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 .
Status of Claims
This communication is in response to remarks and amendments filed 03/31/2026. Claims 1-17 and 19-21 are pending.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-16 and 21 along with their dependent claims 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.
Claim 1 recites, in part, “entering the image as input into the selected energy transformation model, the selected energy transformation model trained to output, based on the image, a transformed image at a second energy level, different than the first energy level, the transformed image in the same contrast phase as the image”. The claim recites “a transformed image at a second energy level, different than the first energy level, the transformed image in the same contrast phase as the image” but is unclear where it states “as the image”. That is, the claim recites “the transformed image in the same contrast phase as the image” which is confusing, as the sentence is awkwardly constructed, leaving the reader to interpret what the “as the image” constitutes. For purposes of compact prosecution, the Examiner will be interpreting the terms “as the image” to “of the initially selected image”, thereby reciting a transformed image at a second energy level, different than the first energy level, the transformed image in the same contrast phase of the initially selected image.
Claim 11 recites identical limitations and is therefore rejected with the same reasons as stated in claim 1 above.
Claim(s) 2-10, 21 and 12-16 depend on independent claims 1 and 11, respectively, are therefore rejected based on dependency.
Response to Remarks
Applicant’s arguments filed 03/31/2026 with respect to independent claims 1 and 11 have been carefully and respectfully considered in light of the instant amendment, but are not persuasive. Accordingly, this action has been made FINAL.
Drawings
The drawing objections are removed in accordance with the remarks and amendments filed 03/31/2026.
Claim Objections
The claim objection(s) with respect to claim 12 are removed in accordance with the remarks and amendments filed 03/31/2026.
Claim Rejections - 35 USC § 103
On page 13 of the remarks filed 03/31/2026, Applicant argues the “cited combination of Abrol and Muhamedrahimov fails to disclose ‘selecting an energy transformation model from among a plurality of energy transformation models based on the contrast phase’ and ‘entering the image as input into the selected energy transformation model, the selected energy transformation model trained to output, based on the image, a transformed image at a second energy level, different than the first energy level, the transformed image in the same contrast phase as the image,’ as claimed in amended claim 1.” More specifically, Applicant asserts that “Abrol does not disclose that the model is selected from among a plurality of models based on the contrast phase of the image” and further indicates “[t]he only time the contrast phase is taken into account in Abrol is when the machine learning model is used to transform the CT image from one contrast phase to a different contrast phase, and, even then, Abrol does not disclose that the model is selected from a plurality of models based on the contrast phase”.
The Examiner respectfully disagrees. Abrol teaches a selection to be made in order to obtain a desired predicted image using a domain transformation machine learning model. That is, Abrol is required to specify what the type of input image is to obtain a desired predicted outcome, “a particular medical scanning domain is preferred and/or desired to capture/generate a medical image” please see Abrol ¶28. It is apparent that Abrol requires a selection of a specific scanning domain such that a different electrical energy level is obtained. After the first scanning domain is selected as input, a contrast phase is inherent to the scanning domain, as CT images containing an energy level have its specific contrast phase. To remedy this, Muhamedrahimov discloses an identification of contrast phase, please see Muhamedrahimov ¶52. Additionally, Abrol teaches that a medical scanning domain (i.e., the first medical scanning domain) can be any “any suitable configurable setting, configurable control, and/or configurable parameter of a medical imaging device/modality, where changing/adjusting the setting, control, and/or parameter can affect how a medical image visually looks once captured/generated”. Such that, for example, “a medical image can be captured/generated in one medical scanning domain and can be transformed into another medical scanning domain. This can be particularly useful when the domain is an electrical energy level. For example, a CT image can be captured at a low electrical energy level (e.g., 70 kVp) so that the patient is not exposed to excessive radiation. Then, the CT image can be substantively transformed and/or converted to a high electrical energy level (e.g., 120 kVp)” please see Abrol ¶127. Therefore, as shown by the teachings of Abrol in view of Muhamedrahimov, the combination of references teach selecting an energy transformation model from among a plurality of energy transformation models based on the contrast phase and entering the image as input into the selected energy transformation model, the selected energy transformation model trained to output, based on the image, a transformed image at a second energy level, different than the first energy level, the transformed image in the same contrast phase as the image.
