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 .
Disposition of the Claims
In response to applicant’s amendment received on 07/20/2026, all requested changes to the claims have been entered.
Claim(s) 1-24 were previously pending.
No Claim(s) have been added.
No Claim(s) have been cancelled.
Claim(s) 1-24 are currently pending.
Response to Amendment
Claim Interpretations under 35 U.S.C. § 112(f)
The “illumination system” of claim 20 was previously interpreted under 35 U.S.C. § 112(f). Applicant has amended this limitation to now recite “an illumination system including a plurality of light sources”. This amendment provides sufficient structure for the illumination system and no longer invokes 35 U.S.C. § 112(f).
Applicant has also amended claims 15, 17-20 to recite circuitry configured to perform distinct recited functions as opposed to reciting an explicit function. The examiner understands these amendments were performed to avoid potential interpretation under 35 U.S.C. § 112(f).
Claim Rejections under 35 U.S.C. § 112(b)
Applicant has amended claims 9 & 14 to now recite “wherein at least one direction of the plurality of directions corresponds to an angle larger than a maximum acceptance angle associated with a numerical aperture…”. This amendment overcomes the previous rejection under 35 U.S.C. § 112(b) for comparing an angle to a dimensionless value.
Rejections under 35 U.S.C. § 112(b) for claims 9 & 14 have been withdrawn.
Response to Arguments
Claim Rejections under 35 U.S.C. § 101
Applicant’s arguments, see pgs. 11-13, filed 07/20/2026, with respect to claim rejections under 35 U.S.C. § 101 have been fully considered and are persuasive. The rejections of claims 1-3, 7, 8, 10-13, 15-17, 19 & 24 under 35 U.S.C. § 101 have been withdrawn.
Claim rejections under 35 U.S.C. § 102(a)(2) and 35 U.S.C. § 103
Applicant's arguments filed 07/20/2026 have been fully considered but they are not persuasive. Applicant argues that Bokodia (US 2022/0334371 A1) fails to disclose “training the machine learning model to analyze the sample using the training set of digital images” recited in claim 1. Applicant states that Bokodia is directed towards optimizing tunable optical hardware elements for reducing a task-specific loss, which is conducted by simulating optical hardware to generate a simulated output that is provided to the digital-layer of the computational model. Applicant further asserts that this model is previously trained, without disclosure of how the classification model itself is trained. Applicant later states that Bokodia’s training process is directed towards identifying optimal physical hardware settings to converge onto an optimal illumination pattern for a given inference task as opposed to classification of raw images. The examiner respectfully disagrees with this conclusion.
The examiner acknowledges that a key element of Bokodia’s invention is directed towards utilizing machine learning to optimize illumination settings, however, as indicated by the applicant, this optimization is conducted to achieve a particular inference task, such as classification of a malaria parasite [¶0067]. The claim language of “training the machine learning model to analyze the sample using the training set of digital images and the received ground truth”, is particularly broad, only necessitating the use of a training set of digital images and a ground truth to train a machine learning model to analyze a sample.
Turning to Bokodia’s disclosure in reference to Figs. 7A-7D, Bokodia explains that they employ a learning sensing network (LSN) 706 (which may include a convolutional neural network or “CNN”) that is optimized specifically for detecting the presence of infection [¶0062], further stating that the microscope 704 is optimized to sense the presence or absence of a malaria parasite, through task-optimized hardware training [¶0063]. This task is still fundamentally directed towards classifying or otherwise analyzing a sample of blood. As noted in the previous office action, during Network Training (b) of Fig. 7A, variably-illuminated training data (i.e., a training set of digital images) is used for supervised training of LSN 706, wherein the infection status of the training data is already known (indicative of a ground truth) [¶0064-67]. Bokodia further explains that after completion of network training, the inference process (in this case, infection classification) can then be applied using images acquired with low-resolution under optimal illumination conditions [¶0080]. This categorically demonstrates that the disclosure teaches not-only the use of a pretrained model, but an actual model training strategy.
