DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
This action is in response to Applicant’s remarks, filed on 4/7/2026. The amendments to claim(s) 1, 4, 8, 10, 12, 14, 16, 26, 100 and 102 have been entered. Claim(s) 2-3, 6-7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27-99, 101, 103 and 105 is/are previously cancelled by Applicant and therefore withdrawn from further consideration pursuant to 37 CFR 1.142(b). New claim(s) 106 has been entered. Accordingly, claim(s) 1, 4-5, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 100, 102, 104 and 106 remain pending for examination.
Response to Arguments
Applicant’s arguments, see p. 8-12, with respect to claim(s) 1, 4-5, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 100, 102 and 104 have been fully considered.
After review of the Applicant’s amendments regarding the objection to claim(s) 4, Examiner respectfully agrees with Applicant and the objection has been withdrawn.
Regarding the rejection(s) under 35 U.S.C. § 112(b), Examiner respectfully agrees with the Applicant and the prior 35 U.S.C. § 112(b) rejections have been withdrawn.
New grounds of rejection are made in view of the following: new amendments provided by Applicant and attached remarks; updated search and review of pertinent, eligible prior art; newly added claims; and/or different interpretation of the previously applied references. Applicant’s arguments with respect to claim(s) 1, 4-5, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 100, 102, 104 and 106 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.
Examiner respectfully notes that Applicant’s arguments only address independent claim(s) 1, and no remarks regarding the subject matter of the dependent claim(s) have been presented. Accordingly, the rejections to dependent claims 4-5, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 100, 102, 104 and 106 are modified to address Applicant’s amendments and the new rejection to independent claim(s) 1 and are sustained. The rejections of claim(s) 1, 4-5, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 100, 102, 104 and 106 under 35 U.S.C. § 102 and 35 U.S.C. §103 are maintained.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claim(s) 1, 4-5, 8, 10, 12, 14, 16, 18, 20, 24, 26, 102, 104 and 106 is/are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Zhang et al. (“Deep Fusion of Brain Structure-Function in Mild Cognitive Impairment,” Medical Image Analysis, Vol 72, (2021) 102082, ISSN 1361-8415, 4-23-2021; hereinafter “Zhang”) as provided by Applicant.
Regarding claim 1, Zhang teaches a method of determining a condition of a subject (“a graph-based deep neural network to simultaneously model brain structure and function in Mild Cognitive Impairment (MCI): the topology of the graph is initialized using structural network (from diffusion MRI) and iteratively updated by incorporating functional information (from functional MRI) to maximize the capability of differentiating MCI patients from elderly normal controls” [abst]; [p.1-3], [fig. 1-2]), the method comprising:
receiving functional magnetic resonance (MR) data and structural MR data, each describing the brain of the subject in a respective native space of said brain (“a graph-based deep model (GBDM) ( Fig. 2 ) to analyze brain structure-function abnormalities in MCI patients by integrating both structural and functional data.” [p.4]; “model is composed of four components: 1) learning of functional profile to parameterize pairwise functional relations between any two brain regions; 2) brain structure-function fusion for seeking to best combine both structural network and the learned functional profile as new topology of the graph; […] A major goal of this work is to examine MCI related network alterations via deep fusion of brain structural and functional data.” [p.4]; “For rs-fMRI images, we applied spatial smoothing, slice time correction, temporal pre-whitening, global drift removal and band pass filtering […] For DTI images, we applied eddy current correction via FSL and fiber tracking via MedINRIA. For T1-weighted images, we registered them to DTI space by FSL FLIRT and then conducted segmentation” [p.7]; [p.4-8], [fig. 1-4]);
receiving images of an anatomical atlas defined over a standardized space (“The T1-weighted MRI data has 240 × 256 × 208 voxels […] The DTI data has 54 gradient directions, each volume has 116 × 116 × 80 voxels […] The rs-fMRI data has 197 volumes, each volume has 64 × 64 × 48 voxels” [p.6]; “For DTI images, we applied eddy current correction via FSL and fiber tracking via MedINRIA. For T1-weighted images, we registered them to DTI space by FSL FLIRT and then conducted segmentation using FreeSurfer package (Fischl (2012)). After the segmentation, we adopted the Destrieux Atlas (Destrieux et al. (2010)) for ROI labeling and the brain cortex is partitioned into 148 regions after removing two unknown areas and two empty areas.” [p.7]; [p.4-8], [fig. 1-4]);