Furthermore, Applicant argues “Abrol has no recognition that the mapping performed by the machine learning model from the image to a different electrical energy level may be different for different contrast phases”
The Examiner notes that there is no claim language which indicates the plurality of transformation models requires a “mapping” from the image to a different electrical energy level that “may be different for different contrast phases” besides the initial/selected contrast phase image. The mere selection of a contrast phase does not indicate “an energy transformation model specifically trained for that contrast phase” including other contrast phases to obtain a mapping that may be different for different contrast phases.
On page 14 of the remarks filed 03/31/2026, Applicant argues “Abrol necessarily cannot disclose that the energy transformation model is selected from among a plurality of energy transformation models based on the contrast phase”.
The Examiner respectfully disagrees. It is noted that the rejection of independent claim(s) is a 103 combination of Abrol in view of Muhamedrahimov. In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
Furthermore, Abrol, in fact, discloses a plurality of energy transformation models in which require a specific contrast phase as required by a user selection, “a particular medical scanning domain is preferred and/or desired to capture/generate a medical image” please see Abrol ¶28. That is, Abrol uses a first medical scanning domain, such as a first electrical energy level and/or a first contrast phase, and a second electrical energy level that is different from the first electrical energy level, to obtain a second medical scanning domain. Please see Abrol ¶31. To the degree that, to establish a plurality of energy transformation models, Abrol Fig. 9 discloses a domain transformation machine learning model with a set of training images in a first medical scanning domain and a respectively corresponding set of images in a second medical scanning domain (step 904) to then train the model to output a predicted image (in the second medical scanning domain) with variations of different energy levels and/or contrast phases based on the desired input.
Therefore, the argued limitations were written broad such that they read upon the cited references or are shown explicitly by the references. As a result, the claims stand rejected as follows.
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(s) 1-2, 4, 6-8, 11-12, and 14-15 are rejected under 35 U.S.C. 103 as being unpatentable over Abrol et al. (US 20230071535 A1) in view of MUHAMEDRAHIMOV et al. (US 20220318567 A1).
Regarding claim 1, Abrol discloses method, comprising: obtaining an image at a first energy level acquired with a single-energy computed tomography (CT) imaging system (“the first medical scanning domain 106 can be both an electrical energy level and a contrast phase (e.g., a single CT image can be captured/generated at both a given electrical energy level and a given contrast phase)” Abrol, [0059]; [0022]; [0057]); selecting an energy transformation model from among a plurality of energy transformation models based on the contrast phase (“a particular medical scanning domain is preferred and/or desired to capture/generate a medical image” Abrol, [0028]; i.e., Abrol requires a selection of a specific scanning domain such that a different electrical energy level is obtained, wherein “the first medical scanning domain can be any suitable medical scanning domain as desired, such as a first electrical energy level and/or a first contrast phase” Abrol, [0031]);
entering the image as input into the selected energy transformation model (“the first medical scanning domain can be any suitable medical scanning domain as desired, such as a first electrical energy level and/or a first contrast phase” Abrol, [0031]), the selected energy transformation model trained to output, based on the image, a transformed image at a second energy level, different than the first energy level (“Accordingly, the machine learning model 202 can be configured and/or trained (as described in more detail below with respect to FIGS. 8-9) to transform and/or convert CT scans taken at 70 kVp to CT scans taken at 120 kVp. Thus, in various cases, the transformation component 114 can feed the medical image 104 to the machine learning model 202 as input, and the machine learning model 202 can produce as output the predicted image 204... That is, the second medical scanning domain 206 in this non-limiting example can be an electrical energy level of 120 kVp” Abrol, [0077]; “As an example, if the first medical scanning domain 106 is a particular electrical energy level (e.g., 70 kVp), the second medical scanning domain 206 can be a different electrical energy level (e.g., 120 kVp). As another example, if the first medical scanning domain 106 is a particular contrast phase (e.g., one of non-contrast, early arterial phase, late arterial phase, hepatic phase, nephrogenic phase, venous phase, and/or delayed phase), the second medical scanning domain 206 can be a different contrast phase (e.g., a different one of non-contrast, early arterial phase, late arterial phase, hepatic phase, nephrogenic phase, venous phase, and/or delayed phase).” Abrol, [0071]; see additionally, [0127]), the transformed image in the same contrast phase as the image (“In various other embodiments, the machine learning model 202 can be configured and/or trained to transform and/or convert CT images from any electrical energy level to any other electrical energy level. In still other embodiments, the machine learning model 202 can be configured and/or trained to transform and/or convert CT images from any contrast phase to any other contrast phase.” Abrol, [0079]); and displaying a final transformed image and/or saving the final transformed image in memory, wherein the final transformed image is the transformed image or is generated based on the transformed image (“In various embodiments, the computer-readable memory 110 can store computer-executable components (e.g., receiver component 112, transformation component 114), and the processor 108 can execute the computer-executable components.” Abrol, [0062]; i.e., the final transformed image is generated via the transformation component where “the transformation component can generate, via execution of a machine learning model, a predicted image based on the medical image” Abrol, [0005] may be stored/saved in memory).