Bokodia, in summary, leverage optimization of tunable hardware elements (in the form of an LED array) to classify cells as infected or not infected using a training dataset with a known status of infection. The tuning of LED settings during training is needed to properly leverage the complex illumination system, which does not teach away from the ultimate function of analyzing samples in the form of infection classification using ground truth training images. Mappings to this limitation have been updated in light of the above arguments in the following claim rejections for clarity of record.
Claim Rejections - 35 USC § 102
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 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)(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.
Claims 1-16, 19-21, & 24 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Bokadia et al (US 2022/0334371 A1).
Regarding claim 1, Bokadia et al (hereinafter referred to as “Bokadia”) disclose a computational imaging system implements a supervised learning algorithm in fourier ptychography to classify cells in microscopy images. More specifically, Bokadia teach A method for training a machine learning model to analyze a sample (the method outlined in Fig. 3 which can be used to classify microscopy samples [¶0051]), the method comprising:
receiving a ground truth including a classification of at least one portion of a sample (the computational algorithm 206 can be a supervised machine learning algorithm [0047], “variably-illuminated training data” of cells classified as “infected” or “not infected” (ground truth images of known classification) are used for Network training (b) [Fig. 7A]);
acquiring a training set of digital images of the sample by illuminating the sample from a plurality of directions and capturing a digital image for each of the plurality of directions (raw data is acquired for machine learning [¶0051; Fig. 3], the light source 506, may be an LED array, which can illuminate the sample (in this case, blood slide 508) from a plurality of directions to acquire training data [¶0057-58; Fig. 5], images are captured for each of the unique illumination conditions [¶0067]); and
training the machine learning model to analyze the sample using the training set of digital images and the received ground truth (the learning sensing network 706 (which may include a convolutional neural network (CNN)) is trained for an inference task (in the form of image classification of blood cells infected with a malaria parasite) during network training (b), wherein variably-illuminated training data of a known infection status (i.e. ground truth) is provided to the LSN to optimize LED illumination schemes for accurate classification using low-resolution images [¶0061-67; 80; Fig. 7A-D]).
Regarding claim 2, Bokodia teach The method according to claim 1 (as described above), wherein the sample is an unstained sample (The optical hardware system 110 and visual detector system 120, may be configured for multiple imaging modalities, like brightfield imaging [¶0033; Fig. 1] – the examiner notes that brightfield imaging inherently can be performed without the use of any stains – additionally disclosed system 802 utilizes tomographic imaging which eliminates the need for staining [¶0104]).
Regarding claim 3, Bokodia teach The method according to claim 1 (as described previously), wherein the training set is acquired by illuminating the sample with white light (The optical hardware system 110 and visual detector system 120, may be configured for multiple imaging modalities, like brightfield imaging [¶0033; Fig. 1] – the examiner notes that brightfield imaging inherently uses white light for illumination) from the plurality of directions and capturing a digital image for each of the plurality of directions (the sample is illuminated from various angles and a corresponding image (812) is captured for each angle [¶0095; Fig. 8]).
Regarding claim 4, Bokodia teach The method according to claim 1 (as described previously), wherein the machine learning model is a convolutional neural network (the computational algorithm 206 may include a convolutional neural network (CNN) [¶0047]).
Regarding claim 5, Bokodia teach The method according to claim 1 (as described above), wherein the ground truth further includes a position of the at least one portion in the sample (the imaging system 100 can identify and capture images of regions of interest (which denote positional information) in the sample and can be used to improve the accuracy of automatic computational decisions for local inferencing [¶0053], which can be used to inform the task 301 of the computational algorithm 206 [¶0051]), and wherein the step of training the machine learning model further includes:
training the machine learning model using the training set of digital images and the received ground truth until a difference between a position output of the machine learning model is smaller than an additional predetermined threshold, thereby training the machine learning model to determine a position of the at least one portion in the sample (the imaging system 100 can identify and capture images of regions of interest (which denote positional information) in the sample and can be used to improve the accuracy of automatic computational decisions for local inferencing [¶0053], which can be used to inform the task 301 (e.g. classification of malaria infection [¶0067]of the computational algorithm 206, which is optimized via minimization of loss function 300 [¶0051]).