applying a transformation of said images of said anatomical atlas onto said respective native space, to provide respective parcellated functional MR data and structural MR data over said respective native space (“Brain structural network is used to initialize the topology of the graph, i.e., the adjacency matrix in GCN. An individual functional profile is learned and combined with structural network iteratively. […] 2) brain structure-function fusion for seeking to best combine both structural network and the learned functional profile as new topology of the graph; 3) brain network convolution conducted upon the updated graph topology;” [p.4]; “After the segmentation, we adopted the Destrieux Atlas (Destrieux et al. (2010)) for ROI labeling and the brain cortex is partitioned into 148 regions after removing two unknown areas and two empty areas. […] For each subject, the whole brain is divided into 148 regions using Destrieux Atlas. We calculate averaged fMRI signal for each brain region and created brain structural network (AS) and Pearson Correlation Coefficient matrix (P) for each subject.” [p.7]; The brain of each subject is divided using Destrieux atlas into regions, wherein fMRI signals and brain structural networks are generated for each region of subject brain [p.4-8], [fig. 1-4]);
constructing a subject-specific functional connectome (FC) using said parcellated functional MR data, and a subject-specific structural connectome (SC) using said parcellated structural MR data (“a novel graph-based deep model (GBDM) to study brain structure-function fusion at connectome level. We construct a multi-layer GCN with trainable graph topology. This graph is parameterized by both DTI-derived brain structural network and functional activities so that the learned graph becomes a deeply hybrid connectome by retaining brain structural substrate and simultaneously taking into account the functional influences as a complementary cross-modal information. […] Because this predicted graph evolves from both structural and functional connectome in a deep manner, we named it as Deep Brain Connectome.” [p.2]; Deep brain connectomes based on structural DTI and functional fMRI were generated for each subject [p.4-8], [fig. 1-5; see fig. 2 reproduced below]); and
PNG
media_image1.png
518
560
media_image1.png
Greyscale
Functional and structural connectomes are generated from MRI data of the subject (Zhang [fig. 2])
analyzing said FC and said SC to estimate a condition of the subject (“Through this way, the disease-related knowledge drives the training process to learn a new brain connectome […] which represents a deep fusion of brain structure and function, that is the deep brain connectome.” [p.6]; “The proposed GBDM is based on MCI/NC classification task. […] GBDM is designed to learn a deeply combined structural-functional connectome that can be used to achieve higher MCI/NC classification performance.” [p.7]; “The core idea of deep brain connectome is to identify the disease related brain network by learning the connectome topology (using both structural and functional information) instead of fixing the predefined brain network. In this work, our proposed model has been applied to MCI/NC classification task, however, it can be easily extended to other tasks. […] our deep brain connectome can be a promising approach to explore the underlying relations between brain structural and functional perturbations at network level in both neurological and psychiatric diseases.” [p.15]; The resulting model is applied to subjects to classify mild cognitive impairment, and may be further applied to classify and predict brain diseases [p.4-8], [fig. 1-5]).
Regarding claim 4, Zhang teaches the method according to claim 1,
Zhang further teaching comprises calculating said transformation by receiving a mean MR image of said brain, and registering a template image defined over said standardized space onto said mean MR image (“We calculated the average fMRI signal for each brain region as the representative for later analysis. […] We normalized the averaged fMRI signal using the standard Z-score normalization” [p.4]; “the disease-related knowledge (from classification) is passed to functional profile (AF) and then
transferred to the new brain connectome (Aˆ), by combining with structural network (AS).” [p.5]; [p.4-8], [fig. 1-5]).
Regarding claim 5, Zhang teaches the teaches the method according to claim 4,
Zhang further teaching wherein said mean MR image is based on a volume average of at least one of said structural and functional MR data (“The DTI data has 54 gradient directions, each volume has 116 × 116 × 80 voxels […] The rs-fMRI data has 197 volumes, each volume has 64 × 64 × 48 voxels” [p.6]; “We calculate averaged fMRI signal for each brain region and created brain structural network (AS) and Pearson Correlation Coefficient matrix (P) for each subject.” [p.7]; [p.4-8], [fig. 1-5], [see claim 1 rejection]).
