DETAILED ACTION
This action is pursuant to claims filed on 09/26/2025. Claims 19-28 are pending. A first action on the merits of claims 19-28 is as follows.
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 09/26/2025 has been entered.
Response to Amendment
The response filed on 9/26/2025 is not in conformance with the Office's rules and regulations regarding claim amendments. In particular, the canceled claims 1-11 and 13-18 should not have text. Also, the withdrawn claim 12 should be presented with its text. Further, claim 21 is missing its status identifier. In an effort to continue prosecution, the amendments have been entered, but the Applicant should be mindful of the proper format for making amendments.
Claim Objections
Claims 19-20, 22, 24, and 26-28 are objected to because of the following informalities:
In claim 19, page 10, line 6, “define ROIs” should read “define the ROIs”
In claim 19, page 10, line 8, “retain ROIs” should read “retain the ROIs”
In claim 19, page 10, line 8, the acronym PPG should be spelled out when it is first introduced in the claims
In claim 19, page 10, line 12, the acronym OCT should be spelled out when it is first introduced in the claims
In claim 19, page 10, line 18, “iPPG” should either read “imaging-PPG” to align with the terminology used earlier in the claim, or iPPG should be introduced as an acronym for imaging-PPG when it is introduced in the claim
In claim 19, page 11, line 1, “the glucose value” should read “the blood glucose value”
In claim 20, lines 5-6, “the region of interest” should read “the ROI”
In claim 20, line 7, “that region of interest” should read “that ROI”
In claim 22, line 15, the line “(supported by specification” should be deleted from the claim
In claim 24, line 19, “the blood glucose level” should read “the blood glucose value” to align with the language used in claim 19
In claim 24, line 21, “feature fusion)” should read “feature fusion”
In claim 26, line 6, “signal quality)” should read “signal quality”
In claim 27, line 7, “the blood glucose level” should read “the blood glucose value” to align with the language used in claim 19
In claim 27, line 9, “the blood glucose level” should read “the blood glucose value” to align with the language used in claim 19
In claim 27, lines 8-9, “features (optionally in combination with other extracted features)” should read “features, optionally in combination with other extracted features,”
In claim 28, line 12, “the blood glucose level” should read “the blood glucose value” to align with the language used in claim 19
Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 19-28 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Regarding claim 19, Applicant has added the limitation “extracting imaging-PPG features from each retained ROI, including pulse amplitude, phase and harmonic ratios, and perfusion index statistics” in lines 9-10, which is not described in the originally filed claims, specification, or drawings to support this newly added limitation. Thus, the newly added limitation is deemed to be new matter. In particular, the limitation of extracting pulse amplitude, phase and harmonic ratios, and perfusion index statistics are not described in the claim or the specification. Therefore, the claim fails the new matter requirement and is rejected under 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph. Claims 20-28 are also rejected due to their interpreted dependence from claim 19.
Further regarding claim 19, Applicant has added the limitation “the network having been trained on paired RGB-video and OCT datasets to learn a mapping from video-derived spatial/temporal features to OCT-variation features” in lines 13-15, which is not described in the originally filed claims, specification, or drawings to support this newly added limitation. Thus, the newly added limitation is deemed to be new matter. In particular, the limitation of training the network with paired RGB-video and OCT datasets to learn the mapping are not described in the claim or the specification. Therefore, the claim fails the new matter requirement and is rejected under 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph. Claims 20-28 are also rejected due to their interpreted dependence from claim 19.
Further regarding claim 19, Applicant has added the limitation “fusing the imaging-PPG features and the OCT-variation feature map to (i) select a glucose-sensitive subregion and (ii) construct a feature vector combining temporal iPPG descriptors with depth/texture statistics” in lines 16-18, which is not described in the originally filed claims, specification, or drawings to support this newly added limitation. Thus, the newly added limitation is deemed to be new matter. In particular, the limitation of fusing the imaging-PPG features and the OCT-variation feature map and constructing a feature vector combining temporal iPPG descriptors with depth/texture statistics are not described in the claim or the specification. Therefore, the claim fails the new matter requirement and is rejected under 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph. Claims 20-28 are also rejected due to their interpreted dependence from claim 19.
Regarding claim 20, Applicant has added the limitation “identifying the region of interest in the video frames corresponding to a blood-perfused tissue area of the subject” in lines 5-6, which is not described in the originally filed claims, specification, or drawings to support this newly added limitation. Thus, the newly added limitation is deemed to be new matter. In particular, the limitation of identifying the region of interest corresponding to a blood-perfused tissue area is not described in the claim or the specification. Therefore, the claim fails the new matter requirement and is rejected under 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph.