Abrol discloses all of the subject matter as described above except for specifically teaching identifying a contrast phase of the image. However, MUHAMEDRAHIMOV in the same field of endeavor teaches identifying a contrast phase of the image (“providing ML models that are automatically trained to classify medical images into contrast states, and/or by increasing accuracy of classification of the ML models. Classification may include a category label indicating the contrast phase” MUHAMEDRAHIMOV, [0052]).
Therefore, it would have been obvious to one of ordinary skill in the art to combine Abrol and MUHAMEDRAHIMOV before the effective filing date of the claimed invention. The motivation for this combination of references would have been to highlight certain anatomy structures such as blood vessels that are difficult to identify without contrast but that appear in the contrast images (MUHAMEDRAHIMOV, [0002]). This motivation for the combination of Abrol and MUHAMEDRAHIMOV is supported by KSR exemplary rationale (G) Some teaching, suggestion, or motivation in the prior art that would have led one of ordinary skill to modify the prior art reference or to combine prior art reference teachings to arrive at the claimed invention. MPEP 2141 (III).
Regarding claim 2, Abrol and MUHAMEDRAHIMOV disclose the method of claim 1, wherein identifying the contrast phase of the image comprises identifying the contrast phase of the image with a contrast phase classifier, the contrast phase classifier comprising a deep learning model trained with a plurality of training triads, each training triad including a set of projection images generated from a 3D volume of a subject (“each of the plurality of medical images comprises a three (3D) dataset, and further comprising extract an input representation of the 3D dataset, wherein the training dataset includes the input representation, and a target input representation of the target medical image is fed into the trained ML mode” MUHAMEDRAHIMOV, [0018]; “Imaging device 112 provides images, which may be included in training dataset(s) 116 optionally with labels... CT contrast images depicting various phases of injected intravenous (IV) contrast. 3D images, and/or slices of 3D images, and/or 4D images may be converted to 2D images for training and/or inference, for example, by selecting 2D slices from a 3D scan, and/or converting the 3D image into a 2D image such as by maximum pixel intensity (MPI) approaches.” MUHAMEDRAHIMOV, [0072]).
Therefore, combining Abrol and MUHAMEDRAHIMOV would meet the claim limitations for the same reasons as previously discussed in claim 1.
Regarding claim 4, Abrol and MUHAMEDRAHIMOV disclose the method of claim 1, wherein the energy transformation model is trained with training pairs, each training pair including a first training image at the first energy level and a second training image at the second energy level (“the training component can electronically train the machine learning model on the training dataset to accurately shift, covert, and/or transform the first medical scanning domain 106 to the second medical scanning domain. More specifically, the training dataset can include a set of training images, each of which is captured/generated via the first medical scanning domain 106, and a respectively corresponding set of annotation images, each of which is captured/generated via the second medical scanning domain. Consider a given training image in the set of training images, where the given training image depicts one or more anatomical structures according to the first medical scanning domain 106. In such case, there can be a given annotation image in the set of annotation images that corresponds to the given training image, where the given annotation image depicts the same one or more anatomical structures as the given training image, but where the given annotation image depicts such one or more anatomical structures according to the second medical scanning domain instead of the first medical scanning domain 106. In other words, the given annotation image can be considered as representing how the given training image would actually look if the given training image had been captured/generated via the second medical scanning domain.” Abrol, [0068]), and wherein the first training image and the second training image are monochromatic images acquired with a dual-energy CT imaging system (“For example, in some cases, the particular training image can be co-registered with the particular annotation image through simulation, through scanning physical phantoms, and/or through dual-energy scans.” Abrol, [0042]).