Regarding claim 8, Zhang teaches the method according to claim 1,
Zhang further teaching wherein said analyzing comprises:
accessing a computer readable medium storing a trained convolutional neural network (CNN) having a first set of layers trained for separately processing FC and SC and for providing combined outputs pertaining to said separate processing, and a second set of layers trained for processing said combined outputs from said first set of layers (“our model is composed of four components: 1) learning of functional profile to parameterize pairwise functional relations between any two brain regions; […] 3) brain network convolution conducted upon the updated graph topology; 4) MCI-NC classification with fully connected neural network.” [p.4]; “we combine the normalized structural network (AS) and functional profile (AF) […] the disease-related knowledge (from classification) is passed to functional profile ( AF ) and then transferred to the new brain connectome ( Aˆ ) , by combining with structural network ( AS ). This is an iterative process and at each iteration, Aˆ will be used as the new topology for graph convolution
of node-associated features.” [p.5]; “brain network based convolutions were implemented using a two-layer GCN.” [p.6]; [p.4-8], [fig. 1-5], [see claim 1 rejection]); and
feeding said CNN with said subject-specific FC and SC (“In the last part of GBDM, we designed a two-layer fully connected neural network to perform binary classification for two classes – MCI and NC. […] For each sample (subject) in training
data, the individual structural network is used to initialize the adjacency matrix with (6). Individual functional signals are used for functional profile learning ( see (3) ), brain structure-function fusion ( see (6) ) as well as node features in (9).” [p.6]; “After applying the trained GBDM to each testing subject, the individual deep brain connectome – Aˆ can be computed” [p.8]; [p.4-8], [fig. 1-5], [see claim 1 rejection]);
wherein said first set of layers comprises one hidden convolutional layer trained for separately processing FC, and one hidden convolutional layer trained for separately processing SC (“According to the input adjacency matrix Aˆ and feature matrix P, a GCN layer creates a hidden representation for each node by combining features from its neighbor nodes based on Wl. After the combination, a nonlinear transformation is applied to the hidden representation.” [p.7, fig. 3 inset]; “Deep neural networks contain multiple non-linear hidden layers which makes them powerful when learning complicated relationships among data samples.” [p.13]; [p.4-8], [fig. 1-5], [see claim 1 rejection]).
Regarding claim 10, Zhang teaches the method according to claim 1,
Zhang further teaching wherein said analyzing comprises:
accessing a computer readable medium storing a trained convolutional neural network (CNN) having a first set of layers trained for separately processing FC and SC and for providing combined outputs pertaining to said separate processing, and a second set of layers trained for processing said combined outputs from said first set of layers (“our model is composed of four components: 1) learning of functional profile to parameterize pairwise functional relations between any two brain regions; […] 3) brain network convolution conducted upon the updated graph topology; 4) MCI-NC classification with fully connected neural network.” [p.4]; “we combine the normalized structural network (AS) and functional profile (AF) […] the disease-related knowledge (from classification) is passed to functional profile ( AF ) and then transferred to the new brain connectome ( Aˆ ) , by combining with structural network ( AS ). This is an iterative process and at each iteration, Aˆ will be used as the new topology for graph convolution
of node-associated features.” [p.5]; “brain network based convolutions were implemented using a two-layer GCN.” [p.6]; [p.4-8], [fig. 1-5], [see claim 1 rejection]); and
feeding said CNN with said subject-specific FC and SC (“In the last part of GBDM, we designed a two-layer fully connected neural network to perform binary classification for two classes – MCI and NC. […] For each sample (subject) in training
data, the individual structural network is used to initialize the adjacency matrix with (6). Individual functional signals are used for functional profile learning ( see (3) ), brain structure-function fusion ( see (6) ) as well as node features in (9).” [p.6]; “After applying the trained GBDM to each testing subject, the individual deep brain connectome – Aˆ can be computed” [p.8]; [p.4-8], [fig. 1-5], [see claim 1 rejection]);
wherein said first set of layers comprises more than one hidden convolutional layer trained for separately processing FC (“According to the input adjacency matrix Aˆ and feature matrix P, a GCN layer creates a hidden representation for each node by combining features from its neighbor nodes based on Wl. After the combination, a nonlinear transformation is applied to the hidden representation.” [p.7, fig. 3 inset]; “Deep neural networks contain multiple non-linear hidden layers which makes them powerful when learning complicated relationships among data samples.” [p.13]; [p.4-8], [fig. 1-5], [see claim 1 rejection]).