Regarding claim 21, Applicant has added the limitation “detecting pulsatile intensity variations over time in the region of interest” in lines 9-10, which is not described in the originally filed claims, specification, or drawings to support this newly added limitation. Thus, the newly added limitation is deemed to be new matter. In particular, the limitation of detecting pulsatile intensity variations over time is not described in the claim or the specification. Therefore, the claim fails the new matter requirement and is rejected under 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph.
Regarding claim 22, Applicant has added the limitation “wherein the OCT-like features comprise a depth-resolved tissue reflectance profile or a synthetic tomographic image analogous to an OCT scan of the region of interest” in lines 13-15, which is not described in the originally filed claims, specification, or drawings to support this newly added limitation. Thus, the newly added limitation is deemed to be new matter. In particular, the limitation of the features comprising a depth-resolved tissue reflectance profile or a synthetic tomographic image is not described in the claim or the specification. Therefore, the claim fails the new matter requirement and is rejected under 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph.
Regarding claim 23, Applicant has added the limitation “detecting pulsatile intensity variations over time in the region of interest” in lines 17-18, which is not described in the originally filed claims, specification, or drawings to support this newly added limitation. Thus, the newly added limitation is deemed to be new matter. In particular, the limitation of detecting pulsatile intensity variations over time is not described in the claim or the specification. Therefore, the claim fails the new matter requirement and is rejected under 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph.
Regarding claim 27, Applicant has added the limitation “a machine-learning regression model” in line 8, which is not described in the originally filed claims, specification, or drawings to support this newly added limitation. Thus, the newly added limitation is deemed to be new matter. In particular, the limitation of a machine-learning regression model is not described in the claim or the specification. Therefore, the claim fails the new matter requirement and is rejected under 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph.
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 19-28 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 19, the phrase “(e.g., a deep learning regressor or trained classifier)” in lines 19-20 renders the claim indefinite because it is unclear whether the limitation(s) following the phrase are part of the claimed invention. See MPEP § 2173.05(d). Claims 20-28 are also rejected due to their interpreted dependence from claim 19.
Further regarding claim 19, the claim recites the limitation “fusing the imaging-PPG features and the OCT-variation feature map” in line 16. It is unclear what it means to fuse the imaging-PPG features and the OCT-variation feature map. It is also unclear how this function is performed. The broad and indefinite scope of the limitation fails to inform a person of ordinary skill in the art with reasonable certainty of the metes and bounds of the claimed invention, therefore the claim is rendered indefinite. For purposes of examination, it is being interpreted as any combination of the i-PPG features and the OCT-variation feature map to be used for determining the blood glucose level of the user. Claims 20-28 are also rejected due to their interpreted dependence from claim 19.
Further regarding claim 19, the claim recites the limitation “select a glucose-sensitive subregion” in lines 16-17. It is unclear what a glucose-sensitive subregion is and how it is determined. The broad and indefinite scope of the limitation fails to inform a person of ordinary skill in the art with reasonable certainty of the metes and bounds of the claimed invention, therefore the claim is rendered indefinite. For purposes of examination, it is being interpreted as referring to any region of the region of interest. Claims 20-28 are also rejected due to their interpreted dependence from claim 19.
Regarding claim 20, the claim is dependent on claim 1, however claim 1 has been cancelled. A claim that depends from a cancelled claim is indefinite since it is not clear what the metes and bounds of the claim are. For purposes of examination, it is being interpreted as depending on claim 19, the only independent claim that has not been cancelled.
Regarding claim 21, the claim is dependent on claim 1, however claim 1 has been cancelled. A claim that depends from a cancelled claim is indefinite since it is not clear what the metes and bounds of the claim are. For purposes of examination, it is being interpreted as depending on claim 19, the only independent claim that has not been cancelled.
Regarding claim 22, the claim is dependent on claim 1, however claim 1 has been cancelled. A claim that depends from a cancelled claim is indefinite since it is not clear what the metes and bounds of the claim are. For purposes of examination, it is being interpreted as depending on claim 19, the only independent claim that has not been cancelled.
Regarding claim 23, the claim is dependent on claim 1, however claim 1 has been cancelled. A claim that depends from a cancelled claim is indefinite since it is not clear what the metes and bounds of the claim are. For purposes of examination, it is being interpreted as depending on claim 19, the only independent claim that has not been cancelled.
