DETAILED ACTION
In response to the amendment filed on 06/05/2026, all the amendments to the claims have been entered and the action follows:
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 § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 20-22 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding claim 20:
i) The limitation “having an increased capacity” renders the claim indefinite. It is unclear and confusing a) what the increased capacity is relative to, and b) whether such capacity is actually quantified and compared directly to show the increase. Please amend the claim for verification.
ii) The limitation “to enable reconstruction” renders the claim indefinite. It is unclear and confusing whether the reconstruction is actually performed as part of the claimed invention.
Regarding claim 22:
iii) The limitation “neural network building blocks” renders the claim indefinite. It is unclear and confusing exactly what constitutes a neural network building block, as there are many components to even a well-known neural network.
For example, a well-known structure of a neural network can be described as interconnected artificial neurons forming an input layer, hidden layer, and output layer. Each of these layers can also be broken down into smaller layers (e.g., the output layer alone includes pooling layers, fully connected layers, normalization layers, etc.). Which of these layers would be equivalent to “neural network building blocks”? The applicant’s specification does not provide sufficient enough details to clearly and precisely define the metes and bounds of the claimed invention.
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 1-8, 12, 16, and 25 are rejected under 35 U.S.C. 103 as being unpatentable over Langoju et al. (US 2023/0177747) in view of Lebel (US 2020/0126190).
Regarding claim 1, Langoju discloses:
a memory configured to store machine executable instructions and a super resolution neural network (see para [73], a neural network 314),
wherein the super resolution neural network is configured to receive an initial magnetic resonance image descriptive of a subject (see para [52] and [73], the neural network receives an initial MRI 104 of a patient),
wherein the initial magnetic resonance image has a first resolution and contains an image distortion artifact (see para [73], wherein the initial MRI 104 has unenhanced resolution and includes visual noise),
wherein the super resolution neural network is configured to output an enhanced magnetic resonance image descriptive of the subject in response to receiving the initial magnetic resonance image, wherein the enhanced magnetic resonance image has a second resolution and has a reduction or removal of the image distortion artifact, wherein the second resolution is higher than the first resolution (see para [73], the neural network 314 outputs an enhanced MRI 202 in response to receiving the initial MRI 104, wherein the enhanced MRI 202 has enhanced resolution and includes reduced visual noise);
a computational system (see fig 2, a computer system of a processor 106 and memory 108), wherein execution of the machine executable instructions causes the computational system to:
receive the initial magnetic resonance image (see para [52], [61], and fig 2, the computer system receives the initial MRI 104 from a scanner); and
receive the enhanced magnetic resonance image in response to inputting the initial magnetic resonance image into the super resolution neural network (see para [61] and fig 2, the computer system receives the enhanced MRI 202 from the neural network 314; and see para [136], the neural network 314 may be remotely located).
However, Langoju does not disclose: wherein the image distortion artifact is a Gibbs ringing image artifact; and wherein: the super resolution neural network is trained according to a supervised learning method; and the supervised learning method comprises using a set of ground truth images that include photographic images (i.e., Langoju discloses denoising and enhancing resolution of an initial MRI via a trained neural network, however, does not specify that the trained neural network is trained based on photographic images to eliminate Gibbs artifacts as part of the denoising).
In a similar field of endeavor of denoising and enhancing resolution of an initial MRI, Lebel discloses:
wherein the image distortion artifact is a Gibbs ringing image artifact (see para [27] and [36]-[40], denoising and enhancing resolution of an initial MRI, wherein the denoising includes eliminating Gibbs ringing by a neural network); and
wherein: the super resolution neural network is trained according to a supervised learning method; and the supervised learning method comprises using a set of ground truth images that include photographic images (see para [39] and [49], the neural network is trained using a set of MRI images with a ground truth label of “without Gibbs ringing”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Langoju with Lebel, and denoise and enhance resolution of an initial MRI, as disclosed by Langoju, wherein denoising includes eliminating Gibbs ringing trained using MRI images without Gibbs ringing, as disclosed by Lebel, for the purpose of improving image quality (see Lebel para [2]).
Regarding claim 2, Langoju further discloses: wherein the super resolution neural network is configured to reconstruct the enhanced magnetic resonance image using, at least in part, information contained in the image distortion artifact (see para [87], the neural network 314 is part of an overall MRI reconstruction using the enhanced MRI, which is generated using noise in the initial MRI 104).
Regarding claim 3, Lebel further disclose: wherein image distortion artifact comprises information descriptive of the subject (see para [47], denoising and enhancing resolution of the initial MRI, wherein the denoising includes eliminating noise descriptive of motion of the patient).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Langoju with Lebel, and denoise and enhance resolution of an initial MRI, as disclosed by Langoju, wherein denoising includes eliminating noise descriptive of motion of the patient, as disclosed by Lebel, for the purpose of improving image quality (see Lebel para [2]).
