copied below for reference. The input, hidden, and softmax layers are labeled, and the arrows between their nodes represent connections in one direction. The output values of the hidden layer nodes are the input values to the subsequent nodes of the softmax layer.)
PNG
media_image3.png
180
368
media_image3.png
Greyscale
Regarding Claim 7, Dribus, Middlemas, and Aggarwal teach The method of claim 6, as shown above.
Aggarwal also teaches wherein nodes in a same layer of the trained feedforward neural network have a same activation function. (See figure 1.9 copied above in regard to claim 6. The nodes of the softmax layer all have the same softmax activation function.)
Regarding Claim 11, Dribus and Middlemas teach The method of claim 1, as shown above.
Dribus and Middlemas do not appear to explicitly disclose the remaining features of claim 11.
However, Aggarwal teaches wherein a final layer of the trained feedforward neural network has one or more nodes with linear activation for a regression or logistic function for binary classification. (Pg. 12, section 1.2.1.3: “The linear activation function is often used in the output node, when the target is a real value.” Linear activation is used in the output node (i.e. final layer) for a real-valued target (i.e. regression).)
Claims 15-16 are system claims containing substantially the same elements as method claims 5-6, respectively. Dribus, Middlemas, and Aggarwal teach the elements of claims 5-6, as shown above.
Claims 10 and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Dribus in view of Middlemas, and further in view of
Boguñá et al. (hereinafter Boguñá), “Generalized percolation in random directed networks”.
Regarding Claim 10, Dribus and Middlemas teach The method of claim 9, as shown above.
Dribus and Middlemas do not appear to explicitly disclose the remaining features of claim 10.
However, Boguñá teaches wherein bidirectional connections between two nodes of the trained feedforward neural network enable an exchange of information that allows for a mixing of information that reaches different regions of the trained feedforward neural network. (Pg. 1, section I: “In this paper we present a general theory for percolation in directed random networks with general two point correlations and bidirectional links.” Pg. 2, section II: “The giant connected component in undirected graphs becomes internally structured in the case of directed networks so that four different types of giant components may arise. Whether giant or not, these components are characterized as follows… The strongly connected component, SCC, the set of vertices reachable from its every vertex by a directed path.” Pg. 7, section VI: “In this case, we have also shown that bidirectional edges act as a catalyst for percolation, favoring the emergence of the GSCC [giant strongly connected component]…” Bidirectional edges between nodes of a directed network facilitate the emergence of strong connectedness (i.e. mixing of information that reaches different regions of the network).)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Dribus, Middlemas, and Boguñá. Dribus teaches a network architecture construction algorithm for constructing a sparse neural network topology, where connectivity is based on the hyperuniform tetrahedral-octahedral honeycomb. Middlemas teaches that hyperuniformity arises in a variety of systems, including the particle distribution and bond connectivity of quasicrystals. Boguñá teaches the impacts of bidirectional edges on percolation in directed networks. One of ordinary skill would have motivation to combine Dribus, Middlemas, and Boguñá because Dribus is concerned with “focusing on three technical problems elaborated below, namely, the problems of achieving sparsity, locality, and connectivity in such networks” (Dribus, 0002), and seeks to “achiev[e] sufficient connectivity in an ANN to ensure flexible training potential” (Dribus, 0041), and according to Boguñá, “bidirectional edges act as a catalyst for percolation, favoring the emergence of the GSCC [giant strongly connected component]” (Boguñá, pg. 7, section VI).
Claim 19 is a system claim containing substantially the same elements as method claim 10. Dribus, Middlemas, and Boguñá teach the elements of claim 10, as shown above.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BENJAMIN M ROHD whose telephone number is (571)272-6445. The examiner can normally be reached Mon-Thurs 8:00-6:00 EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Viker Lamardo can be reached at (571) 270-5871. 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.
/B.M.R./Examiner, Art Unit 2147
/ERIC NILSSON/Primary Examiner, Art Unit 2151
Prosecution Insights