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 .
Response to Arguments
Applicant's arguments filed March 12, 2026 have been fully considered but they are not persuasive.
Claim 95 has been amended by the present amendment to incorporate the subject matter of claim 108 into claim 95 and claim 108 has been canceled. In the previous nonfinal Office Action, claim 108 was rejected as being unpatentable over Huang et al. In arguing the patentability of amended claim 95 over Huang et al., Applicant argues “Huang fails to disclose the feature 'determining, by the processor, a hotspot-specific threshold value for the particular hotspot based on (i) the corresponding hotspot intensity and (ii) at least one of the one or more reference value(s),’ as recited in claim 95.” Specifically, Applicant argues that paras. [0115]-[0118] of Huang et al., which were cited by the examiner as teaching these this limitation, “describe a procedure whereby a Gaussian function fit is used to determine a threshold value, but the threshold that Huang determines via this approach is based solely on parameters of the Gaussian fit, according to the equation T = y(b - Kc). Huang at [0117]. None of the parameters in Huang's equation at paragraph [0117] are or could be interpreted as a reference value as recited in claim 95.”
Regarding the term “reference value”, the BRI for the limitation “determining one or more reference values, each based on a measure of intensities of voxels of the 3D functional image located within a particular reference volume corresponding to a particular reference tissue region” was set forth in the previous Office Action as follows: “the BRI for this limitation, based on para. [0096] of the present specification, is that different threshold values, i.e., reference values, are determined for particular regions because different regions such as different organs have different ranges of SUV values; for example, the range of SUV values that indicate hotspot candidates for the aorta region is different from the range of SUV values that indicate hotspot candidates for the liver region”.
It should be noted that Applicant has not disputed the examiner’s BRI, but argues that “[t]he region-specific thresholds that Huang describes at paragraph [0114] are themselves thresholds. They are not used, together with a hotspot intensity, to determine a hotspot-specific threshold value that is then used to segment a particular hotspot, as recited in claim 95.”
The above BRI is precisely what Huang et al. discloses in para. [0114] of Huang:
“In PET images, because FDG uptakes by different organs or tissues have large variations, a global thresholding on the converted SUV volumes often fails to provide good hot-spot candidates. For example, tumors in the lung may have lower SUV values than normal tissue in the liver. Hence to automatically generate good hot-spot candidates, organ-specific or region-specific thresholding is very attractive. Using a whole-body context according to an embodiment of the invention, one can first detect, segment and separate organs or regions that have different ranges in SUV values, then apply organ- or region-specific thresholding. For instance, one can first detect and segment the lung, the liver, and other organs, or detect and separate the thorax and abdomen regions. Then a threshold is chosen for the lung, such as the mean SUV value in the lung region, and hot-spot candidates can be generated in the lung which have SUV values above the threshold. A different threshold can be chosen for the liver, such as the mean SUV value in the liver, and hot-spot candidates can be generated in the liver.”
Therefore, the examiner properly interpreted the region-specific threshold values of Huang as constituting “reference values”.
Applicant argues further that the region-specific threshold values of Huang et al. “are not used, together with a hotspot intensity, to determine a hotspot-specific threshold value that is then used to segment a particular hotspot, as recited in claim 95.” It should be noted that amended claim 1 is not specific regarding how the reference values or the hotspot intensity are used to determine a hotspot-specific threshold value. Amended claim 1 recites only that the hotspot-specific threshold value is determined “based on” the corresponding hotspot intensity and at least one of the one or more reference values.
This is taught by Huang et al. In Huang et al., the region-specific threshold values are determined for the different regions before the adaptive thresholding phase takes place. Once determined, the region-specific thresholds are used to detect and segment the specific regions. Then, during the adaptive thresholding phase, experts delineate the segmented specific regions and the adaptive thresholds are generated and used to perform hotspot segmentation within the specific regions. (Huang, paras. [114]-[0115]). During the adaptive thresholding process, a Gaussian function is used to determine the adaptive threshold values, but they are nonetheless determined “based on” the region-specific threshold reference values and the hotspot intensity values because the region-specific threshold values are used to first segment the specific regions and then hotspot intensities within the specific regions are used with the Gaussian function to determine the adaptive threshold values.