Regarding claim 4, Lebel further discloses:
wherein the memory further stores an image filter module configured to remove random image errors from images with the first resolution (see para [27] and [36], denoising the initial MRI is performed by a denoising neural network prior to performing resolution enhancement separately, wherein the denoising includes eliminating random noise),
wherein execution of the machine executable instructions further causes the computational system to remove random image errors from the initial magnetic resonance image by inputting the initial magnetic resonance image into the image filter module before inputting the initial magnetic resonance image into the super resolution neural network (see para [27] and [36], denoising the initial MRI by the denoising neural network, wherein the denoising includes eliminating random noise by inputting the initial MRI into the neural network prior to performing resolution enhancement separately).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Langoju with Lebel, and denoise and enhance resolution of an initial MRI, as disclosed by Langoju, wherein denoising the initial MRI includes eliminating random noise, as disclosed by Lebel, for the purpose of improving image quality (see Lebel para [2]).
Regarding claim 5, Langoju and Lebel further disclose: wherein the image filter module is an algorithmic image filter module to denoise images with the first resolution (see rejection of claim 4, the denoising neural network denoising the initial MRI prior to performing resolution enhancement separately).
Regarding claim 6, Langoju and Lebel further discloses: wherein the denoising filter module is an image filtering neural network configured to remove the random image errors from images with the first resolution, wherein the image filtering neural network is a denoising neural network (see rejection of claim 4, the denoising neural network denoising the initial MRI prior to performing resolution enhancement separately).
Regarding claim 7, Lebel further discloses:
a magnetic resonance imaging system, wherein the memory further contains pulse sequence commands configured to control the magnetic resonance imaging system to acquire k-space data (see para [26], pulse sequence sampling of k-space), wherein execution of the machine executable instructions further causes the computational system to:
acquire the k-space data by controlling the magnetic resonance imaging system with the pulse sequence commands (see para [26], pulse sequence sampling of k-space); and
reconstruct the initial magnetic resonance image from the k-space data (see para [15] and [47], reconstructing the initial MRI by Fourier transforming the k-space data).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Langoju with Lebel, and denoise and enhance resolution of an initial MRI, as disclosed by Langoju, wherein the initial MRI is captured by pulse sequence sampling of k-space data and includes Gibbs ringing to be removed, as disclosed by Lebel, for the purpose of improving image quality (see Lebel para [2]).
Regarding claim 8, Lebel further discloses: wherein reconstruction of the initial magnetic resonance image causes the image distortion artifact (see para [39], Gibbs ringing is caused by processing of the k-space data).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Langoju with Lebel, and denoise and enhance resolution of an initial MRI, as disclosed by Langoju, wherein the initial MRI is captured by pulse sequence sampling of k-space data and includes Gibbs ringing to be removed, as disclosed by Lebel, for the purpose of improving image quality (see Lebel para [2]).
Regarding claims 12 and 16, Langoju and Lebel disclose everything claimed as applied above (see rejection of claim 1).
Regarding claim 25, Langoju and Lebel disclose everything claimed as applied above (see rejection of claim 4, eliminating random noise).
Claims 20-22 are rejected under 35 U.S.C. 103 as being unpatentable over Langoju and Lebel in view of Wang et al. (US 2020/0085290).
Regarding claim 20, Langoju further discloses: wherein the super resolution neural network is structured according to a predetermined super resolution neural network architecture having a capacity to enable reconstruction of the enhanced magnetic resonance image using information contained in the image distortion artifact (see para [87], the neural network is part of an overall MRI reconstruction using the enhanced MRI, which is generated using noise in the initial MRI 104; and see para [34], a neural network with any suitable number of layers/neurons is used).
However Langoju and Lebel do not disclose: having an increased capacity (i.e., although Langoju discloses flexibility in the number of layers/neurons, Langoju does not disclose increasing the number of layers increases the capacity of the neural network).
In a similar field of endeavor of employing a neural network, Wang discloses: having an increased capacity to enable reconstruction of the enhanced magnetic resonance image using information contained in the image distortion artifact (see para [91], increasing the number of convolutional layers increases the performance of the neural network).
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention to combine Langoju and Lebel, and employ a neural network to denoise an MRI, as disclosed by Langoju and Lebel, while increasing the number of convolutional layers in the neural network, as disclosed by Wang, for the purpose of increasing its performance (see Wang para [91]).
Regarding claim 21, Langoju, Lebel, and Wang further disclose: wherein the increased capacity is provided by increasing a number of kernels per convolutional layer, increasing a number of convolutional layers, or both (see rejection of claim 20, increasing the number of convolutional layers).