Regarding claim 12, Zhang teaches the method according to claim 1,
Zhang further teaching wherein said analyzing comprises:
accessing a computer readable medium storing a trained convolutional neural network (CNN) having a first set of layers trained for separately processing FC and SC and for providing combined outputs pertaining to said separate processing, and a second set of layers trained for processing said combined outputs from said first set of layers (“our model is composed of four components: 1) learning of functional profile to parameterize pairwise functional relations between any two brain regions; […] 3) brain network convolution conducted upon the updated graph topology; 4) MCI-NC classification with fully connected neural network.” [p.4]; “we combine the normalized structural network (AS) and functional profile (AF) […] the disease-related knowledge (from classification) is passed to functional profile ( AF ) and then transferred to the new brain connectome ( Aˆ ) , by combining with structural network ( AS ). This is an iterative process and at each iteration, Aˆ will be used as the new topology for graph convolution
of node-associated features.” [p.5]; “brain network based convolutions were implemented using a two-layer GCN.” [p.6]; [p.4-8], [fig. 1-5], [see claim 1 rejection]); and
feeding said CNN with said subject-specific FC and SC (“In the last part of GBDM, we designed a two-layer fully connected neural network to perform binary classification for two classes – MCI and NC. […] For each sample (subject) in training
data, the individual structural network is used to initialize the adjacency matrix with (6). Individual functional signals are used for functional profile learning ( see (3) ), brain structure-function fusion ( see (6) ) as well as node features in (9).” [p.6]; “After applying the trained GBDM to each testing subject, the individual deep brain connectome – Aˆ can be computed” [p.8]; [p.4-8], [fig. 1-5], [see claim 1 rejection]);
wherein said first set of layers comprises more than one hidden convolutional layer trained for separately processing SC (“According to the input adjacency matrix Aˆ and feature matrix P, a GCN layer creates a hidden representation for each node by combining features from its neighbor nodes based on Wl. After the combination, a nonlinear transformation is applied to the hidden representation.” [p.7, fig. 3 inset]; “Deep neural networks contain multiple non-linear hidden layers which makes them powerful when learning complicated relationships among data samples.” [p.13]; [p.4-8], [fig. 1-5], [see claim 1 rejection]).
Regarding claim 14, Zhang teaches the method according to claim 1,
wherein said analyzing comprises:
accessing a computer readable medium storing a trained convolutional neural network (CNN) having a first set of layers trained for separately processing FC and SC and for providing combined outputs pertaining to said separate processing, and a second set of layers trained for processing said combined outputs from said first set of layers (“our model is composed of four components: 1) learning of functional profile to parameterize pairwise functional relations between any two brain regions; […] 3) brain network convolution conducted upon the updated graph topology; 4) MCI-NC classification with fully connected neural network.” [p.4]; “we combine the normalized structural network (AS) and functional profile (AF) […] the disease-related knowledge (from classification) is passed to functional profile ( AF ) and then transferred to the new brain connectome ( Aˆ ) , by combining with structural network ( AS ). This is an iterative process and at each iteration, Aˆ will be used as the new topology for graph convolution
of node-associated features.” [p.5]; “brain network based convolutions were implemented using a two-layer GCN.” [p.6]; [p.4-8], [fig. 1-5], [see claim 1 rejection]); and
feeding said CNN with said subject-specific FC and SC (“In the last part of GBDM, we designed a two-layer fully connected neural network to perform binary classification for two classes – MCI and NC. […] For each sample (subject) in training
data, the individual structural network is used to initialize the adjacency matrix with (6). Individual functional signals are used for functional profile learning ( see (3) ), brain structure-function fusion ( see (6) ) as well as node features in (9).” [p.6]; “After applying the trained GBDM to each testing subject, the individual deep brain connectome – Aˆ can be computed” [p.8]; [p.4-8], [fig. 1-5], [see claim 1 rejection]);
wherein said second set of layers comprises at least two hidden convolutional layers (“According to the input adjacency matrix Aˆ and feature matrix P, a GCN layer creates a hidden representation for each node by combining features from its neighbor nodes based on Wl. After the combination, a nonlinear transformation is applied to the hidden representation.” [p.7, fig. 3 inset]; “Deep neural networks contain multiple non-linear hidden layers which makes them powerful when learning complicated relationships among data samples.” [p.13]; [p.4-8], [fig. 1-5], [see claim 1 rejection]).
