DETAILED ACTION
Response to Amendment
Applicant’s preliminary amendments filed September 27, 2024, and March 17, 2025, have been entered in full. Claims 1-2, 4-8, 10-12, 14-15, 18-21, and 24-26 are pending.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on September 27, 2024, and March 9, 2026, are being considered by the examiner.
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.
Claim(s) 1 and 12 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by ‘Li’ (“Quantitative phase imaging via a cGAN network with dual intensity images under centrosymmetric illumination,” 2019).
Regarding claim 1, Li discloses a quantitative phase imaging method comprising:
imaging a sample to obtain one or more oblique illumination raw captures (e.g., Figure 3, imaging with sequential oblique illumination from LED array; Also see, e.g., the paragraph below equation 5);
inputting one or more of the oblique illumination raw captures into a deep learning neural network (DLNN) (e.g., Fig. 1, input at top-left; The DLNN is the cGAN); and
generating, using the DLNN, a quantitative phase image of the sample based on the inputted oblique illumination raw captures (e.g., Fig. 1, Generated phase image output from decoder of generator).
Regarding claim 12, Li discloses the method of claim 1, wherein the sample comprises at least one of blood tissue (e.g., Page 2881, right column, center, blood smear slide shown in Figs. 4(a1)-4(d1)) or brain tissue.
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.
Claim(s) 1-2, 4-8, 11-12, 14-15, 18-21, and 24-26 is/are rejected under 35 U.S.C. 103 as being unpatentable over ‘Ledwig’ (“Epi-mode tomographic quantitative phase imaging in thick scattering samples,” 2019; cited and copy provided in IDS filed 09 March 2026) in view of Li.
Regarding claim 1, Examiner notes that the claim recites a method that is generic to the method performed by the system of claim 14. The system of claim 14 is obvious over Ledwig in view of Li (see below). Accordingly, claim 1 is also rejected under 35 U.S.C. 103 as being unpatentable over Ledwig in view of Li for substantially the same reasons as claim 14.
Regarding claim 2, Examiner notes that the claim recites a method that is generic to the method performed by the system of claim 15. The system of claim 15 is obvious over Ledwig in view of Li (see below). Accordingly, claim 2 is also rejected under 35 U.S.C. 103 as being unpatentable over Ledwig in view of Li for substantially the same reasons as claim 15.
Regarding claim 4, Examiner notes that the claim recites a method that is generic to the method performed by the system of claim 15. The system of claim 15 is obvious over Ledwig in view of Li (see below). Accordingly, claim 4 is also rejected under 35 U.S.C. 103 as being unpatentable over Ledwig in view of Li for substantially the same reasons as claim 15.
Regarding claim 5, Ledwig in view of Li teaches the method of claim 2, and Li further teaches training the DLNN (e.g., Pg. 2881, 1st full par.).
Regarding claim 6, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 18. Ledwig in view of Li teaches the invention of claim 18 (see below). Accordingly, claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over Ledwig in view of Li for substantially the same reasons as explained in the rejection of claim 18.
Regarding claim 7, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 19. Ledwig in view of Li teaches the invention of claim 19 (see below). Accordingly, claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Ledwig in view of Li for substantially the same reasons as explained in the rejection of claim 19.
Regarding claim 8, Examiner notes that the claim recites limitations that are substantially the same as limitations recited in claim 20. Ledwig in view of Li teaches the invention of claim 20 (see below). Accordingly, claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Ledwig in view of Li for substantially the same reasons as explained in the rejection of claim 20.
Regarding claim 11, Ledwig in view of Li teaches the method of claim 5, and Li further teaches that training the DLNN comprises training the DLNN to obtain the quantitative phase image from the first oblique illumination raw capture and the second oblique illumination raw capture using a training data set to create a trained neural network (e.g., Pg. 2881, 1st full par., “We train our cGAN … using 1301 intensity pair and the corresponding ground truth phase”; This is a training dataset and it is used to train the cGAN DLNN to accept first and second oblique illumination raw captures – i.e., the intensity pair – and output a quantitative phase image; Also see, e.g., Fig. 1).