Regarding claim 6, Abrol and MUHAMEDRAHIMOV disclose the method of claim 1, wherein the energy transformation model is a first energy transformation model and the transformed image is a first transformed image, and wherein the final transformed image is generated based on the first transformed image by entering the first transformed image as input to a second energy transformation model trained to output the final transformed image at a third energy level (“transform and/or convert CT scans taken at 70 kVp to CT scans taken at 120 kVp. Thus, in various cases, the transformation component 114 can feed the medical image 104 to the machine learning model 202 as input, and the machine learning model 202 can produce as output the predicted image 204... That is, the second medical scanning domain 206 in this non-limiting example can be an electrical energy level of 120 kVp” Abrol, [0077]; “As another example, if the first medical scanning domain 106 is a particular contrast phase (e.g., one of non-contrast, early arterial phase, late arterial phase, hepatic phase, nephrogenic phase, venous phase, and/or delayed phase), the second medical scanning domain 206 can be a different contrast phase (e.g., a different one of non-contrast, early arterial phase, late arterial phase, hepatic phase, nephrogenic phase, venous phase, and/or delayed phase).” Abrol, [0071]), the second energy level being different than the third energy level (“In various other embodiments, the machine learning model 202 can be configured and/or trained to transform and/or convert CT images from any electrical energy level to any other electrical energy level. In still other embodiments, the machine learning model 202 can be configured and/or trained to transform and/or convert CT images from any contrast phase to any other contrast phase.” Abrol, [0080]).
Regarding claim 7, Abrol and MUHAMEDRAHIMOV disclose the method of claim 1, wherein the contrast phase is a first contrast phase and wherein identifying the contrast phase of the image comprises identifying the first contrast phase and a second contrast phase of the image and a ratio of the first contrast phase relative to the second contrast phase (“FIG. 15, which is a schematic depicting overlap of clinical indicators associated with different phases demonstrated by a quantitative analysis, for the experiment, in accordance with some embodiments of the present invention. An example abdominal CT axial slice 1502, and a graph 1504 of Hounsfield units as a function of contrast phases for different organs (aorta 1506, inferior vena cava 1508, liver 1510, and kidneys 1512) depicted in CT slice 1502 is presented.” MUHAMEDRAHIMOV, [0176]; i.e., Fig. 15 illustrates a ratio of phases in the image including non-contrast, arterial, venous and delayed; Alternatively, using time to determine ratio of contrast phases in images, “The respective time interval may be based on empirical clinical evidence, such as when the contrast phase is expected to appear in a population of subjects. The mapping may be performed, for example, according to a predefined mapping function, and/or a mapping dataset. For example: a non-enhanced contrast phase is mapped to a value of zero seconds, an arterial contrast phase is mapped to a value of 27.5 seconds within an approximate time interval of 20-35 seconds, a venous contrast phase is mapped to a value of 75 seconds within an approximate time interval of 60-90 seconds, and a delayed contrast phase is mapped to a value of 480 seconds within an approximate time interval of 6-10 minutes.” MUHAMEDRAHIMOV, [0091]).
Therefore, combining Abrol and MUHAMEDRAHIMOV would meet the claim limitations for the same reasons as previously discussed in claim 1.
Regarding claim 8, Abrol and MUHAMEDRAHIMOV disclose the method of claim 7, wherein the energy transformation model is a first energy transformation model and the transformed image is a first transformed image, and further comprising entering the image as input to a second energy transformation model trained to output a second transformed image at the second energy level, the second energy transformation model selected from among the plurality of energy transformation models based on the second contrast phase (“As another example, if the first medical scanning domain 106 is a particular contrast phase (e.g., one of non-contrast, early arterial phase, late arterial phase, hepatic phase, nephrogenic phase, venous phase, and/or delayed phase), the second medical scanning domain 206 can be a different contrast phase (e.g., a different one of non-contrast, early arterial phase, late arterial phase, hepatic phase, nephrogenic phase, venous phase, and/or delayed phase).” Abrol, [0071]; “In various other embodiments, the machine learning model 202 can be configured and/or trained to transform and/or convert CT images from any electrical energy level to any other electrical energy level. In still other embodiments, the machine learning model 202 can be configured and/or trained to transform and/or convert CT images from any contrast phase to any other contrast phase.” Abrol, [0079])).