Regarding claim 16, Zhang teaches the method according to claim 1,
Zhang further teaching wherein said constructing said subject-specific FC, comprises extracting from said functional MR data a plurality of time-ordered series of values, each series corresponding to a different region of said brain and each value representing a blood-oxygenation-level-dependent (BOLD) signal acquired from a respective region at a respective time point (“Functional information (Fig. 1(a)) is used in two ways: the fMRI BOLD signals of each pair of brain regions are parameterized to form trainable functional profiles which are integrated into the current graph at each iteration; the functional connectivity are treated as features associated with the nodes that represent different brain regions.” [p.2]; “By using Destrieux atlas (Destrieux et al. (2010)) along with DTI and resting state fMRI data, we extracted the averaged BOLD signal of each brain region (148 regions in total) (Fig. 1(a)) and constructed brain structural network (Fig. 1(b)).” [p.3, fig. 1 inset]; [p.4-8], [fig. 1-5]), and
constructing a correlation matrix describing correlation among said plurality of series, wherein said subject-specific FC is said correlation matrix (“An individual functional profile is learned and combined with structural network iteratively. Pearson’s correlation coefficients of averaged BOLD signals are treated as the features associated with the nodes of the graph.” [p.4]; “Using functional connectivity (defined with Pearson’s correlation coefficient) as features associated to the network nodes (brain regions), we conduct graph convolution based on the hybrid brain network.” [p.5, fig. 2 inset]; “We calculated Pearson Correlation Coefficient of every pair of averaged fMRI signals as feature matrix” [p.6]; [p.4-8], [fig. 1-5]).
Regarding claim 18, Zhang teaches the method according to claim 16,
Zhang further teaching wherein said correlation is selected from the group consisting of a pairwise correlation, a partial correlation, and a distance correlation (“We defined the parameterized functional-pairwise distance between region- i and region- j” [p.4]; “Using functional connectivity (defined with Pearson’s correlation coefficient) as features associated to the network nodes (brain regions), we conduct graph convolution based on the hybrid brain network.” [p.5, fig. 2 inset]; “We calculated Pearson Correlation Coefficient of every pair of averaged fMRI signals as feature matrix” [p.6]; [p.4-8], [fig. 1-5]).
Regarding claim 20, Zhang teaches the method according to claim 1,
Zhang further teaching wherein said constructing said subject-specific SC, comprises applying whole brain tractography to define a plurality of streamlines or fractional anisotropy values between pairs of regions of said brain, and converting said plurality of streamlines or fractional anisotropy values to a connectivity matrix, wherein said subject-specific SC is said connectivity matrix (“we calculate individual structural network matrix – AS ∈ RN×N , which is a symmetric matrix and ASij ∈ R is the number of DTI-derived fibers connecting brain region i and region j.” [p.4]; “For rs-fMRI images, we applied spatial smoothing, slice time correction, temporal pre-whitening, global drift removal and band pass filtering […] For DTI images, we applied eddy current correction via FSL and fiber tracking via MedINRIA. For T1-weighted images, we registered them to DTI space by FSL FLIRT and then conducted segmentation” [p.7]; “By comparing the learned deep brain connectome with the original brain structural network, we are able to examine the functional influences on the structural connectivity (fusion of structural and functional data) when conducting MCI classification task.” [p.8]; [p.4-8], [fig. 1-5]).
Regarding claim 24, Zhang teaches the method according to claim 1,
Zhang further teaching comprising predicting a clinical outcome of the condition (“In the last part of GBDM, we designed a two-layer fully connected neural network to perform binary classification for two classes – MCI and NC.” [p.6]; [p.4-8], [fig. 1-5], [see claim 1 rejection]).
Regarding claim 26, Zhang teaches the method according to claim 1,
Zhang further teaching comprising predicting, based on said analysis of said FC and said SC, a likelihood for at least one of: brain concussion, depressive disorder, stroke, traumatic brain injury, post-traumatic stress disorder, epilepsy, Parkinson, multiple sclerosis, agitation, abuse, Alzheimer's disease, anxiety, panic, phobic disorder, bipolar disorder, borderline personality disorder, behavior control disorder, body dysmorphic disorder, cognitive impairment, dissociative disorder, eating disorder, fatigue, impulse-control disorder, irritability, obsessive-compulsive disorder, personality disorders, psychotic disorder, sexual disorders, sleep disorder, stuttering, Tourette's Syndrome, Trichotillomania, self- destructive behavior, fibromyalgia, tremor, schizophrenia, attention-deficit disorder, hyperactivity disorder, and learning disorder (“In this work, we developed a graph-based deep neural network to simultaneously model brain structure and function in Mild Cognitive Impairment (MCI)” [abst]; “In the last part of GBDM, we designed a two-layer fully connected neural network to perform binary classification for two classes – MCI and NC.” [p.6]; [p.4-8], [fig. 1-5], [see claim 1 rejection]).
