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 .
Claim Rejections - 35 USC § 103
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 (i.e., changing from AIA to pre-AIA ) 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.
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 text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claims 1-3, 6, 10-12, 16-19, 21-25, 28, 32-33, and 36 are rejected under 35 U.S.C. 103 as being unpatentable over Guo et al. NPL “Predicting 15O-Water PET cerebral blood flow maps from multi-contrast MRI using a deep convolutional neural network with evaluation of training cohort bias” in view of Parsey et al. (US 20160239966 A1).
Regarding claim 1, Guo et al. teaches a computer-implemented method of training a model to generate positron emission tomography (PET) images from magnetic resonance imaging (MRI) images (see Abstract; “a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images to predict gold-standard 15O-water PET CBF images obtained on a simultaneous PET/MRI scanner”, see also page 2442, right col., 1st para; “We developed a dCNN to synthesize high-quality, PET like CBF maps from only MRI inputs including ASL”); the method comprising: selecting a healthy dataset, the healthy dataset including paired MR and PET images from healthy subjects (see Abstract; “16 healthy controls (HC) and 16 cerebrovascular disease patients (PT) with 4-fold cross-validation”), creating a training dataset from the combination of healthy dataset and the target dataset (see Abstract; “The dCNN was trained and tested on 64 scans in 16 healthy controls (HC) and 16 cerebrovascular disease patients (PT) with 4-fold cross-validation…The dCNN trained with the mixed HC and PT cohort performed the best”); training a machine learning model using the training dataset (see page 2248, Fig. 4; “deep convolutional neural networks (dCNNs) trained on different datasets”). However, Guo et al. does not teach selecting a target dataset, the target dataset including paired MR and PET images from subjects having a condition corresponding to a biomarker of interest.
In the same field of endeavor, Parsey et al. selecting a target dataset (see para [0059]; “individuals in the control group and the reference group are separated into groups according to the presence or absence of a neurological disorder….. the neurological disorder is Alzheimer's disease)”, see also para [0094]; “the groups can be selected such that one group comprises individuals having, or at risk of having a neurological disorder, and the other group corresponds to reference group not having a neurological disorder. The groups can be also selected such that one group consists of individuals having, or at risk of having a neurological disorder, and the other group consists of individuals not having a neurological disorder”), the target dataset including paired MR and PET images from subjects having a condition corresponding to a biomarker of interest (see para [0078]; “the collection of MRI and PET images one or more subjects from two groups of subjects (e.g. a group comprising individuals having a neurological disorder and a reference group comprising individuals not having a neurological disorder”, and para [0082]; “Such a robust metric can improve diagnosis or can be used to select patients with confirmed amyloid deposition for clinical trials involving novel drugs aimed at reducing amyloid load in AD…. an automated basis for making the distinction between Alzheimer's and healthy control subjects using [.sup.11C]PIB. Such a robust metric can improve diagnosis or can be used to select patients with confirmed amyloid deposition for clinical trials involving novel drugs aimed at reducing amyloid load in AD”). Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention to modify a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images of Guo et al. in view of methods for diagnosing a neurological disorder in a subject of Parsey et al. in order to diagnose and making treatment decisions in subjects having a neurological disorder (see para [0059]).
Regarding claim 2, the rejection of claim 1 is incorporated herein.
Guo et al. in the combination further teach wherein MR images comprise at least one MR sequence image (see Fig. 1a, page 2241, left col. 3rd para; “we propose a deep learning-based
method that takes multiple MRI scans”).
Regarding claim 3, the rejection of claim 2 is incorporated herein.
Guo et al. in the combination further teach wherein the MR sequence image comprises an MR T1 image, or MR T2 image, or a combination of images acquired using more than one MR sequences (see page 2242, right col. last para; “(2) structural scans: T1w andT2w FLAIRimages”).
Regarding claim 6, the rejection of claim1 is incorporated herein.
Paresy et al. in the combination further teach wherein the condition is a brain disorder (see para [0014]; “the standard brain atlas is a specific brain atlas created for a particular neurological disorder”).
Regarding claim 10, the rejection of claim1 is incorporated herein.
Guo et al. in the combination further teach wherein the machine learning model comprises a deep learning model (see Fig. 1; “Conceptual framework of the deep convolutional neural network (dCNN)”).
Regarding claim 11, the rejection of claim 10 is incorporated herein.