Regarding claim 11, Abrol and MUHAMEDRAHIMOV disclose a system, comprising: one or more processors; and memory storing instructions executable by the one or more processors to (“The computer-readable memory 110 can store computer-executable instructions which, upon execution by the processor 108, can cause the processor 108 and/or other components of the learning-based domain transformation system 102 (e.g., receiver component 112, transformation component 114) to perform one or more acts” Abrol, [0062]): obtain an image at a first energy level, the image reconstructed from projection data acquired at a single peak energy level (“the first medical scanning domain 106 can be both an electrical energy level and a contrast phase (e.g., a single CT image can be captured/generated at both a given electrical energy level and a given contrast phase)” Abrol, [0059]; [0022]; [0057]); identify a contrast phase of the image with a contrast phase classifier model (“providing ML models that are automatically trained to classify medical images into contrast states, and/or by increasing accuracy of classification of the ML models. Classification may include a category label indicating the contrast phase” MUHAMEDRAHIMOV, [0052]); select an energy transformation model from among a plurality of energy transformation models based on the contrast phase (“a particular medical scanning domain is preferred and/or desired to capture/generate a medical image” Abrol, [0028]; i.e., Abrol requires a selection of a specific scanning domain such that a different electrical energy level is obtained, wherein “the first medical scanning domain can be any suitable medical scanning domain as desired, such as a first electrical energy level and/or a first contrast phase” Abrol, [0031]); enter the image as input the selected energy transformation model (“the first medical scanning domain can be any suitable medical scanning domain as desired, such as a first electrical energy level and/or a first contrast phase” Abrol, [0031]), the selected energy transformation model trained to output, based on the image, a transformed image at a second energy level, different than the first energy level (“Accordingly, the machine learning model 202 can be configured and/or trained (as described in more detail below with respect to FIGS. 8-9) to transform and/or convert CT scans taken at 70 kVp to CT scans taken at 120 kVp. Thus, in various cases, the transformation component 114 can feed the medical image 104 to the machine learning model 202 as input, and the machine learning model 202 can produce as output the predicted image 204... That is, the second medical scanning domain 206 in this non-limiting example can be an electrical energy level of 120 kVp” Abrol, [0077]; “As an example, if the first medical scanning domain 106 is a particular electrical energy level (e.g., 70 kVp), the second medical scanning domain 206 can be a different electrical energy level (e.g., 120 kVp). As another example, if the first medical scanning domain 106 is a particular contrast phase (e.g., one of non-contrast, early arterial phase, late arterial phase, hepatic phase, nephrogenic phase, venous phase, and/or delayed phase), the second medical scanning domain 206 can be a different contrast phase (e.g., a different one of non-contrast, early arterial phase, late arterial phase, hepatic phase, nephrogenic phase, venous phase, and/or delayed phase).” Abrol, [0071]; see additionally, [0127]), the transformed image in the same contrast phase as the image (“In various other embodiments, the machine learning model 202 can be configured and/or trained to transform and/or convert CT images from any electrical energy level to any other electrical energy level. In still other embodiments, the machine learning model 202 can be configured and/or trained to transform and/or convert CT images from any contrast phase to any other contrast phase.” Abrol, [0079]); and display a final transformed image and/or save the final transformed image in memory, wherein the final transformed image is the transformed image or is generated based on the transformed image(“In various embodiments, the computer-readable memory 110 can store computer-executable components (e.g., receiver component 112, transformation component 114), and the processor 108 can execute the computer-executable components.” Abrol, [0062]; i.e., the final transformed image is generated via the transformation component where “the transformation component can generate, via execution of a machine learning model, a predicted image based on the medical image” Abrol, [0005] may be stored/saved in memory).