Regarding claim 24, the claim is dependent on claim 5, however claim 5 has been cancelled. A claim that depends from a cancelled claim is indefinite since it is not clear what the metes and bounds of the claim are. For purposes of examination, it is being interpreted as depending on claim 23.
Further regarding claim 24, the claim recites the limitation “the iPPG rich regions of interest” in line 20. There is insufficient antecedent basis for this limitation in the claim. Additionally, it is unclear what is considered an iPPG rich region of interest. For purposes of examination, it is being interpreted as referring to the region of interests identified in claim 19.
Regarding claim 25, the claim is dependent on claim 1, however claim 1 has been cancelled. A claim that depends from a cancelled claim is indefinite since it is not clear what the metes and bounds of the claim are. For purposes of examination, it is being interpreted as depending on claim 19, the only independent claim that has not been cancelled.
Further regarding claim 25, the claim recites the limitation “the deep learning model” in line 22. It is unclear if this is referring to the trained deep neural network from claim 10, or the deep learning regressor from claim 19. The broad and indefinite scope of the limitation fails to inform a person of ordinary skill in the art with reasonable certainty of the metes and bounds of the claimed invention, therefore the claim is rendered indefinite. For purposes of examination, it is being interpreted as referring to the trained deep neural network.
Regarding claim 26, the claim is dependent on claim 1, however claim 1 has been cancelled. A claim that depends from a cancelled claim is indefinite since it is not clear what the metes and bounds of the claim are. For purposes of examination, it is being interpreted as depending on claim 19, the only independent claim that has not been cancelled.
Further regarding claim 26, the claim recites the limitation “the deep learning model” in line 5. It is unclear if this is referring to the trained deep neural network from claim 10, or the deep learning regressor from claim 19. The broad and indefinite scope of the limitation fails to inform a person of ordinary skill in the art with reasonable certainty of the metes and bounds of the claimed invention, therefore the claim is rendered indefinite. For purposes of examination, it is being interpreted as referring to the trained deep neural network.
Regarding claim 27, the claim is dependent on claim 1, however claim 1 has been cancelled. A claim that depends from a cancelled claim is indefinite since it is not clear what the metes and bounds of the claim are. For purposes of examination, it is being interpreted as depending on claim 19, the only independent claim that has not been cancelled.
Further regarding claim 27, the claim recites the limitation “a machine learning regression model”. It is unclear if this is referring to the calibrated regression model from claim 19, or a different regression model. If it is referring to the calibrated regression model from claim 19, it needs to refer back to it. If it is referring to a different regression model, it needs to be distinguished from the candidate from claim 19. For purposes of examination, it is being interpreted as referring to the calibrated regression model from claim 19.
Regarding claim 28, the claim is dependent on claim 1, however claim 1 has been cancelled. A claim that depends from a cancelled claim is indefinite since it is not clear what the metes and bounds of the claim are. For purposes of examination, it is being interpreted as depending on claim 19, the only independent claim that has not been cancelled.
The following is a quotation of 35 U.S.C. 112(d):
(d) REFERENCE IN DEPENDENT FORMS.—Subject to subsection (e), a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
The following is a quotation of pre-AIA 35 U.S.C. 112, fourth paragraph:
Subject to the following paragraph [i.e., the fifth paragraph of pre-AIA 35 U.S.C. 112], a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
Claims 20-28 are rejected under 35 U.S.C. 112(d) or pre-AIA 35 U.S.C. 112, 4th paragraph, as being of improper dependent form for failing to further limit the subject matter of the claim upon which it depends, or for failing to include all the limitations of the claim upon which it depends. Claims 20-28 depend from canceled claims, which is improper.
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 factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 19, 21, 23-25, and 27-28 are rejected under 35 U.S.C. 103 as being unpatentable over Nour (US 11324406) in further view of Mejía-Mejía (“Photoplethysmography Signal Processing and Synthesis”), Medeiros (“From Machine to Machine: An OCT-trained Deep Learning Algorithm for Objective Quantification of Glaucomatous Damage in Fundus Photographs”), Veltz (US 20170173262), and Jayaraman (US 10772569).