Regarding claim 22, Langoju, Lebel, and Wang disclose: wherein the predetermined super resolution neural network architecture comprises neural network building blocks, and wherein the increased capacity is provided by an additional repetition of one or more of the neural network building blocks (see rejection of claim 20, increasing the number of convolutional layers, thereby repeating convolutional layers).
Claim 24 is rejected under 35 U.S.C. 103 as being unpatentable over Langoju and Lebel in view of Venkatesan et al. (US 2020/0400767).
Regarding claim 24, Langoju and Lebel disclose everything claimed as applied above (see rejection of claim 7), however, does not disclose: wherein the k-space data is acquired from an accelerated scan using SENSE, Compressed SENSE, or HyperSense.
In a similar field of endeavor of obtaining an MRI, Venkatesan discloses: wherein the k-space data is acquired from an accelerated scan using SENSE, Compressed SENSE, or HyperSense (see para [45]-[46], Compressed Sensing or HyperSense).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Langoju and Lebel with Venkatesan, and obtain the k-space data for MRI reconstruction, as disclosed by Langoju and Lebel, via Compressed Sensing or HyperSense, as disclosed by Venkatesan, for the purpose of increasing efficiency (see Venkatesan para [46]).
Allowable Subject Matter
Claims 10, 14, 18, 23, and 26 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims. The following is a statement of reasons for the indication of allowable subject matter:
Regarding claim 10, Langoju further discloses: wherein the super resolution neural network is trained using a supervised learning method that comprises: receiving a set of ground truth images with the second resolution (see para [74], ground-truth images used to train the neural network 314 represent how a denoised and resolution enhanced MRI is supposed to look).
However, Langoju does not disclose: constructing a set of trial images from the set of ground truth images, wherein constructing the set of trial images comprises: providing a set of down sampled images by down sampling the ground truth images to the first resolution; providing a set of k-space representation of the set of down sampled images by transforming the set of down sampled ground truth images to k-space; providing a set of modified k-space representations by truncating by cropping the set of k-space representation of the set of down sampled images; and provide the set of trial images by transforming the set of modified k-space representations to image space; and training the super resolution neural network with the set of ground truth images and the set of trial images. Similar reasons apply to claims 14 and 18.
Regarding claim 23, Langoju and Lebel disclose everything claimed as applied above (see rejection of claim 12), however, do not disclose: removing random image errors from the initial magnetic resonance image suing an image filter module before inputting the initial magnetic resonance image into the super resolution neural network (i.e., Langoju and Lebel disclose eliminating random noises by the neural network, however, do not disclose a separate module performing it before inputting into the neural network). The applicant has also provided an explicit definition to the claimed “random image errors” in specification p4 lines 32-34, which has been considered. Similar reasons apply to claim 26 (i.e., Langoju and Lebel disclose denoising by the neural network, however, do not disclose a denoising step employing a separate neural network).
Response to Arguments
Arguments regarding 101
In view of the claim amendments, the rejection under 101 has been withdrawn.
Arguments regarding 112(b)
In view of the claim amendments and arguments, the 112(b) rejection has been withdrawn for claim 2. However, newly added claims 20-22 are now rejected under 112(b).
Arguments regarding prior art
Regarding claim 1, the applicant argues that Langoju and Lebel do not disclose the subject matter of the claim, specifically because Lebel discloses using medical images to train a neural network instead of using “photographic images”.
The examiner respectfully disagrees. Following are references that refer to an MRI as a photograph:
Lee et al. (US 2013/0060129): see abstract, “MRI photographing a patient”
Hirano et al. (US 6,687,533): see col 1 lines 47-48, “(MRI) photography”
Utsumi et al. (US 2016/0116562): see para [26], “MRI device for photographing”
Miyashita et al. (US 6,522,908): see col 5 47, “photographed by an MRI device”
It is clear, from the above references, that the term “photographic images” do not exclude MRIs, which contradicts with the applicant’s argument. Lebel further discloses that the MRIs used to train the neural network include ground-truth labels (i.e., “image with Gibbs ringing” and “image without Gibbs ringing” are ground-truth labels). Therefore, Lebel discloses training the neural network using ground-truth labeled photographic images, which reads on the broadest reasonable interpretation of the claimed “using a set of ground truth images that include photographic images”. Similar reasons apply to claims 12 and 16.
No particular arguments have been received regarding newly added claims 20-26.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SJ PARK whose telephone number is (571)270-3569. The examiner can normally be reached M-F 8:00 AM - 5:00 PM.
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, EMILY TERRELL can be reached at 571-270-3717. 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.
/SJ Park/Primary Examiner, Art Unit 2675
Prosecution Insights