Guo et al. in the combination further teach wherein the deep learning model comprises an encoder-decoder with 3D convolution layers with ReLU activation operators (see page 2242, right col., 1st para; “We created a U-Net dCNN35 which includes three encoder layers and three decoder layers. Each encoder layer consists of three 2 D convolutional layers with a 3 × 3 kernel, a batch normalization layer, 36 a rectifier linear unit (ReLU) activation layer”).
Regarding claim 12, the rejection of claim 1 is incorporated herein.
Guo et al. in the combination further teach wherein the machine learning model comprises a deep convolution neural network or a diffusion network (see Fig. 1; “Conceptual framework of the deep convolutional neural network (dCNN)”).
Regarding claim 16, the rejection of claim 1 is incorporated herein.
Guo et al. in the combination further teach wherein the MR and PET images are subjected to preprocessing prior to training the machine learning model (see page 2243, right col. 1st para; “total of 64 PET/MRI datasets were included in this study, consisting of 62,720 images”, see also para 2242, left col. last para; “All images were co-registeredtotheT1wstructural images using Statistical Parametric Mapping software (SPM12), and then normalized to the Montreal Neurological Institute (MNI) brain template32 with 2mm isotropic resolution using Advanced Normalization Tools (ANTs) software.33 The brain was extracted using FSL software.34 To reduce computation requirements, all images were resized to 96 X 96”).
Regarding claim 17, the rejection of claim 16 is incorporated herein.
Parsey et al. in the combination further teach wherein the preprocessing comprises segmentation and co-registration of the MR and PET images (see para [0094]; “the voxel image can be transformed into a subject's MRI-space by performing a PET to MRI alignment (i.e., a “co-registration”). …. the MRI image can be segmented by statistical parametric mapping (e.g., using SPM5 software). For example, such segmenting can be performed to generate a gray matter image map, a white matter image map or a cerebrospinal fluid (CSF) matter image map”).
Regarding claim 18, the rejection of claim 1 is incorporated herein.
Guo et al. in the combination further teach method of generating positron emission tomography (PET) images from magnetic resonance imaging (MRI), the method comprising: obtaining a MR image from a subject; and applying the model trained to the MR image (see Abstract; “a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images to predict gold-standard 15O-water PET CBF images obtained on a simultaneous PET/MRI scanner”, see also page 2442, right col., 1st para; “We developed a dCNN to synthesize high-quality, PET like CBF maps from only MRI inputs including ASL”).
Regarding claim 19, the rejection of claim 18 is incorporated herein.
Guo et al. in the combination further teach wherein the MR image comprises at least one MR T1 image or at least one MR T2 image (see page 2242, right col. last para; “(2) structural scans: T1w andT2w FLAIRimages”).
Regarding claim 21, Guo et al. teaches a computer-implemented method of training a model to generate positron emission tomography (PET) images from multimodal input images ((see Abstract; “a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images to predict gold-standard 15O-water PET CBF images obtained on a simultaneous PET/MRI scanner”, see also page 2441, left col., 3rd para; “we propose a deep learning-based method that takes multiple MRI scans, including ASL, as inputs to predict a simultaneously obtained 15O-water PET CBF map”, Note: under specification’s definition, “multimodal input images” include images from different modalities, image types, or acquisition sequences. Therefore, single-delay ASL, multi-delay ASL, and structural MRI images therefore constitute multimodal input images), the method comprising: selecting a healthy dataset, the healthy dataset including paired PET and multimodal images from healthy subjects (see Abstract; “16 healthy controls (HC) and 16 cerebrovascular disease patients (PT) with 4-fold cross-validation”), creating a training dataset from the combination of healthy dataset and the target dataset (see Abstract; “The dCNN was trained and tested on 64 scans in 16 healthy controls (HC) and 16 cerebrovascular disease patients (PT) with 4-fold cross-validation…The dCNN trained with the mixed HC and PT cohort performed the best”); training a machine learning model using the training dataset (see page 2248, Fig. 4; “deep convolutional neural networks (dCNNs) trained on different datasets”). However, Guo et al. does not teach selecting a target dataset, the target dataset including paired PET and multimodal images from subjects having a condition corresponding to a biomarker of interest.