Applicant argues further that “Huang fails to disclose, or even suggest, a hotspot-specific threshold that is a function of both a reference intensity value and an intensity of the corresponding hotspot, as recited in claim 95.” Claim 1 does not recite determining a hotspot-specific threshold that is a function of both a reference intensity value and an intensity of the corresponding hotspot. Paras. [0096]-[0098] of the present specification recite such functions, such as a variable percentage that decreases with hotspot intensity, but this is not recited in amended claim 1. Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993).
Claim Interpretation
The claims in this application are given their broadest reasonable interpretation (BRI) using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. Some of the terms in the claims have been given the BRIs in light of the specification. Should applicant wish different interpretations, Applicant should point to the portions of the specification that clearly support a different interpretation.
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)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claims 95-101 and 122 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by U.S. Publ. Appl. No. 2007/0081712 A1 to Huang et al. (hereinafter referred to as “Huang et al.”).
Regarding claim 95, Huang et al. discloses a method for automatically processing 3D images (paras. [0061] discussing the term “image” as meaning 2D or 3D images; para. [0125] discussing automatically processing PET images and CT images) of a subject to identify and/or characterize cancerous lesions within the subject (para. [0136]-[0144], Figs. 16-19 discuss the method being used to perform cancerous lesion detection), the method comprising:
receiving, by a processor of a computing device, a 3D functional image of the subject obtained using a functional imaging modality (Fig. 16, step 162 and para. [0125] discusses receiving the PET image, which is a 3D functional image obtained using the PET functional imaging modality; para. [0149], Fig. 19, the CPU 192 is the processor of computing device 191 that receives and processes the PET image);
receiving, by the processor, a 3D anatomical image of the subject obtained using an anatomical imaging modality (Fig. 16, step 161 and para. [0125] discusses receiving the CT image, which is a 3D anatomical image obtained using the CT anatomical imaging modality; para. [0149], Fig. 19, the CPU 192 is the processor of computing device 191 that receives and processes the CT image);
receiving, by the processor, a 3D segmentation map identifying one or more particular tissue region(s) or group(s) of tissue regions within the 3D functional image and/or within the 3D anatomical image (paras. [0111] and [0112] discuss translating observed Voxel standard uptake value (SUV) values into different state values corresponding to particular tissue regions, such as for lungs, liver and kidney; para. [0112], the voxel is classified and labeled according to whichever state is dominant for the particular Voxel and the label is assigned to a segmentation map; since Voxels represent points in 3D space, the resulting segmentation map is a 3D segmentation map);
automatically detecting, by the processor, using one or more machine learning module(s), a set of one or more hotspots within the 3D functional image, each hotspot corresponding to a local region of elevated intensity with respect to its surrounding and representing a potential cancerous lesion within the subject (paras. [0014] and [0062] discusses one or more machine learning models that are trained and executed by CPU 192 discussed in paras. [0148]-[0149]; para. [0084], Fig. 8, steps 81 and 82 represent the process performed by one or more trained machine learning modules to detect hotspots, which are local regions of high intensity with respect to their surroundings representing potential cancerous lesions within the subject; the high intensity corresponds to a high SUV value, which is determined at step 81; see also Fig. 18, step 181 detecting hotspot regions having high uptake values within the PET 3D functional image) thereby creating a preliminary 3D hotspot map, identifying, for each hotspot, a corresponding preliminary 3D hotspot volume within the 3D functional image (as indicated above, paras. [0111] and [0112] discuss generating the 3D hotspot segmentation map based on the automatic hotspot detection; the 3D hotspot segmentation map and 3D hotspot volume are preliminary in that they are later refined during hotspot segmentation using adaptive thresholding to generate hotspot-specific threshold values, as discussed in paras. [00114]-[0118]),