Regarding claim 6, Bokodia teach The method according to claim 1, wherein the ground truth further includes dimensions of the at least one portion in the sample (parameters for each cell, such as volume, perimeter, diameter, height, etc., can be simultaneously calculated [¶0099], which can be used to inform the task 301 of the computational algorithm 206 [¶0051]), and wherein the step of training the machine learning model further includes:
training the machine learning model using the training set of digital images and the received ground truth until a difference between a dimensions output of the machine learning model is smaller than a predetermined dimensions threshold, thereby training the machine learning model to determine dimensions of the at least one portion in the sample (parameters for each cell, such as volume, perimeter, diameter, height, etc., can be simultaneously calculated [¶0099], which can be used to inform the task 301 of the computational algorithm 206, which is optimized via minimization of loss function 300 [¶0051]).
Regarding claim 7, Bokodia teach The method according to claim 1 (as described previously), wherein the ground truth includes a respective classification of a plurality of portions of the sample (“variably-illuminated training data” of cells classified as “infected” or “not infected” (ground truth images of known classification) are used for Network training (b) [Fig. 7A]).
Regarding claim 8, Bokodia teach The method according to claim 7 (as described previously), wherein the ground truth further includes a respective position of the plurality of portions in the sample (“variably-illuminated training data” of cells classified as “infected” or “not infected” (ground truth images of known classification) are used for Network training (b) [Fig. 7A], the imaging system 100 can also identify and capture images of regions of interest (which denote positional information) in the sample and can be used to improve the accuracy of automatic computational decisions for local inferencing based on the training data [¶0053]).
Regarding claim 9, Bokodia teach The method according to claim 1 (as described previously), wherein the training set of digital images is acquired using a microscope objective (low-magnification objective 806 of system 802 [Fig. 8]) and an image sensor (camera 810 of system 802 [Fig. 8]), and wherein at least one direction of the plurality of directions corresponds to an angle larger than a maximum acceptance angle associated with a numerical aperture of the microscope objective (system 802 performs Fourier ptychographic tomography [¶0095-96; Fig. 8] – the examiner notes that fourier ptychography inherently captures images from an illumination angle larger than the numerical aperture of the objective. This is evidenced by Guo et al. (“Fourier Ptychography for Brightfield, Phase, Darkfield, Reflective, Multi-Slice, and Fluorescence Imaging”, 2015), who disclose that in LED array illumination for Fourier ptychography, the maximum illumination angle is higher than that of the collection numerical aperture [Section IIA – LED Array Illumination; ¶01]. Fourier ptychography takes advantage of Fourier reconstruction to create high-resolution images without being limited by the usual constraints of the numerical aperture [Section IIA – LED Array Illumination; ¶03-5]).
Regarding claim 10, Bokodia teach A method for analyzing a sample (the disclosed system can be configured for different types of analyses [¶0046], with an exemplary analysis method for blood sample analysis outlined in Fig. 6 [0058]), the method comprising:
receiving an input set of digital images of the sample, wherein the input set of digital images is acquired by illuminating the sample from a plurality of directions and capturing a digital image for each of the plurality of directions (the light source 506, may be an LED array, which can illuminate the sample (in this case, blood slide 508) from a plurality of directions [¶0057-58; Fig. 5], images are captured for each of the unique illumination conditions [¶0067]); analyzing the sample by:
inputting the input set of digital images into a machine learning model being trained (raw data is acquired for machine learning [¶0051; Fig. 3]) according to the method of claim 1 (as described previously), and
receiving, from the machine learning model, an output including a classification of at least one portion of the sample (the process of Fig. 3 can be used to train the computational algorithm 206 to output a classification of each cell in the sample [¶0051], such as classifying cells based on whether they exhibit signs of an infection on a cell-by-cell basis [¶0067; Fig. 7A & Fig. 7D]).