In the same field of endeavor, Parsey et al. selecting a target dataset (see para [0059]; “individuals in the control group and the reference group are separated into groups according to the presence or absence of a neurological disorder….. the neurological disorder is Alzheimer's disease)”, see also para [0094]; “the groups can be selected such that one group comprises individuals having, or at risk of having a neurological disorder, and the other group corresponds to reference group not having a neurological disorder. The groups can be also selected such that one group consists of individuals having, or at risk of having a neurological disorder, and the other group consists of individuals not having a neurological disorder”), the target dataset including paired MR and multimodal images from subjects having a condition corresponding to a biomarker of interest (see para [0078]; “the collection of MRI and PET images one or more subjects from two groups of subjects (e.g. a group comprising individuals having a neurological disorder and a reference group comprising individuals not having a neurological disorder”, and para [0082]; “Such a robust metric can improve diagnosis or can be used to select patients with confirmed amyloid deposition for clinical trials involving novel drugs aimed at reducing amyloid load in AD…. an automated basis for making the distinction between Alzheimer's and healthy control subjects using [.sup.11C]PIB. Such a robust metric can improve diagnosis or can be used to select patients with confirmed amyloid deposition for clinical trials involving novel drugs aimed at reducing amyloid load in AD”). Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention to modify a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images of Guo et al. in view of methods for diagnosing a neurological disorder in a subject of Parsey et al. in order to diagnose and making treatment decisions in subjects having a neurological disorder (see para [0059]).
Regarding claim 22, the rejection of claim 21 is incorporated herein.
Parsey et al. in the combination further teach wherein the multimodal input images comprise anatomical imaging modalities and/or functional imaging modalities see para [0090]; “the methods described herein relate to methods useful for establishing an ideal diagnostic anatomical VOI”, see also para 0122; “anatomical regions of interest, including a reference region, can be identified with a scan (e.g. a PET scan)” and para 0167]; “Studying these disorders generally involves acquiring a PET and sometimes an MRI for an individual. The PET is used as a measure of the protein and the MRI is used to identify anatomical structures”).
Regarding claim 23, the rejection of claim 21 is incorporated herein.
Parsey et al. in the combination further teach wherein the multimodal input images are selected from at least two imaging modalities selected from the group consisting of magnetic resonance imaging (MRI), PET, computed tomography (CT), single photon emission computed tomography (SPECT), electroencephalography data (EEG), magnetic encephalography data (MEG), near-infrared spectroscopy imaging (NIRS), and functional near-infrared spectroscopy imaging (fNIRS) (see para [0167]; “Studying these disorders generally involves acquiring a PET and sometimes an MRI for an individual. The PET is used as a measure of the protein and the MRI is used to identify anatomical structures by manual or processor-driven determination of regions of interest (ROIs). The PET is then spatially aligned to the MRI after which the PET can be used to quantify the levels of a protein in identified anatomical structures”).
Regarding claim 24, the rejection of claim 23 is incorporated herein.
Guo et al. in the combination further teach wherein MR images comprise at least one MR sequence image (see Fig. 1a, page 2241, left col. 3rd para; “we propose a deep learning-based
method that takes multiple MRI scans”).
Regarding claim 25, the rejection of claim 24 is incorporated herein.
Guo et al. in the combination further teach wherein the MR sequence image comprises an MR T1 image, or MR T2 image, or a combination of images acquired using more than one MR sequences (see page 2242, right col. last para; “(2) structural scans: T1w andT2w FLAIRimages”).
Regarding claim 28, the rejection of claim 21 is incorporated herein.
Paresy et al. in the combination further teach wherein the condition is a brain disorder (see para [0014]; “the standard brain atlas is a specific brain atlas created for a particular neurological disorder”).
Regarding claim 32, the rejection of claim 31 is incorporated herein.
Guo et al. in the combination further teach wherein the machine learning model comprises a deep learning model (see Fig. 1; “Conceptual framework of the deep convolutional neural network (dCNN)”).
Regarding claim 33, the rejection of claim 32 is incorporated herein.
Guo et al. in the combination further teach wherein the deep learning model comprises a multi-stage U-net with 3D convolutional layers (see page 2242, right col., 1st para; “We created aU-NetdCNN35whichincludes three encoder layers and three decoder layers”).
Regarding claim 36, the rejection of claim 21 is incorporated herein.
Guo et al. in the combination further teach wherein the machine learning model comprises a deep convolution neural network or a diffusion network (see Fig. 1; “Conceptual framework of the deep convolutional neural network (dCNN)”).