wherein at least one of the one or more machine learning module(s) receives, as input (i) the 3D functional image, (ii) the 3D anatomical image, and (iii) the 3D segmentation map (the segmentation map is used by the machine learning model implemented by the CPU 192 in step 185 of Fig. 18 to compute the likelihood that a hotspot region corresponds to lesion tissue of a particular type, para. [0139]);
for each hotspot in the preliminary 3D hotspot map, segmenting, by the processor, the hotspot using an adaptive thresholding method to create a refined 3D hotspot map (sections entitled “Organ/Region-Specific Thresholding for Automatic Generation of Hot-Spot Candidates” and “Adaptive Threshold Selection and Segmentation of Hot-Spots”) by:
determining, by the processor, one or more reference values, each based on a measure of intensities of voxels of the 3D functional image located within a particular reference volume corresponding to a particular reference tissue region (the BRI for this limitation, based on para. [0096] of the present specification, is that different threshold values, i.e., reference values, are determined for particular regions because different regions such as different organs have different ranges of SUV values; for example, the range of SUV values that indicate hotspot candidates for the aorta region is different from the range of SUV values that indicate hotspot candidates for the liver region; Huang et al., para. [0114], discloses detecting and segmenting different tissue regions and determining respective threshold values that are region-specific based on the intensities of the corresponding voxels in the regions; these region-specific threshold values constitute reference values, each based on a measure of intensities of voxels of the 3D functional image located within a particular reference volume corresponding to a particular reference tissue region); and
for each particular preliminary 3D hotspot volume of the preliminary 3D hotspot map:
determining, by the processor, a corresponding hotspot intensity based on intensities of voxels within the particular preliminary 3D hotspot volume (paras. [0111]-[0112] disclose determining voxel SUV values, which correspond to hotspot intensities for particular hotspot volumes);
determining, by the processor, a hotspot-specific threshold value for the particular hotspot based on (i) the corresponding hotspot intensity and (ii) at least one of the one or more reference value(s) (paras. [0115]-[0118] discuss adaptively determining hotspot-specific threshold values based on the hotspot intensity and based on the corresponding reference value; para. [0114], the region-specific threshold values are determined for the different regions before the adaptive thresholding phase takes place; once determined, the region-specific thresholds are used to detect and segment the specific regions, thereby generating the preliminary 3D hotspot map; paras. [0115]-[0118], after the preliminary 3D hotspot map has been generated using the region-specific threshold reference values, then, during the adaptive thresholding phase, experts further delineate the segmented specific regions of the preliminary 3D hotspot map and the adaptive thresholds are generated and used to perform hotspot segmentation within the specific regions to generate the refined 3D hotspot maps; during the adaptive thresholding process, a Gaussian function is used to determine the adaptive threshold values, but they are nonetheless determined “based on” the region-specific threshold reference values and the hotspot intensity values because the region-specific threshold values are used to first detect and segment the specific regions and then hotspot intensities within the specific regions are used with the Gaussian function to determine the adaptive threshold values); and
segmenting, by the processor, the hotspot using the hotspot- specific threshold value, thereby determining a refined hotspot volume for inclusion in the refined 3D hotspot map (paras. [0115]-[0118] disclose using the hotspot-specific threshold values to segment the hotspots to determine the refined hotspot volume for inclusion in the refined 3D hotspot map); and
storing and/or providing, for display and/or further processing, the refined 3D hotspot map (para. [0149], Fig. 19, the system includes memory 197 and display 195; Fig. 15(a) shows a tumor hotspot displayed; para. [0077], Fig. 6, step 64, the segmentation results are provided to the user via a user interface to allow the user to make manual adjustments; see also para. [0078] discussing “providing a visualization of the segmented results. VOIs returned from different time points can be rendered as surfaces in different colors, or the segmented volumes can be rendered together with the fused PET/CT image for visual inspection and comparison”).