Regarding independent claim 19, Nour teaches a computer-implemented method for non-invasive estimation of blood glucose using only a video camera (Abstract: “Provided are systems and methods for determining a subject's physiological parameters such as heart rate, blood glucose level, variation in heart rate and/or oxygen saturation in a non-contact manner”), the method comprising:
(a) acquiring, via a monocular RGB video camera operating at >30fps (Column 10, lines 15-16: “Images obtained at 200 fps are combined to produce color RGB video recording”; Claim 1: “capturing by a camera a red-green-blue (RGB) video of the subject, the RGB video being of at least one area of the subject's body selected from the group consisting of at least a part of a head of the subject, at least a part of an arm of the subject, and at least a part of a leg of the subject”), a time series of facial frames of a subject under controlled illumination conditions (Column 3, lines 64-65: “a photoplethysmography signal derived from any point of video frames in the time domain”; Column 1, lines 25-27: “A photoplethysmography is obtained by illuminating the relevant area of the body and receiving the reflected or transmitted light signal”);
(b) identifying candidate skin regions of interest (ROIs) by:
(i) segmenting facial landmarks to define ROIs (Column 10, lines 10-14: “object detection algorithms are used to detect the face, finger, palm, or ear from the images obtained from the camera. And the relevant region is determined (Region of Interest-ROI)”)
and separating the intensity traces based on the channels (Column 8, lines 44-48: “Glucose concentration can be determined by analysis of changes in wavelength or intensity of passing light. The blood sugar content can be calculated by measuring with infrared light of different wavelengths (from 700 nm to 2500 nm)”).
However, Nour does not teach computing a temporal signal-to-noise or quality metric on channel-separated intensity traces to retain ROIs exhibiting stable imaging-PPG pulsatile components.
Mejía-Mejía discloses photoplethysmography signal processing steps. Specifically, Mejía-Mejía discloses computing a temporal signal-to-noise or quality metric on signals to retain ROIs exhibiting stable imaging-PPG pulsatile components (Page 13: “A wide range of pulse wave characteristics have been used to assess signal quality. Amplitude characteristics include: pulse wave amplitude [49], [50], and the Perfusion Index (PI, which measures the ratio of the ‘AC’ to ‘DC’ component of the signal) [51]. Timing characteristics include the systolic phase duration, the ratio of systolic to diastolic phase duration, inter-beat-intervals [52], the average pulse rate. Shape characteristics include: the number of diastolic peaks; the number of times the signal changes from positive to negative, or vice versa, also called the zero-crossing rate; the signal-to-noise ratio (SNR); and the comparison of the accuracy of different systolic wave detectors for isolating events (e.g., beats or noise artefacts) [51]. Higher order statistics, such as skewness and kurtosis, give an indication of the distribution of the data, therefore serve as an indication of how the pulse wave is distributed over time [30], [51]. These measurements are especially good at identifying PPG pulses with outliers generated by noise [30]. Shannon entropy has been proposed as another measurement of the presence of ‘disorder’ in a PPG pulse wave, since it is a measure of uncertainty in a system, which increases in noisy PPG signals”). Nour and Mejía-Mejía are analogous arts as they are both related to systems that process and analyze photoplethysmography signals.
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to include the signal-to-noise ratio and quality metric from Mejía-Mejía into the method from Nour as it allows the method to determine that the signals are producing quality measurements, which will ensure that the method can provide the most accurate result.
The Nour/Mejía-Mejía combination teaches (c) extracting imaging-PPG features from each retained ROI, including pulse amplitude (Column 16, lines 10-22: “generating with the data processor, a measurement of the oxygen saturation level of the subject using said calibration parameters, the red band signal and the blue band signal from the decomposed RGB video; simultaneously generating with the data processor a measurement of the blood glucose level of the subject, using said green band signal by generating a three-dimensional (3D) RGB matrix from said RGB video; wherein the 3D RGB matrix is generated by combining the red band signal, the blue band signal, and the green band signal generating a one-dimensional (1D) photoplethysmogram (PPG) signal by a 3D to 1D dimensional reduction of the 3D RGB matrix, creating a blood glucose level measurement model”; Fig. 6).
However, the Nour/Mejía-Mejía combination does not teach extracting phase and harmonic ratios and perfusion index statistics.