Regarding claim 12, Examiner notes that the claim recites a method that is generic to the method performed by the system of claim 24. The system of claim 24 is obvious over Ledwig in view of Li (see below). Accordingly, claim 12 is also rejected under 35 U.S.C. 103 as being unpatentable over Ledwig in view of Li for substantially the same reasons as claim 24.
Regarding claim 14, Ledwig teaches a quantitative phase imaging system (e.g., Figure 1a) comprising:
an imager (e.g., Fig. 1, camera) configured to take one or more quantitative oblique back-illumination microscopy (qOBM) raw captures of a sample (e.g., Fig. 1, four qOBM images are captured, one for each sequential oblique-angle back-illumination via LEDs);
a processing resource (see Note Regarding Computer Hardware below); and
a non-transitory computer-readable medium, coupled to the processing resource, having stored therein instructions that when executed by the processing resource cause the processing resource to (see Note Regarding Computer Hardware below):
generate, based on one or more of the qOBM raw images, a quantitative phase image of the sample (e.g., Fig. 2g, generated qOBM quantitative phase image) using a deep learning neural network (DLNN) trained to obtain the quantitative phase image (see Note Regarding DLNN below) using incoherent illumination from one or more of the qOBM raw images (e.g., Fig. 1a, Section 2.1, LED illumination is used, which is incoherent).
Note Regarding Computer Hardware. Ledwig’s teachings certainly imply the use of computer hardware to generate its quantitative phase images. For example, Ledwig teaches using a digital camera, custom software, and a data acquisition block (e.g., paragraph above Sec. 2.1.1). Nevertheless, Ledwig’s disclosure focuses on its image processing algorithms and does not provide a detailed explanation of what computer hardware is used to implement those algorithms. In particular, Ledwig does not explicitly teach its image processing algorithms being implemented using a processing resource and a non-transitory computer-readable medium, coupled to the processing resource, having stored therein instructions that when executed by the processing resource cause the processing resource to perform the algorithms.
However, Examiner takes Official Notice that it is old and well-known in the art of image analysis to implement image processing algorithms using computer hardware including a processing resource and a non-transitory computer-readable medium, coupled to the processing resource, having stored therein instructions that when executed by the processing resource cause the processing resource to perform the algorithms. Use of a processing resource (e.g., a processor, CPU, GPU, etc.) advantageously allows processing of digital data and fast and efficient execution of image processing algorithms. Use of a non-transitory computer-readable medium (e.g., a memory) advantageously allows computer code instructions for performing the image processing algorithms to be made accessible for execution in the processor and to be preserved over time for continued and/or repetitive performance of the image processing algorithms.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to implement the image processing algorithms of Ledwig using computer hardware including a processing resource and a non-transitory computer-readable medium, coupled to the processing resource, having stored therein instructions that when executed by the processing resource cause the processing resource to perform the algorithms with the reasonable expectation that this would result in a system that could perform the image processing applications quickly and efficiently over time.
Note Regarding DLNN. Ledwig builds an analytical model of image formation and inverts it to generate a quantitative phase image from qOBM raw captures (e.g., Sec. 2, especially the determination of system transfer function at Sec. 2.1.1 and reconstruction of quantitative phase at Sec. 2.1.3).
Ledwig does not explicitly teach using a trained deep learning neural network (DLNN) to generate the quantitative phase image.
Li recognizes that building an image formation model and inverting it (i.e., the type of approach used by Ledwig) is a common approach to generating quantitative phase images (e.g., Page 2879, right column, middle paragraph). However, Li teaches that “Another solution to tackle such inverse problems is to use a deep learning network” (Id.). Li teaches a specific example, where images of a sample are captured under oblique illumination (e.g., Fig. 3; e.g., Pg. 2881, below equation 5) and input to a trained DLNN (e.g., Fig. 1, especially the Generator; e.g., Pg. 2881, 1st full par., describes the training) to obtain a quantitative phase image (e.g., Fig. 1, Generated phase image).