Regarding claim 102, Zhang teaches a computer software product, comprising a computer-readable medium in which program instructions are stored, which instructions, when read by an image processor, cause the data processor to execute the method according to claim 1, including said receiving of said functional MR data and said structural MR data, each describing the brain of said subject (“we developed a graph-based deep neural network to simultaneously model brain structure and function in Mild Cognitive Impairment (MCI): the topology of the graph is initialized using structural network (from diffusion MRI) and iteratively updated by incorporating functional information (from functional MRI) to maximize the capability of differentiating MCI patients from elderly normal controls” [abst]; “a graph-based deep model (GBDM) ( Fig. 2 ) to analyze brain structure-function abnormalities in MCI patients by integrating both structural and functional data.” [p.4]; “The T1-weighted MRI data has 240 × 256 × 208 voxels […] The DTI data has 54 gradient directions, each volume has 116 × 116 × 80 voxels […] The rs-fMRI data has 197 volumes, each volume has 64 × 64 × 48 voxels” [p.6]; [p.4-8], [fig. 1-5], [see claim 1 rejection]).
Regarding claim 104, Zhang teaches a magnetic resonance imaging (MRI) system for imaging a brain of a subject (“we developed a graph-based deep neural network to simultaneously model brain structure and function in Mild Cognitive Impairment (MCI): the topology of the graph is initialized using structural network (from diffusion MRI) and iteratively updated by incorporating functional information (from functional MRI) to maximize the capability of differentiating MCI patients from elderly normal controls” [abst]; [fig. 1-5], [see claim 1 rejection]), the system comprising:
an MRI scanner configured for scanning the brain to provide functional MR data and structural MR data, each describing the brain (“we jointly analyzed three modalities of brain imaging data in ADNI dataset, including structure MRI (T1-weighted), resting state fMRI (rs-fMRI) and DTI.” [p.6]; [fig. 1-5], [see claim 1 rejection]); and
an image processor configured for executing the method according to claim 1 (“a graph-based deep model (GBDM) ( Fig. 2 ) to analyze brain structure-function abnormalities in MCI patients by integrating both structural and functional data.” [p.4]; “model is composed of four components: 1) learning of functional profile to parameterize pairwise functional relations between any two brain regions; 2) brain structure-function fusion for seeking to best combine both structural network and the learned functional profile as new topology of the graph; […] A major goal of this work is to examine MCI related network alterations via deep fusion of brain structural and functional data.” [p.4]; [p.4-8], [fig. 1-5], [see claim 1 rejection]).
Regarding claim 106, Zhang teaches the method according to claim 8,
Zhang further teaching wherein said analyzing comprises generating a plurality of FC activation maps and a plurality of SC activation maps, and wherein said combined outputs comprise a concatenation between a respective FC activation map and respective SC activation map (“the disease-related knowledge (from classification) is passed to functional profile ( A F ) and then transferred to the new brain connectome ( Aˆ ) , by combining with structural network ( A S ).” [p.5]; “The feature vector of node […] is the concatenation of correlations to all the other nodes. The input graphs in our model are individual-level graphs which have the same number of nodes representing the corresponding brain regions.” [p.6]; [p.4-8], [fig. 1-5]).
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.
Claim(s) 22 and 100 is/are rejected under 35 U.S.C. 103 as being obvious over Zhang as applied to claim 1 above, in view of Madhavan et al. (US20200230413A1, 2020-07-23; hereinafter “Madhavan”).
Regarding claim 22, Zhang teaches the method according to claim 1,
but Zhang may fail to teach predicting a response to a treatment for the condition.