Mejía-Mejía discloses extracting phase and harmonic ratios and perfusion index statistics (Page 27: “The harmonics of the PPG signal can also be studied”; Page 13: “A wide range of pulse wave characteristics have been used to assess signal quality. Amplitude characteristics include: pulse wave amplitude [49], [50], and the Perfusion Index (PI, which measures the ratio of the ‘AC’ to ‘DC’ component of the signal) [51]. Timing characteristics include the systolic phase duration, the ratio of systolic to diastolic phase duration, inter-beat-intervals [52], the average pulse rate. Shape characteristics include: the number of diastolic peaks; the number of times the signal changes from positive to negative, or vice versa, also called the zero-crossing rate; the signal-to-noise ratio (SNR); and the comparison of the accuracy of different systolic wave detectors for isolating events (e.g., beats or noise artefacts) [51]. Higher order statistics, such as skewness and kurtosis, give an indication of the distribution of the data, therefore serve as an indication of how the pulse wave is distributed over time [30], [51]. These measurements are especially good at identifying PPG pulses with outliers generated by noise [30]. Shannon entropy has been proposed as another measurement of the presence of ‘disorder’ in a PPG pulse wave, since it is a measure of uncertainty in a system, which increases in noisy PPG signals”).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to include the phase and harmonic ratios and perfusion index statistics from Mejía-Mejía into the Nour/Mejía-Mejía combination as it allows the device to analyze additional aspects of the photoplethysmography signals, which can provide helpful information that can be used to provide a more accurate and comprehensive analysis of the signals and the user’s health state.
However, the Nour/Mejía-Mejía combination does not teach generating, by executing a trained deep neural network on the acquired frames, an OCT-variation feature map that represents optical-coherence-tomography-like depth/texture/coherence signatures for each ROI, the network having been trained on paired RGB-video and OCT datasets to learn a mapping from video-derived spatial/temporal features to OCT-variation features.
Veltz discloses a medical system and device for determining blood glucose of a user. Specifically, Veltz teaches using both PPG and OCT measurements to determine blood glucose for a user ([0042]-[0043]: “An optical or electrochemical detector (glucometer) can be used to analyze the blood sample and can give a numerical glucose reading. Recently, non-invasive glucose measuring devices that monitor BG through infrared technology and optical sensing have become available; blood pressure sensor can be a non-invasive sensor designed to measure systolic and diastolic human blood pressure utilizing the oscillometric technique”; [0260]: “Imaging methods associated with encapsulating devices can include … optical coherence tomography (OCT)”; [0387]: “It is also suggested to combine described embodiments or combinations thereof with one or more of the following words, suggesting techniques or technologies or concepts or ideas or paradigms or states or the like: … photoplethysmograph”). Veltz and Nour are analogous arts as they are both related to systems that determine blood glucose levels of a user through imaging techniques.
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to include using both PPG and OCT to determine glucose levels of a user as it allows multiple different techniques to be used, which can provide the method with more information and can allow for a more accurate and comprehensive result.
However, the Nour/Mejía-Mejía/Veltz combination does not teach the OCT features being determined through a neural network analyzing the acquired frames.
Medeiros discloses an OCT-trained deep learning algorithm. Specifically, Medeiros teaches (d) generating, by executing a trained deep neural network on the acquired frames, an OCT-variation feature map that represents optical-coherence-tomography-like depth/texture/coherence signatures for each ROI, the network having been trained on paired RGB-video and OCT datasets to learn a mapping from video-derived spatial/temporal features to OCT-variation features (Page 4: “A deep learning algorithm was trained to predict SDOCT average RNFL thickness from assessment of optic disc photographs. The target value, i.e., the variable we wanted to predict from analysis of optic disc photographs was the SDOCT average RNFL thickness.”; Page 4: “Therefore, for training the neural network, a pair of train-target consisted of the optic disc photograph and the SDOCT average RNFL thickness value. The sample of pairs of photos OCT was split into a training plus validation set (80%) and test sample (20%)”). Nour, Veltz, and Medeiros are analogous arts as they are all related to using imaging techniques to determine health parameters of a user.
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to include the OCT technology from Medeiros into the Nour/Mejía-Mejía/Veltz combination as it allows the method to perform the OCT-like analysis, but does not require all the OCT technology and instead only requires receiving the images, which reduces the amount of technology that is required to perform the analysis.