Li teaches that “The deep learning approach has found important applications in microscopy” and that the cGAN DLNN architecture it uses “has been used for many pixel-to-pixel mapping applications (Pg. 2879, rt. col., middle par.). Li further teaches that “the cGAN is able to directly learn the mapping relationship from the intensity pair to the targeted phase distribution” (Pg. 2879, last par.). Li also teaches that its DLNN-based approach is advantageous because the network performs well on both simulated and experimental data and it has an advantage in processing speed (Pg. 2882, last par.).
As noted above, Li suggests using a DLNN as an alternative to the types of analytical inverse problem solutions used by Ledwig. One of ordinary skill in the art would have had a reasonable expectation of success given Li’s success using a DLNN for quantitative phase image generation and Li’s recognition that DLNNs have been used successfully in microscopy and image mapping applications.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to modify the system of Ledwig as applied above with the DLNN-based mapping of Li in order to improve system with the reasonable expectation that this would result in a system that could directly learn a mapping relationship, without requiring an analytical model, and could generate quantitative phase images with good performance and an advantage in processing speed. This technique for improving the system of Ledwig was within the ordinary ability of one of ordinary skill in the art based on the teachings of Li.
Therefore, it would have been obvious to one of ordinary skill in the art to combine the teachings of Ledwig and Li to obtain the invention as specified in claim 14.
Regarding claim 15, Ledwig in view of Li teaches the system of claim 14, wherein two of the qOBM raw captures comprise a first qOBM raw capture taken at a first wavelength and a second qOBM raw capture taken at a second wavelength (e.g., par. spanning pages 3607-3608, red and green wavelengths; e.g., Fig. 2a-b) and orthogonal to the first qOBM raw capture (e.g., Pg. 3607, Fig. 1 and par. below caption, red and green illumination are “positioned perpendicularly to one another (Fig. 1(b)) to produce images with orthogonal shear directions”).
Regarding claim 18, Ledwig in view of Li teaches the system of claim 14, and Li further teaches that the DLNN comprises a generative adversarial network (GAN) (e.g., Fig. 1).
Regarding claim 19, Ledwig in view of Li teaches the system of claim 18, and Li further teaches that the GAN is an independent U-Net GAN (e.g., Fig. 1, Generator of GAN has a “U” shape of progressively down-scaling encoders, corresponding progressively upscaling decoders, and skip connections between; For at least these reasons the GAN is within the scope of being a “U-Net GAN”; Furthermore, the GAN can be considered “independent” at least because it is only trained to generate phase images, and thus is independent of other tasks; Additionally, or alternatively, the U-Net encoder in the GAN is distinct and independent from the discriminator component of the GAN).
Regarding claim 20, Ledwig in view of Li teaches the system of claim 18, and Li further teaches that:
the GAN comprises a discriminator and a generator (Fig. 1);
the generator is configured to create fake training images (e.g., Fig. 1, output from generator given input training images is a “fake” generated phase image); and
the discriminator is configured to classify the fake training images from real images (e.g., Fig. 1, discriminator attempts to discriminate (i.e., classify separately) “fake” phase image from generator and real images from ground truth).
Regarding claim 21, Ledwig in view of Li teaches the system of claim 20, and Li further teaches that the generator comprises 8 encoding layers and 8 decoding layers (e.g., Fig. 1, generator’s encoder and decoder both include 8 blocks, each block including at least one layer; Note the open-ended nature of “comprises”).
Regarding claim 24, Ledwig in view of Li teaches the system of claim 14, and Ledwig further teaches that the sample comprises at least one of blood tissue (e.g., Sec. 3.2) or brain tissue (e.g., Sec. 3.3).
Regarding claim 25, Examiner notes that the claim recites a system that is generic to the system of claim 14 in that it more-generally requires OBM raw captures (instead of qOBM raw captures) and does not require use of incoherent illumination. Ledwig in view of Li teaches the system of claim 14 (see above). Accordingly, claim 25 is also rejected under 35 U.S.C. 103 as being unpatentable over Ledwig in view of Li for substantially the same reasons as claim 14.