Regarding claim 11, Bokodia teach The method according to claim 10 (as described above), wherein the sample is an unstained sample (The optical hardware system 110 and visual detector system 120, may be configured for multiple imaging modalities, like brightfield imaging [¶0033; Fig. 1] – the examiner notes that brightfield imaging inherently can be performed without the use of any stains – additionally disclosed system 802 utilizes tomographic imaging which eliminates the need for staining [¶0104]).
Regarding claim 12, Bokodia teach The method according to claim 10 (as described previously), wherein the input set of digital images is acquired by illuminating the sample with white light (The optical hardware system 110 and visual detector system 120, may be configured for multiple imaging modalities, like brightfield imaging [¶0033; Fig. 1] – the examiner notes that brightfield imaging inherently uses white light for illumination) from the plurality of directions and capturing a digital image for each of the plurality of directions (the sample is illuminated from various angles and a corresponding image (812) is captured for each angle [¶0095; Fig. 8]).
Regarding claim 13, Bokodia teach The method according to claim 10 (as described previously), wherein the output further includes a position of the at least one portion in the sample (the imaging system 100 can identify and capture images of regions of interest (which denote positional information [¶0053]).
Regarding claim 14, Bokodia teach The method according to claim 10 (as described previously), wherein the input set of digital images of the sample is acquired using a microscope objective (low-magnification objective 806 of system 802 [Fig. 8]) and an image sensor (camera 810 of system 802 [Fig. 8]), and wherein at least one direction of the plurality of directions corresponds to an angle larger than a maximum acceptance angle associated with a numerical aperture of the microscope objective (system 802 performs Fourier ptychographic tomography [¶0095-96; Fig. 8] – the examiner notes that Fourier ptychography inherently captures images from an illumination angle larger than the numerical aperture of the objective. This is evidenced by Guo et al. (“Fourier Ptychography for Brightfield, Phase, Darkfield, Reflective, Multi-Slice, and Fluorescence Imaging”, 2015), who disclose that in LED array illumination for Fourier ptychography, the maximum illumination angle is higher than that of the collection numerical aperture [Section IIA – LED Array Illumination; ¶01]. Fourier ptychography takes advantage of Fourier reconstruction to create high-resolution images without being limited by the usual constraints of the numerical aperture [Section IIA – LED Array Illumination; ¶03-5]).
Regarding claim 15, Bokodia teach A device for training a machine learning model (the automated imaging system 100 of Fig. 1, which is used to train the computational algorithm 206 [¶0047; Fig. 2]) comprising circuitry (computing system 130, comprising CPU 132, storage 140, and memory 134 [¶0034; Fig. 1]) configured to:
receive a training set of digital images (the optical hardware model 302 receives raw data for the loss function training set [¶0051-52; Fig. 3]), wherein the training set of digital images is acquired by illuminating a sample from a plurality of directions and capturing a digital image of at least one portion of the sample for each of the plurality of directions (raw data is acquired for machine learning [¶0051; Fig. 3], the light source 506, may be an LED array, which can illuminate the sample (in this case, blood slide 508) from a plurality of directions to acquire training data [¶0057-58; Fig. 5], images are captured for each of the unique illumination conditions [¶0067]);
receive a ground truth (the computational algorithm 206 can be a supervised machine learning algorithm [0047], “variably-illuminated training data” of cells classified as “infected” or “not infected” (ground truth images of known classification) are used for Network training (b) [Fig. 7A]); and
train a machine learning model (the loss function 300 is used to train computational algorithm 206 [¶0051; Fig. 3]) according to the method of claim 1 (as described previously) using the received ground truth and the acquired training set of digital images (the computational model 206 is trained via minimization of the loss function 300 for a classification task [¶0051; Fig. 3], “variably-illuminated training data” of cells classified as “infected” or “not infected” (ground truth images of known classification) are used for Network training (b) [Fig. 7A]).