Claims 4-5, 7-9, 26-27, 29-30 and 31 are rejected under 35 U.S.C. 103 as being unpatentable over Guo et al. in view of Parsey et al. as applied in claim 1 above and further in view of Huang et al. (US 20200010447 A1).
Regarding claim 4, the rejection of claim 1 is incorporated herein. The combination of Guo et al. and Parsey et al. as a whole does not teach wherein the biomarker of interest is a synaptic density tracer.
In the same field of endeavor, Huang et al. teach wherein the biomarker of interest is a synaptic density tracer (see para [0024]; “a method of detecting and/or measuring synaptic density in at least one region of a subject's body”, see also para [0048]; “PET imaging of SV2A as a synaptic density biomarker is a valuable tool for the accurate diagnosis”). Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention to modify a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images of Guo et al. in view of methods for diagnosing a neurological disorder in a subject of Parsey et al. and a method of detecting and/or measuring the amount of synaptic vesicle glycoprotein 2A (SV2A) in a subject's body of Huang et al. in order to evaluate effectiveness of treatment and/or therapy for a disease (see para [0024]).
Regarding claim 5, the rejection of claim 4 is incorporated herein.
Huang et al. in the combination further teach wherein the biomarker of interest is synaptic vesicle protein 2A (SV2A) (see para [0024]; “a method of detecting and/or measuring synaptic density in at least one region of a subject's body……. the method comprises determining a level of SV2A in the at least one region of the body of the subject”).
Regarding claim 7, the rejection of claim 6 is incorporated herein.
Huang et al. in the combination further teach wherein the brain disorder a neuropsychiatric disorder (see para [0044]; “to diagnose a neurodegenerative, neurological, psychiatric, and/or metabolic disease”). Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention to modify a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images of Guo et al. in view of methods for diagnosing a neurological disorder in a subject of Parsey et al. and a method of detecting and/or measuring the amount of synaptic vesicle glycoprotein 2A (SV2A) in a subject's body of Huang et al. in order to evaluate effectiveness of treatment and/or therapy for a disease (see para [0024]).
Regarding claim 8, the rejection of claim 7 is incorporated herein.
Huang et al. in the combination further teach wherein the neuropsychiatric disorder is selected from the group consisting of Alzheimer’s disease, mild cognitive impairment, major depressive disorder, schizophrenia, post-traumatic stress disorder, autism spectrum disorders, anxiety disorders, bipolar disorder and substance use disorders (see para [0044]; “measure changes in SV2A binding as a surrogate for synaptic density, or to diagnose a neurodegenerative, neurological, psychiatric, and/or metabolic disease, such as but not limited to AD, Parkinson's disease (PD), multiple sclerosis (MS), autism, epilepsy, stroke, traumatic brain injury (TBI), schizophrenia, post-traumatic stress disorder (PTSD), depression, and diabetes, or any diseases/disorders where involvement of synaptic disruptions and/or abnormalities are present”).
Regarding claim 9, the rejection of claim 8 is incorporated herein.
Huang et al. in the combination further teach wherein the substance use disorder is cannabis use disorder (see para [0028]; “the disease or disorder is a psychiatric disease, such as …..and/or substance abuse disorder”).
Regarding claim 26, the rejection of claim 21 is incorporated herein.
Huang et al. in the combination further teach wherein the biomarker of interest is a synaptic density tracer (see para [0024]; “a method of detecting and/or measuring synaptic density in at least one region of a subject's body”, see also para [0048]; “PET imaging of SV2A as a synaptic density biomarker is a valuable tool for the accurate diagnosis). Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention to modify a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images of Guo et al. in view of methods for diagnosing a neurological disorder in a subject of Parsey et al. and a method of detecting and/or measuring the amount of synaptic vesicle glycoprotein 2A (SV2A) in a subject's body of Huang et al. in order to evaluate effectiveness of treatment and/or therapy for a disease (see para [0024]).
Regarding claim 27, the rejection of claim 26 is incorporated herein.
Huang et al. in the combination further teach wherein the biomarker of interest is synaptic vesicle protein 2A (SV2A) (see para [0024]; “a method of detecting and/or measuring synaptic density in at least one region of a subject's body……. the method comprises determining a level of SV2A in the at least one region of the body of the subject”).