Regarding claim 26, Ledwig in view of Li teaches the system of claim 25, and Ledwig further teaches that:
the imager uses incoherent illumination (e.g., Fig. 1, LED illumination; e.g., Page 3617, below equation 14) and comprises:
light sources comprising light-emitting devices (Sec. 2.1 and Fig. 1a, four LEDs); and
an image-capturing device (Sec. 2.1 and Fig. 1a, camera); and
the light sources illuminate the sample sequentially (e.g., Fig. 1, caption) and the oblique back-illumination microscopy (OBM) raw captures are acquired by the image-capturing device (e.g., Fig. 1, caption; Also see pg. 3608, 1st par.).
Claim(s) 10 is/are rejected under 35 U.S.C. 103 as being unpatentable over Li in view of ‘Bhatt’ (“High-resolution single-shot phase-shifting interference microscopy using deep neural network for quantitative phase imaging of biological samples,” 2021).
Regarding claim 10, Li teaches the method of claim 1.
Li teaches training the DLNN to obtain the quantitative image from a pair of oblique illumination raw captures using a training data set to create a trained neural network (e.g., Fig. 1, top-left).
Li does not explicitly describe training the DLNN to obtain the quantitative phase image from a single oblique illumination raw capture.
However, at Page 2881 Li states the following:
PNG
media_image1.png
198
400
media_image1.png
Greyscale
PNG
media_image2.png
194
400
media_image2.png
Greyscale
This statement would have suggested to one of ordinary skill in the art that it would be possible, but challenging, to train the DLNN to obtain the quantitative phase image from a single raw capture.
Bhatt is similar to Li in that it is also focused on obtaining quantitative phase images (QPIs). While Bhatt uses a different imaging modality (WL-PSIM), Bhatt faces a similar problem in that WL-PSIM conventionally requires capturing multiple images (i.e., multiple raw captures) in order to recover a QPI (e.g., Page 2, par. spanning columns; e.g., Abstract). Specifically, multiple raw captures are required in order to introduce systematic phase shifts between them (Page 3, par. spanning columns). Bhatt explains that requiring multiple raw captures is disadvantageous because, for example, it cannot be used to image dynamic scenes (e.g., Page 2, par. spanning columns). Bhatt teaches an approach to solving this problem by training a GAN DLNN to directly obtain a QPI from a single raw image capture (e.g., Sec. 3.2; e.g., Page 3, above Sec. 2.2). Bhatt teaches that its single-capture approach is advantageous because, for example, it reduces data acquisition time (Sec. 4, last par.). Bhatt also teaches that its single-capture approach performs on par with another approach that trains a DLNN to use multiple (a single real and four simulated) raw captures to obtain a QPI (Sec. 4, 2nd-to-last par.).
Taken together, the teachings of Li suggest that single-image QPI reconstruction may be possible, albeit challenging; the teachings of Bhatt provide further motivation to use single-image QPI reconstruction to reduce data acquisition time and enable dynamic scene imaging; and the teachings of Bhatt would provide a reasonable expectation that using a DLNN to generate QPIs from single raw captures, even in imaging modalities that conventionally rely on multiple raw captures to provide phase information, would be successful.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to modify the method of Li with a single-raw-capture-based DLNN in order to improve the method with the reasonable expectation that this would result in a method that could produce quantitative phase images of dynamic scenes and/or with reduced acquisition time. This technique for improving the method of Li was within the ordinary ability of one of ordinary skill in the art based on the teachings of Li and Bhatt.
Therefore, it would have been obvious to one of ordinary skill in the art to combine the teachings of Li and Bhatt to obtain the invention as specified in claim 10.
Conclusion
The following prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
‘Angelopoulos’ (“Image-to-Image Regression with Distribution-Free Uncertainty Quantification and Applications in Imaging,” 10 Feb. 2022)
Secs. 3.2-3.3 describe using a DLNN to take two obliquely illuminated cell intensity images as input and output a quantitative phase image
Any inquiry concerning this communication or earlier communications from the examiner should be directed to GEOFFREY E SUMMERS whose telephone number is (571)272-9915. The examiner can normally be reached Monday-Friday, 7:00 AM to 3:30 PM ET.
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, Chan Park can be reached at (571) 272-7409. 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.
/GEOFFREY E SUMMERS/Examiner, Art Unit 2669
Prosecution Insights