Regarding claim 16, Bokodia teach The device according to claim 15 (as described above), wherein the training set is acquired by illuminating the sample with white light (The optical hardware system 110 and visual detector system 120, may be configured for multiple imaging modalities, like brightfield imaging [¶0033; Fig. 1] – the examiner notes that brightfield imaging inherently uses white light for illumination) from the plurality of directions and capturing a digital image for each of the plurality of directions (the sample is illuminated from various angles and a corresponding image (812) is captured for each angle [¶0095; Fig. 8]).
Regarding claim 19, Bokodia teach The device according to claim 17 (as described previously), wherein the circuitry (computing system 130, comprising CPU 132, storage 140, and memory 134 [¶0034; Fig. 1]) is further configured to:
determine a position of the at least one portion of the sample by determining a position of a corresponding portion in the high-resolution digital image of the sample (the imaging system 100 can identify and capture images of regions of interest (which denote positional information) in the sample which can be further imaged at a higher at a higher resolution (40× magnification) [¶0053]); and wherein the ground truth further comprises the determined position of the at least one portion in the sample (the imaging system 100 can identify and capture images of regions of interest (which denote positional information) in the sample and can be used to improve the accuracy of automatic computational decisions for local inferencing [¶0053], which can be used to inform the task 301 of the computational algorithm 206, which is optimized via minimization of loss function 300 [¶0051]).
Regarding claim 20, Bokodia teach A microscope system (microscope system 802 for Fourier ptychographic diffraction tomography [¶0091; Fig. 8]) comprising:
an illumination system including a plurality of light sources (the LED array 804 is quasi-hemispherical [¶0096; Fig. 8]) configured to illuminate a sample from a plurality of directions (provides 360° illumination of the sample [¶0096]);
an image sensor (camera 810 [¶Fig. 8]);
at least one microscope objective arranged to image the sample onto the image sensor (low magnification objective 806 [Fig. 8]); and
circuitry (computing system 130, comprising CPU 132, storage 140, and memory 134 [¶0034]) configured to:
control the illumination system to sequentially illuminate the sample from the plurality of directions (the samples are illuminated across various angles [¶0096] – the examiner notes that sequentially illuminating a sample and capturing an image for each angle is inherent to Fourier ptychography),
control the image sensor to acquire an input set of digital images (the camera 810 captures images across a 360° illumination of the sample [¶0096; Fig. 8]), wherein the input set of digital images is acquired by illuminating the sample from a plurality of directions and capturing a digital image for each of the plurality of directions (the camera 810 captures images across a 360° illumination of the sample [¶0096; Fig. 8]), and
input the input set of digital images into a machine learning model being trained (after network training, only one to two low-resolution images fed into the network are needed for classification [¶0080]) according to the method of claim 1 (as described previously), and receive, from the machine learning model, an output including a classification of at least one portion of the sample into at least one class (the process of Fig. 3 can be used to train the computational algorithm 206 to output a classification of each cell in the sample [¶0051], such as classifying cells based on whether they exhibit signs of an infection on a cell-by-cell basis [¶0067; Fig. 7A & Fig. 7D]).
Regarding claim 21, Bokodia teach The microscope system according to claim 20 (as described above), wherein each light source of the plurality of light sources is configured to emit white light (light source 506, may be an LED array, which can illuminate the sample (in this case, blood slide 508) from a plurality of directions [¶0057-58; Fig. 5]).