The Nour/Mejía-Mejía/Veltz/Medeiros combination teaches (e) fusing the imaging-PPG features and the OCT-variation feature map to (i) select a glucose-sensitive subregion (Veltz: [0042]-[0043]: “An optical or electrochemical detector (glucometer) can be used to analyze the blood sample and can give a numerical glucose reading. Recently, non-invasive glucose measuring devices that monitor BG through infrared technology and optical sensing have become available; blood pressure sensor can be a non-invasive sensor designed to measure systolic and diastolic human blood pressure utilizing the oscillometric technique”; [0260]: “Imaging methods associated with encapsulating devices can include … optical coherence tomography (OCT)”; [0387]: “It is also suggested to combine described embodiments or combinations thereof with one or more of the following words, suggesting techniques or technologies or concepts or ideas or paradigms or states or the like: … photoplethysmograph”; Mejía-Mejía, Page 27: “The harmonics of the PPG signal can also be studied”; Page 13: “A wide range of pulse wave characteristics have been used to assess signal quality. Amplitude characteristics include: pulse wave amplitude [49], [50], and the Perfusion Index (PI, which measures the ratio of the ‘AC’ to ‘DC’ component of the signal) [51]. Timing characteristics include the systolic phase duration, the ratio of systolic to diastolic phase duration, inter-beat-intervals [52], the average pulse rate. Shape characteristics include: the number of diastolic peaks; the number of times the signal changes from positive to negative, or vice versa, also called the zero-crossing rate; the signal-to-noise ratio (SNR); and the comparison of the accuracy of different systolic wave detectors for isolating events (e.g., beats or noise artefacts) [51]. Higher order statistics, such as skewness and kurtosis, give an indication of the distribution of the data, therefore serve as an indication of how the pulse wave is distributed over time [30], [51]. These measurements are especially good at identifying PPG pulses with outliers generated by noise [30]. Shannon entropy has been proposed as another measurement of the presence of ‘disorder’ in a PPG pulse wave, since it is a measure of uncertainty in a system, which increases in noisy PPG signals”).
However, the Nour/Mejía-Mejía/Veltz/Medeiros combination does not teach constructing a feature vector combining temporal iPPG descriptors with depth/texture statistics.
Jayaraman discloses devices and methods to detect diabetes. Specifically, Jayaraman teaches (ii) constructing a feature vector combining temporal iPPG descriptors with depth/texture statistics (Claim 1: “using a machine learning technique that includes selection of features from the signal based on an impact coefficient computed from a ratio of correlation between pulse rate variability features in an input data set of the signal and output being a function of a pulse rate variability features of a number of users, and sum of correlations between the pulse rate variability features in the input data set of the signal with remaining features other than the features correlated in the input data set, and selecting features that are highly correlated with the output and least correlated with the remaining features in the input data set are selected as input feature vector”). Nour, Veltz, Medeiros, and Jayaraman are analogous arts as they are all related to systems that determine blood glucose levels of a user through imaging techniques.
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to include the feature vector from Jayaraman into the Nour/Mejía-Mejía/Veltz/Medeiros combination as it allows for easier processing and a clear input into the machine learning model, which can ensure clear and easy processing for the machine learning.
The Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination teaches (f) inputting the feature vector into a calibrated regression model (e.g., a deep learning regressor or trained classifier) to output an estimated blood glucose value (Nour, Column 10, lines 40-45: “From the PPG signal changes obtained when hungry and full, the person's blood sugar rate is estimated by mathematical or intuitive methods. Linear and nonlinear regression methods are tried as mathematical methods, and the best method is determined”; Column 11, lines 42-52: “Prediction algorithms: Regression and prediction algorithms are used to estimate physiological parameters from the non-contact PPG signal. Methods to Use: Curve fitting algorithms (MATLAB) Linear regression algorithms Nonlinear regression methods Artificial neural networks Fuzzy logic Neural-fuzzy logic network-based hybrid systems Support vector machines”); and
(g) presenting the glucose value together with a correlation label that visually links the selected subregion to the displayed video frame (Nour, Column 4, lines 40-50: “one or more monitors or screens for displaying the analyzed physiological parameter data for a visual display. The present disclosure also includes a method for determining at least one physiological parameter in a contactless manner. The method includes obtaining at least one image and/or video of at least one area of the subject's body, decomposing the image and/or video into three or more color channel components, determining one or more physiological parameters and visually displaying the analyzed physiological parameters in real time”; Fig. 10).
Regarding claim 21, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination teaches the method of claim 1, further comprising extracting a photoplethysmographic (PPG) signal from the video frames by detecting pulsatile intensity variations over time in the region of interest (Nour, Column 10, lines 59-61: “D.C. components were calculated as the red and blue band average intensities at each time point.”).
Regarding claim 23, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination teaches the method of claim 1, further comprising extracting a photoplethysmographic (iPPG) signal from the video frames by detecting pulsatile intensity variations over time in the region of interest (Nour, Column 10, lines 59-61: “D.C. components were calculated as the red and blue band average intensities at each time point.”).