Therefore, combining Abrol and MUHAMEDRAHIMOV would meet the claim limitations for the same reasons as previously discussed in claim 1.
Regarding claim 12, Abrol and MUHAMEDRAHIMOV disclose the system of claim 11, wherein the contrast phase comprises one or more of no contrast, a venous phase, a portal phase, an arterial phase, or a delayed phase (“As another example, if the first medical scanning domain is a particular contrast phase (e.g., arterial phase), then the second medical scanning domain can be a different contrast phase (e.g., hepatic phase)” Abrol, [0096]).
Regarding claim 14, Abrol and MUHAMEDRAHIMOV disclose the system of claim 11, wherein training of the contrast phase classifier model comprises: obtaining a plurality of training triads, each training triad including a set of 3 projection images at a respective contrast phase of a plurality of contrast phases; entering a selected training triad from the plurality of training triads as input to the contrast phase classifier model; receiving, from the contrast phase classifier model, one or more predicted contrast phases included in the selected training triad (“each of the plurality of medical images comprises a three (3D) dataset, and further comprising extract an input representation of the 3D dataset, wherein the training dataset includes the input representation, and a target input representation of the target medical image is fed into the trained ML mode” MUHAMEDRAHIMOV, [0018]; “Imaging device 112 provides images, which may be included in training dataset(s) 116 optionally with labels... CT contrast images depicting various phases of injected intravenous (IV) contrast. 3D images, and/or slices of 3D images, and/or 4D images may be converted to 2D images for training and/or inference, for example, by selecting 2D slices from a 3D scan, and/or converting the 3D image into a 2D image such as by maximum pixel intensity (MPI) approaches.” MUHAMEDRAHIMOV, [0072]); comparing the one or more predicted contrast phases to one or more ground truth contrast phases indicated via annotations of the selected training triad (“FIG. 12, which is a confusion matrix for CT abdomen with contrast phase classification by the classifier ML model with description of the test results within each phase, in accordance with some embodiments of the present invention. Results for the results of the classifier ML model are compared to the ground truth.” MUHAMEDRAHIMOV, [0167]); and adjusting model parameters of the contrast phase classifier model based on the comparison (FIG. 13, which provides some examples of abdominal CT scans misclassified by the classifier ML model, during the experiment, in accordance with some embodiments of the present invention. Typically, the misclassification occurred when a scan might be in between phases, indicating that the misclassification may not necessarily be a full error, but may be partially true. The misclassification in between phases may be due to improper timing of capture of the CT scan itself, when the CT scan was taken in between phases rather than during a target phase. When the regressor ML model is used, the misclassification may be learned as a new contrast phase using the zero-shot learning approach described herein. New contrast phases may be created to represent the misclassification category, enabling the classifier ML model to learn the new category, and correctly label new CT scans in between phases into the new category.” MUHAMEDRAHIMOV, [0168]).
Therefore, combining Abrol and MUHAMEDRAHIMOV would meet the claim limitations for the same reasons as previously discussed in claim 1.
Regarding claim 15, Abrol and MUHAMEDRAHIMOV disclose the system of claim 11, wherein training of the energy transformation model comprises: entering a first image of a training image pair to the energy transformation model, the first image at the first energy level; receiving a first transformed training image output from the energy transformation model (“the training component can electronically train the machine learning model on the training dataset to accurately shift, covert, and/or transform the first medical scanning domain 106 to the second medical scanning domain. More specifically, the training dataset can include a set of training images, each of which is captured/generated via the first medical scanning domain 106, and a respectively corresponding set of annotation images, each of which is captured/generated via the second medical scanning domain. Consider a given training image in the set of training images, where the given training image depicts one or more anatomical structures according to the first medical scanning domain 106. In such case, there can be a given annotation image in the set of annotation images that corresponds to the given training image, where the given annotation image depicts the same one or more anatomical structures as the given training image, but where the given annotation image depicts such one or more anatomical structures according to the second medical scanning domain instead of the first medical scanning domain 106. In other words, the given annotation image can be considered as representing how the given training image would actually look if the given training image had been captured/generated via the second medical scanning domain.” Abrol, [0068]); determining a loss function based on the first transformed training image and a second image of the training image pair, the second image at the second energy level; and updating the energy transformation model based on the loss function(“This can cause the machine learning model to generate a predicted output image, where the predicted output image represents the machine learning model’s estimation of how the given training image would look if captured/generated according to the second medical scanning domain. In various instances, the training component can compute an error/loss between the predicted output image and the given annotation image, and the training component can update (e.g., via backpropagation) internal parameters of the machine learning model based on such error/loss” Abrol, [0068]), wherein the first image and the second image are monochromatic images generated from dual-energy projection data (“For example, in some cases, the particular training image can be co-registered with the particular annotation image through simulation, through scanning physical phantoms, and/or through dual-energy scans.” Abrol, [0042]).