Regarding claim 24, Bokodia teach A non-transitory computer-readable storage medium (the computing system 130 may include a storage device 140 [¶0035-36; Fig. 1]) storing program code portions which, when executed on a device having processing capabilities (computing system 130, comprising CPU 132, storage 140, and memory 134 [¶0034]), performs a method comprising:
receiving an input set of digital images of a sample, wherein the input set of digital images is acquired by illuminating the sample from a plurality of directions and capturing a digital image for each of the plurality of directions (the light source 506, may be an LED array, which can illuminate the sample (in this case, blood slide 508) from a plurality of directions [¶0057-58; Fig. 5], images are captured for each of the unique illumination conditions [¶0067]); and
analyzing the sample by:
inputting the input set of digital images into a trained machine learning model (raw data is acquired for machine learning [¶0051; Fig. 3]) according to the method of claim 1 (as described previously); and receiving, from the trained machine learning model, an output including a classification of at least one portion of the sample (the process of Fig. 3 can be used to train the computational algorithm 206 to output a classification of each cell in the sample [¶0051], such as classifying cells based on whether they exhibit signs of an infection on a cell-by-cell basis [¶0067; Fig. 7A & Fig. 7D]), wherein the trained machine learning model is trained by:
receiving a ground truth including a classification of at least one portion of a sample (the computational algorithm 206 can be a supervised machine learning algorithm [0047], “variably-illuminated training data” of cells classified as “infected” or “not infected” (ground truth images of known classification) are used for Network training (b) [Fig. 7A]);
acquiring a training set of digital images of the sample by illuminating the sample from a plurality of directions and capturing a digital image for each of the plurality of directions (raw data is acquired for machine learning [¶0051; Fig. 3], the light source 506, may be an LED array, which can illuminate the sample (in this case, blood slide 508) from a plurality of directions to acquire training data [¶0057-58; Fig. 5], images are captured for each of the unique illumination conditions [¶0067]); and
training the machine learning model to analyze the sample using the training set of digital images and the received ground truth (the learning sensing network 706 (which may include a convolutional neural network (CNN)) is trained for an inference task (in the form of image classification of blood cells infected with a malaria parasite) during network training (b), wherein variably-illuminated training data of a known infection status (i.e. ground truth) is provided to the LSN to optimize LED illumination schemes for accurate classification using low-resolution images [¶0061-67; 80; Fig. 7A-D]).
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.
Claim 17 is rejected under 35 U.S.C. 103 as being unpatentable over Bokadia et al (US 2022/0334371 A1) in view of Stringer et al (“Cellpose: A generalist algorithm for cellular segmentation”, Nature Methods, 2020).
Regarding claim 17, Bokodia teach The device according to claim 15 (as described previously), wherein the circuitry (computing system 130, comprising CPU 132, storage 140, and memory 134 [¶0034]) is further configured to:
receive a high-resolution digital image of the sample (regions of interest can be imaged at a higher resolution with a greater magnification ¶0053], a high-resolution phase map 816 is obtained via FTP reconstruction 814 [¶0095-97]); and
classify at least one portion of the sample by classifying a corresponding portion of the high-resolution digital image of the sample; (classifying cells based on whether they exhibit signs of an infection on a cell-by-cell basis [¶0067; Fig. 7A & Fig. 7D]).
Bokodia does not explicitly recite a process for forming the ground truth comprising classifications of a portion of a sample. Stringer et al (hereinafter referred to as “Stringer”), however, is analogous art pertinent to the field of endeavor and disclose a generalized algorithm for cellular segmentation. More specifically, Stringer et al teach and form the ground truth comprising the classification of the at least one portion of the sample (Stringer: each voxel is classified as either background, the nucleus, or the cytoplasm in 3D cellular segmentation through manual annotation of a small number of voxels [Sec Methods – Using ilastik for 3D segmentation; ¶01-02]).
Stringer further disclose that their algorithm allows for precisely segmenting a variety of cell types from a range of image types without model retraining or parameter adjustments [Sec - Abstract]. Therefore, it would have been obvious before the effective filing date of the present application to incorporate the ground truth formation proposed by Stringer into the Fourier ptychographic system outlined by Bokodia to arrive at the present invention of the instant application.
Claim 18 is rejected under 35 U.S.C. 103 as being unpatentable over Bokadia et al (US 2022/0334371 A1) in view of Stringer et al (“Cellpose: A generalist algorithm for cellular segmentation”, Nature Methods, 2020), further in view of Zhang et al (“Fourier ptychographic microscopy reconstruction with multiscale deep residual network”, Optics Express, 2019).