Regarding claim 24, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination teaches the method of claim 5, wherein estimating the blood glucose level comprises using the iPPG rich regions of interest with the OCT-like features to improve estimation accuracy regarding feature fusion (Mejía-Mejía: Page 13: “A wide range of pulse wave characteristics have been used to assess signal quality. Amplitude characteristics include: pulse wave amplitude [49], [50], and the Perfusion Index (PI, which measures the ratio of the ‘AC’ to ‘DC’ component of the signal) [51]. Timing characteristics include the systolic phase duration, the ratio of systolic to diastolic phase duration, inter-beat-intervals [52], the average pulse rate. Shape characteristics include: the number of diastolic peaks; the number of times the signal changes from positive to negative, or vice versa, also called the zero-crossing rate; the signal-to-noise ratio (SNR); and the comparison of the accuracy of different systolic wave detectors for isolating events (e.g., beats or noise artefacts) [51]. Higher order statistics, such as skewness and kurtosis, give an indication of the distribution of the data, therefore serve as an indication of how the pulse wave is distributed over time [30], [51]. These measurements are especially good at identifying PPG pulses with outliers generated by noise [30]. Shannon entropy has been proposed as another measurement of the presence of ‘disorder’ in a PPG pulse wave, since it is a measure of uncertainty in a system, which increases in noisy PPG signals”; Veltz, [0042]-[0043]: “An optical or electrochemical detector (glucometer) can be used to analyze the blood sample and can give a numerical glucose reading. Recently, non-invasive glucose measuring devices that monitor BG through infrared technology and optical sensing have become available; blood pressure sensor can be a non-invasive sensor designed to measure systolic and diastolic human blood pressure utilizing the oscillometric technique”; [0260]: “Imaging methods associated with encapsulating devices can include … optical coherence tomography (OCT)”; [0387]: “It is also suggested to combine described embodiments or combinations thereof with one or more of the following words, suggesting techniques or technologies or concepts or ideas or paradigms or states or the like: … photoplethysmograph”; Medeiros, Page 4: “A deep learning algorithm was trained to predict SDOCT average RNFL thickness from assessment of optic disc photographs. The target value, i.e., the variable we wanted to predict from analysis of optic disc photographs was the SDOCT average RNFL thickness.”).
Regarding claim 25, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination teaches the method of claim 1, wherein the deep learning model processes a sequence of consecutive video frames to capture temporal blood flow dynamics when generating the one or more OCT-like features, thereby incorporating pulsation information into the features (Nour, Column 5, lines 31-36: “For the formation of a PPG signal, a sufficient amount of the pixel region, in which the PPG signal obtained from a portion of the subject's skin may vary. For example, a sufficient amount of 20×20 pixels in the sequence of video picture frames in the region where the finger (index finger) tip is located may be collected”).
Regarding claim 27, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination teaches the method of claim 1, wherein estimating the blood glucose level comprises applying a machine-learning regression model to the one or more OCT-like features (optionally in combination with other extracted features) to determine the glucose level of the subject (Nour, Column 10, lines 40-45: “From the PPG signal changes obtained when hungry and full, the person's blood sugar rate is estimated by mathematical or intuitive methods. Linear and nonlinear regression methods are tried as mathematical methods, and the best method is determined”; Column 11, lines 42-52: “Prediction algorithms: Regression and prediction algorithms are used to estimate physiological parameters from the non-contact PPG signal. Methods to Use: Curve fitting algorithms (MATLAB) Linear regression algorithms Nonlinear regression methods Artificial neural networks Fuzzy logic Neural-fuzzy logic network-based hybrid systems Support vector machines”; Veltz, [0042]-[0043]: “An optical or electrochemical detector (glucometer) can be used to analyze the blood sample and can give a numerical glucose reading. Recently, non-invasive glucose measuring devices that monitor BG through infrared technology and optical sensing have become available; blood pressure sensor can be a non-invasive sensor designed to measure systolic and diastolic human blood pressure utilizing the oscillometric technique”; [0260]: “Imaging methods associated with encapsulating devices can include … optical coherence tomography (OCT)”; [0387]: “It is also suggested to combine described embodiments or combinations thereof with one or more of the following words, suggesting techniques or technologies or concepts or ideas or paradigms or states or the like: … photoplethysmograph”; [0182]: “[0042]-[0043]: “An optical or electrochemical detector (glucometer) can be used to analyze the blood sample and can give a numerical glucose reading. Recently, non-invasive glucose measuring devices that monitor BG through infrared technology and optical sensing have become available; blood pressure sensor can be a non-invasive sensor designed to measure systolic and diastolic human blood pressure utilizing the oscillometric technique”; [0260]: “Imaging methods associated with encapsulating devices can include … optical coherence tomography (OCT)”; [0387]: “It is also suggested to combine described embodiments or combinations thereof with one or more of the following words, suggesting techniques or technologies or concepts or ideas or paradigms or states or the like: … photoplethysmograph”).