Regarding claim 29, the rejection of claim 28 is incorporated herein.
Huang et al. in the combination further teach wherein the brain disorder a neuropsychiatric disorder (see para [0044]; “to diagnose a neurodegenerative, neurological, psychiatric, and/or metabolic disease”).
Regarding claim 30, the rejection of claim 29 is incorporated herein.
Huang et al. in the combination further teach wherein the neuropsychiatric disorder is selected from the group consisting of Alzheimer’s disease, mild cognitive impairment, major depressive disorder, schizophrenia, post-traumatic stress disorder, autism spectrum disorders, anxiety disorders, bipolar disorder and substance use disorders (see para [0044]; “measure changes in SV2A binding as a surrogate for synaptic density, or to diagnose a neurodegenerative, neurological, psychiatric, and/or metabolic disease, such as but not limited to AD, Parkinson's disease (PD), multiple sclerosis (MS), autism, epilepsy, stroke, traumatic brain injury (TBI), schizophrenia, post-traumatic stress disorder (PTSD), depression, and diabetes, or any diseases/disorders where involvement of synaptic disruptions and/or abnormalities are present”).
Regarding claim 31, the rejection of claim 30 is incorporated herein.
Huang et al. in the combination further teach wherein the substance use disorder is cannabis use disorder (see para [0028]; “the disease or disorder is a psychiatric disease, such as …..and/or substance abuse disorder”).
Claims 13-15, and 37-38 are rejected under 35 U.S.C. 103 as being unpatentable over Guo et al. in view of Parsey et al. as applied in claim 1, above and further in view of Wang et al. NPL “Generation of synthetic PET images of synaptic densityand amyloid from 18 F- FDG images using deep learning”.
Regarding claim 13, the rejection of claim 1 is incorporated herein. The combination of Guo et al. and Parsey et al. as a whole does not teach wherein the machine learning model comprises a network model trained using: a. MR T1 or T2 to BP 60 PET image; b. MR T1 or T2 to BP 90 PET image; or c. MR T1 or T2 to distribution volume ratio (DVR) PET image; or d. MR T1 or T2 to SUV PET image
In the same field of endeavor, Wang et al. teaches wherein the machine learning model comprises a network model trained using: a. MR T1 or T2 to BP 60 PET image; b. MR T1 or T2 to BP 90 PET image; or c. MR T1 or T2 to distribution volume ratio (DVR) PET image; or d. MR T1 or T2 to SUV PET image (see Table 2, page 5117, right col. 2nd para; “For each subject, a T1-weighted magnetic reso-nance imaging (MRI) scan was obtained at the YaleMRI Center for the registration between PET and MRimages as well as the definition for regions of inter-est (ROI). Registration to MRI was implemented bythe BioImage Suite software for the PET images. 28Specifically, both the 18 F- FDG images of SUVR and K iRatio and the 11C- UCB-J images of SUVR and DVRof the same subject were registered with a rigid trans-formation to each individual's MR image….Similarly, we registered 18 F- FDG datasets (SUVR and K i Ratio) and 11C- PiB datasets(SUVR and K i Ratio) to the MR space and generated ROIs from the MR image of each subject”). Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention to modify a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images of Guo et al. in view of methods for diagnosing a neurological disorder in a subject of Parsey et al. and generation of synthetic PET images of synaptic density and amyloid from 18 F- FDG images using deep learning of Wang et al. in order to assess the feasibility of generating synthetic PET im-ages of less-available tracers from the PET image of another common tracer, in particular 18 F- FDG (see Table 2, page 5117, right col. 2nd para).
Regarding claim 14, the rejection of claim 1 is incorporated herein.
Wang et al. in the combination further teach wherein the training comprises patch-based training (see page 5118, left col. 1st para; “the training data in terms of the number of patches are much larger than the number of images…… During the training, 240 epochs each containing16000 patches were chosen for the convergence”). Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention to modify a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images of Guo et al. in view of methods for diagnosing a neurological disorder in a subject of Parsey et al. and generation of synthetic PET images of synaptic density and amyloid from 18 F- FDG images using deep learning of Wang et al. in order to map the feature map to the label at the final stage (see page 5118, left col. 1st para).
Regarding claim 15, the rejection of claim 14 is incorporated herein.