Regarding claim 18, Bokodia in view of Stringer teach The device according to claim 17 (as described above), wherein the circuitry (Bokodia: computing system 130, comprising CPU 132, storage 140, and memory 134 [¶0034; Fig. 1]) is further configured to:
train the machine learning model (Bokodia: the loss function 300 is used to train computational algorithm 206 [¶0051; Fig. 3]) using a first subset of the training set of digital images (Bokodia: raw data is acquired for machine learning [¶0051; Fig. 3], the light source 506, may be an LED array, which can illuminate the sample (in this case, blood slide 508) from a plurality of directions to acquire training data [¶0057-58; Fig. 5], images are captured for each of the unique illumination conditions [¶0067])
the high-resolution image having a resolution higher than a resolution of the digital images of the training set; (Bokodia: the high-resolution phase map 816 has a higher resolution as it is generated via Fourier ptychographic image processing [¶0095-97], this high resolution phase map is generated via images at a lower resolution / objective [¶0090-93).
Zhang et al (hereinafter referred to as “Zhang”), however, is analogous art pertinent to the field of endeavor and disclose Fourier microscopy reconstruction using subsets of training data. More specifically, Zhang et al teach configured to reconstruct the high-resolution digital image of the sample from a second subset of the training set of digital images, (Zhang: a large training set split into subsets for training, with the high-resolution images reconstructed from training dataset images and patches [Sec 3.2 – Building the datasets; ¶01-03; Fig. 3]).
Zhang et al further disclose that their method yields better reconstruction results in less time [Sec – Abstract]. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the present invention to utilize the reconstruction of digital images from training subsets outlined by Zhang in the Fourier ptychographic system disclosed by Bokodia to arrive at the present invention of the instant application.
Claims 22 & 23 are rejected under 35 U.S.C. 103 as being unpatentable over Bokadia et al (US 2022/0334371 A1) in view of Engel et al (US 2022/0350123 A1).
Regarding claim 22, Bokodia teach The microscope system according to claim 20 (as described previously), but only discloses light sources arranged on a quasi-hemispherical surface and not a curved surface.
Engel et al (hereinafter referred to as “Engel), is analogous art pertinent to the field of endeavor of the present application and disclose a ptychographic imaging system using a curved array of light sources. More specifically, Engel teach wherein the plurality of light sources are arranged on a curved surface being concave along at least one direction along the surface (Engel: light sources 2 are arranged in concentric circles forming a curved, hemi-spherical surface [¶0089-90; Figs. 6 & 7]), and wherein each light source of the plurality of light sources is configured to illuminate the sample from one of the plurality of directions (Engel: light sources 2 illuminate the sample (at sample location 5) from a plurality of directions [¶0083; Fig 2]).
Engel further discloses that selection of such an arrangement of light sources may be decided for a desired polar angle to provide an appropriate overlap in the Fourier spectrum [¶0092]. Therefore, it would have been obvious to one of ordinary skill before the effective filing date of the present application to implement the curved array of light sources disclosed by Engel to the base microscope system of Bokodia to optimize the overlapping Fourier spectrum for each individual image for improved ptychographic image reconstruction.
Regarding claim 23, Bokodia in view of Engel teach The microscope system according to claim 22, wherein the curved surface is formed of facets (Bokodia: according to ¶0060 of the specification, a curved surface formed of facets can be constructed via a plurality of flat surfaces – light sources 2 are comprised of individual point light sources (indicated by a four-point star), which, on the individual level, can be flat LEDs [¶0138; Figs. 6 & 7]).
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 Michael M. Sofroniou whose telephone number is (571)272-0287. The examiner can normally be reached M-F: 8:30 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, John M. Villecco can be reached at (571) 272-7319. 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.
/MICHAEL M SOFRONIOU/Examiner, Art Unit 2661
/JOHN VILLECCO/Supervisory Patent Examiner, Art Unit 2661
Prosecution Insights