Regarding claim 28, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination teaches the method of claim 1.
However, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination does not teach the method further comprising generating an alert or notification when the estimated blood glucose level falls outside a predetermined safe range for the subject.
Veltz teaches the method further comprising generating an alert or notification when the estimated blood glucose level falls outside a predetermined safe range for the subject ([0098]: “if a hypoglycemia probability or event is determined, associated alerts or values can preempt or replace any other display data”. To determine a hypoglycemia event, there must be a predetermined safe range that indicates when the user is not in a hypoglycemic state.).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to include the alert from Veltz into the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination as it allows the method to notify the user if there is a health issue, which can ensure they can take the appropriate action.
Claim 20 is rejected under 35 U.S.C. 103 as being unpatentable over the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination as applied to claim 1 above, and further in view of Hendriks (US 20240081766).
Regarding claim 20, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination teaches the method of claim 1.
However, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination is silent on the specific features identified in the region of interest.
Hendriks discloses a method for imaging peripheral vascular disease using photoplethysmography. Specifically, Hendriks teaches where the method further comprising identifying the region of interest in the video frames corresponding to a blood-perfused tissue area of the subject, wherein the one or more OCT-like features are generated for that region of interest ([0023]: “the processing unit is configured for deriving, over a predefined region of interest in the organ tissue to be perfused during imaging, a quantity from the aligned first and second changes in the perfusion states”. The OCT-like features are generated for the region of interest, therefore the Oct-like features would be generated for the blood-perfused tissue area.). Nour, Veltz, Medeiros, Jayaraman, and Hendriks are analogous arts as they are all related to systems that use imaging techniques to determine health parameters of a user.
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to include identifying a blood-perfused tissue area in the region of interest from Hendriks into the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination as it ensures that the method is sensing an area with good blood flow, which will provide the most accurate measurement of blood glucose levels.
Claim 22 is rejected under 35 U.S.C. 103 as being unpatentable over the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination as applied to claim 1 above, and further in view of Desjardins (US 20070201033).
Regarding claim 22, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination teaches the method of claim.
However, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination does not teach wherein the OCT-like features comprise a depth-resolved tissue reflectance profile or a synthetic tomographic image analogous to an OCT scan of the region of interest.
Desjardins teaches methods and systems for performing optical coherence tomography. Specifically, Desjardins teaches wherein the OCT-like features comprise a depth-resolved tissue reflectance profile or a synthetic tomographic image analogous to an OCT scan of the region of interest ([0054]: “Fourier-domain optical coherence tomographic reconstruction techniques may be applied the vectors, which can generate depth-resolved reflectance profiles”). Nour, Veltz, Medeiros, Jayaraman, and Desjardins are analogous arts as they are all related to systems that use imaging techniques to determine health parameters of a user.
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to include the reflectance profile from Desjardins into the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination as it provides information about the optical properties of the user’s skin, which can be used for analysis and provides a more comprehensive analysis.
Claim 26 is rejected under 35 U.S.C. 103 as being unpatentable over the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination as applied to claim 1 above, and further in view of JP ‘399 (JP 6866399). Citations to JP 6866399 will refer to the English Machine Translation that accompanies this Office Action.
Regarding claim 26, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination teaches the method of claim 1.
However, the Nour/Mejía-Mejía/Veltz/Medeiros/Jayaraman combination does not teach the method further comprising preprocessing the video frames by performing at least one of motion stabilization or illumination normalization before inputting them to the deep learning model, thereby reducing motion artifacts and lighting variability in the generated features for support on region preparation and signal quality.
JP ‘399 discloses a high resolution imaging using a camera. Specifically, JP ‘399 teaches the method further comprising preprocessing the video frames by performing at least one of motion stabilization or illumination normalization before inputting them to the deep learning model, thereby reducing motion artifacts and lighting variability in the gen