Wang et al. in the combination further teach wherein the patch-based training comprises extracting at least four random patches for each epoch from each training image (see page 5117, 2.2; “Since training such a 3D network based on entire im-ages is computationally intensive, we extracted patches randomly from the images within the brain mask as the input and output”).
Regarding claim 37, the rejection of claim 21 is incorporated herein.
Wang et al. in the combination further teach wherein the training comprises patch-based training (see page 5118, left col. 1st para “the training data in terms of the number of patches are much larger than the num-ber of images…… During the training, 240 epochs each containing16000 patches were chosen for the convergence”). Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention to modify a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images of Guo et al. in view of methods for diagnosing a neurological disorder in a subject of Parsey et al. and generation of synthetic PET images of synaptic density and amyloid from 18 F- FDG images using deep learning of Wang et al. in order to map the feature map to the label at the final stage (see page 5118, left col. 1st para).
Regarding claim 38, the rejection of claim 37 is incorporated herein.
Wang et al. in the combination further teach wherein the patch-based training comprises extracting at least four random patches for each epoch from each training image (see page 5117, 2.2; “Since training such a 3D network based on entire im-ages is computationally intensive, we extracted patches randomly from the images within the brain mask as the input and output”).
Claim 20 is rejected under 35 U.S.C. 103 as being unpatentable over Guo et al. in view of Parsey et al. as applied in claims 1, and 18 above, and further in view of Zaharchuk et al. (US 20200311914 A1).
Regarding claim 20, the rejection of claim 18 is incorporated herein. The combination of Guo et al. and Parsey et al. as a whole does not teach wherein the subject is a low-dose subject and the MR image is converted to a higher quality PET image than true PET.
In the same field of endeavor, Zaharchuk et al. teaches wherein the subject is a low-dose subject (see para) and the MR image is converted to a higher quality PET image than true PET (see para [0017]; “the low-dose nuclear medicine image further includes a combination of multiple slices and multiple contrast images as input. Here, the combination of the multiple slices and the multiple contrast images can include T1w MR images, T2w MR images, FLAIR MR images”, see para [0020]; “generating high-quality images for radiology imaging modalities and nuclear medicine applications from low-radiation-dose samples”, and para [0037]; “reconstruct standard-dose PET images from ultra-low-dose images”). Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention to modify a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images of Guo et al. in view of methods for diagnosing a neurological disorder in a subject of Parsey et al. and a method of reducing radiation dose for radiology imaging modalities and nuclear medicine of Zaharchuk et al. in order to restore image quality of reduced radiotracer dose PET images (see para [0017).
Claims 34-35 are rejected under 35 U.S.C. 103 as being unpatentable over Guo et al. in view of Parsey et al. as applied in claims 1, 21, and 32 and above and further in view of Zamir et al. NPL “Multi-Stage Progressive Image Restoration”
Regarding claim 34, the rejection of claim 32 is incorporated herein. The combination of Guo et al. and Parsey et al. as a whole does not teach wherein the deep learning model comprises a cross-stage feature fusion (CSFF) strategy.
In the same field of endeavor, Zamir et al. teaches wherein the deep learning model comprises a cross-stage feature fusion (CSFF) strategy (see page 14817, left col. 2nd para; “A
mechanism of cross-stage feature fusion (CSFF) is added”). Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention to modify a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images of Guo et al. in view of methods for diagnosing a neurological disorder in a subject of Parsey et al. and Multi-Stage Progressive Image Restoration of Zamir et al. in order to help propagating multi-scale contextualized features from the earlier to later stages (see page 14817, left col. 2nd para).
Regarding claim 35, the rejection of claim 32 is incorporated herein.
Zamir et al. in the combination further teach wherein the deep learning model comprises a supervised attention module (SAM) (see page 14817, left col. 2nd para; “A supervised attention module (SAM) is plugged between every two stages to enable progressive learning”). Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention to modify a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images of Guo et al. in view of methods for diagnosing a neurological disorder in a subject of Parsey et al. and Multi-Stage Progressive Image Restoration of Zamir et al. in order to help propagating multi-scale contextualized features from the earlier to later stages (see page 14817, left col. 2nd para).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to WINTA GEBRESLASSIE whose telephone number is (571)272-3475. The examiner can normally be reached Monday-Friday9:00-5:00.
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, Andrew Bee can be reached at 571-270-5180. 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.
/WINTA GEBRESLASSIE/Examiner, Art Unit 2677
Prosecution Insights