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Abrol in view of MUHAMEDRAHIMOV and in further view of Eusemann et al. (“Dual energy CT: How to best blend both energies in one fused image?”, 2008).
Regarding claim 9, the combination of Abrol and MUHAMEDRAHIMOV as a whole does not expressly disclose blending the first transformed image and the second transformed image to generate the final transformed image. However, Eusemann in the same field of endeavor teaches blending the first transformed image and the second transformed image to generate the final transformed image (“optimize the blending of the low and high kV information for display in a way that combines the benefits (contrast and sharpness) of both energies in a single image” Eusemann, abstract).
Therefore, it would have been obvious to one of ordinary skill in the art to combine Abrol, MUHAMEDRAHIMOV and Eusemann before the effective filing date of the claimed invention. The motivation for this combination of references would have been to present image data as a single energy CT image which combines the benefits of contrast and sharpness of both energies (Eusemann, abstract). This motivation for the combination of Abrol, MUHAMEDRAHIMOV and Eusemann is supported by KSR exemplary rationale (G) Some teaching, suggestion, or motivation in the prior art that would have led one of ordinary skill to modify the prior art reference or to combine prior art reference teachings to arrive at the claimed invention. MPEP 2141 (III).
Claim 13 is rejected under 35 U.S.C. 103 as being unpatentable over Abrol in view of MUHAMEDRAHIMOV and in further view of Charyyev et al. (“Learning-based synthetic dual energy CT imaging from single energy CT for stopping power ratio calculation in proton radiation therapy”, 2022).
Regarding claim 13, the combination of Abrol and MUHAMEDRAHIMOV as a whole does not expressly disclose wherein the first energy level is greater than the second energy level. However, Charyyev in the same field of endeavor teaches wherein the first energy level is greater than the second energy level (“machine learning based method to synthesize DECT from SECT... our method synthesized both LECT and HECT from SECT” Charyyev, pg. 8 Col 2; wherein Dual Energy CT is (DECT), Single energy CT is (SECT), High Energy CT is (HECT), and Low Energy CT is (LECT) i.e., the first energy level of a SECT is a HECT to synthesize a LECT).
Therefore, it would have been obvious to one of ordinary skill in the art to combine Abrol, MUHAMEDRAHIMOV and Charyyev before the effective filing date of the claimed invention. The motivation for this combination of references would have been to provide both Low Energy CT and High Energy CT scans from Single Energy CT imaging modalities (Charyyev, pg. 8 Col 2). This motivation for the combination of Abrol, MUHAMEDRAHIMOV and Charyyev is supported by KSR exemplary rationale (G) Some teaching, suggestion, or motivation in the prior art that would have led one of ordinary skill to modify the prior art reference or to combine prior art reference teachings to arrive at the claimed invention. MPEP 2141 (III).
Allowable Subject Matter
Claim 3, 5, 10, 16 and 21 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten to overcome any 35 U.S.C. 112(b) rejections stated above and in independent form including all of the limitations of the base claim and any intervening claims.
Claims 17, 19 and 20 are allowed.
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.
Inquiries
Any inquiry concerning this communication or earlier communications from the examiner should be directed to EMMANUEL SILVA-AVINA whose telephone number is (571)270-0729. The examiner can normally be reached Monday - Friday 11 AM - 8 PM EST.
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, Chineyere Wills-Burns can be reached at (571) 272-9752. 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.
/EMMANUEL SILVA-AVINA/Examiner, Art Unit 2673
/CHINEYERE WILLS-BURNS/Supervisory Patent Examiner, Art Unit 2673
Prosecution Insights