However, in the same field of endeavor, Madhavan teaches a method of determining a condition of a subject (“A computer-based method for identifying a patient-specific neurosurgery target location,” [clm 16]; “personalized functional neurosurgery targeting and brain stimulation programming that is derived from a brain connectivity atlas along with functional and structural imaging and clinical/psychometric testing” [0001]; “resting brain state conditions are observed through changes in blood flow in the brain which creates what is referred to as a blood-oxygen-level dependent (BOLD) signal” [0044]; [fig. 2, 5]);
Madhavan further teaching comprising predicting a response to a treatment for the condition (“determining an optimal set of DBS parameters to be used for patient treatment at the neurosurgery target location that achieves the optimal treatment of the patient.” [clm 21]; “in generating the functional connectivity map, an fMRI series may be analyzed to identify couplings between regions of a brain that may work together to perform a particular type of function or to respond to a specific class of stimulus” [0034]; “as the population-based quantitative connectome atlas includes not only population-based data on ideal neurosurgery target locations for each of various disorders, but also data on (1) the stimulation parameters used/applied during treatment, such as DBS parameters including voltage, frequency, and pulse width of applied DBS signals, as well as electrode contacts activated to apply the DBS signals, for example, and (2) clinical outcomes for the cohort of patients achieved using those stimulation parameters, optimal stimulation parameters may be estimated for the patient.” [0049]; The outcome of a functional neurosurgical treatment (e.g., DBS) can be estimated based on review and comparison of patient MRI data with brain connectivity atlas, which provides the optimal treatment location in brain and the predicted patient outcomes [0032-0066], [fig. 2-5]).
It would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to combine the method of determining a condition of a subject taught by Zhang by predicting a response to a treatment for the condition as taught by Madhavan. With the availability of large-scale multiple types of brain image data, integration of data acquired from different imaging techniques, termed as multimodal data fusion, has gained considerable attention in neuroimaging field. Besides studying general relationships between brain structure and function, multimodal data fusion can provide complementary knowledge when exploring and identifying potential abnormalities occurred in brain disorders (Zhang [p.1]). The quantitative connectome atlas comprises a disease-specific, population-based quantitative connectome atlas that identifies an optimal target location for treatment associated with a maximal clinical improvement for each disease in the population of patients (Madhavan [abst]).
Regarding claim 100, Zhang teaches a method of treating a disorder, comprising:
Zhang further teaching executing the method according to claim 1, to determine said condition for the subject, said condition being a disorder (“a graph-based deep neural network to simultaneously model brain structure and function in Mild Cognitive Impairment (MCI): the topology of the graph is initialized using structural network (from diffusion MRI) and iteratively updated by incorporating functional information (from functional MRI) to maximize the capability of differentiating MCI patients from elderly normal controls” [abst]; [p.1-3], [fig. 1-2], [see claim 1 rejection]);
but Zhang may fail to explicitly teach applying to the subject a treatment selected to specifically treat said determined disorder.
However, in the same field of endeavor, Madhavan teaches applying to the subject a treatment selected to specifically treat said determined disorder (“determining an optimal set of DBS parameters to be used for patient treatment at the neurosurgery target location that achieves the optimal treatment of the patient.” [clm 21]; “a completed quantitative functional connectome atlas is generated/output at STEP 42 that enables identification of a convergent “probabilistic zone of optimal stimulation” and their associated connectomes, which are based on treatment localization (i.e., lesion location for ablative procedure and electrode location for DBS), stimulation parameters (for DBS), and clinical outcomes in various disorders.” [0040];“as the population-based quantitative connectome atlas includes not only population-based data on ideal neurosurgery target locations for each of various disorders, but also data on (1) the stimulation parameters used/applied during treatment, such as DBS parameters including voltage, frequency, and pulse width of applied DBS signals, as well as electrode contacts activated to apply the DBS signals, for example, and (2) clinical outcomes for the cohort of patients achieved using those stimulation parameters, optimal stimulation parameters may be estimated for the patient.” [0049]; The optimal treatment is determined based on comparison and analysis of patient brain MRI data and then applied to patient [0032-0066], [fig. 2-5]).
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to James F. McDonald III whose telephone number is (571)272-7296. The examiner can normally be reached M-F; 8AM-6PM 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, Chris Koharski can be reached at 5712727230. 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.
JAMES FRANKLIN MCDONALD III
Examiner
Art Unit 3797
/CHRISTOPHER KOHARSKI/Supervisory Patent Examiner, Art Unit 3797
Prosecution Insights