Regarding claim 96, Huang et al. discloses:
receiving, by the processor, an initial 3D segmentation map that identifies one or more particular tissue regions within the 3D anatomical image and/or the 3D functional image (as indicated above, paras. [0111] and [0112] discuss generating the 3D segmentation map based on the automatic hotspot detection and segmentation; the segmentation map identifies particular tissue regions based on their SUVs of the identified hotspot regions); and
identifying, by the processor, at least a portion of the one or more particular tissue regions as belonging to a particular one of one or more tissue grouping(s) (para. [0111], particular tissue regions, such as lung, liver, kidney regions, are identified based on their SUVs: “[f]or lungs, SUVs higher than 2.0 are considered abnormal, while for liver and kidney, SUVs higher than 3.0 are considered abnormal”) and updating, by the processor, the 3D segmentation map to indicate the identified particular regions as belonging to the particular tissue grouping (para. [0111] discusses the process of identifying hotspot regions as belonging to particular tissue regions to generate the segmentation map as an initialization step; para. [0112] discusses the process of updating the segmentation map after the initialization process has been performed); and
using, by the processor, the updated 3D segmentation map as input to at least one of the one or more machine learning modules (para. [0113] discusses fine tuning of the segmentation map in the machine learning model by using the updated segmentation map in a mode-seeking region growing algorithm to identify SUV-maxima points “that correspond to the primary locations of true hot-spots”).
Regarding claim 97, Huang et al. discloses that the one or more tissue groupings comprise a soft-tissue grouping, such that particular tissue regions that represent soft-tissue are identified as belonging to the soft-tissue grouping (para. [0111], e.g., hotspots corresponding to lung tissue are labeled as belonging to a particular soft-tissue grouping, namely, lung tissue).
Regarding claim 98, Huang et al. discloses that the one or more tissue groupings comprise a bone tissue grouping, such that particular tissue regions that represent bone are identified as belonging to the bone tissue grouping (para. [0124]-[0128], Fig. 17, e.g., segmented hotspots corresponding to bone tissue are labeled as belonging to a particular bone tissue grouping).
Regarding claim 99, Huang et al. discloses that the one or more tissue groupings comprise a high-uptake organ grouping, such that one or more organs associated with high radiopharmaceutical uptake are identified as belonging to the high uptake grouping (as indicated above, para. [0111] discloses that different organs such as lungs, liver and kidneys are identified as belonging to particular high radiopharmaceutical uptake groupings; the radiopharmaceutical used in Huang et al. is fluorine-18 deoxyglucose; see also para. [0073] discussing different organs associated with high uptake hotspots).
Regarding claim 100, Huang et al. discloses for each detected and/or segmented hotspot, determining, by the processor, a classification for the hotspot (as indicated above, para. [0112] discloses that the voxel is classified and labeled according to whichever state is dominant for the particular Voxel and the label is assigned to a segmentation map).
Regarding claim 101, Huang et al. discloses that the method comprises using at least one of the one or more machine learning modules to determine, for each detected and/or segmented hotspot, the classification for the hotspot (the CPU 192 of Huang et al. executes one or more machine learning modules, including the spatial Hidden-Markov model (HMM) and a Competition-Diffusion (CD) segmentation module that classify and label hotspots, as discussed in paras. [0014] and [0111]-[0113]).
Regarding claim 122, to the extent that claim 122 recites the same limitations that are recited in claim 95, the rejection of claim 95 applies mutatis mutandis to claim 122. The only limitations that are recited in claim 122 that are not also recited in claim 95 are a processor of a computing device and memory having instructions stored thereon for execution by the processor to cause it to perform the steps recited in claims 95 and 122. Huang et al. discloses a processor and memory (Fig. 19, CPU 192 and memory 197) storing instructions for execution by the processor for performing the operations recited in claims 95 and 122 (paras. [0148]-[0149]).
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 102-105 are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. in view of U.S. Publ. Appl. No. 2019/0209116 A1 to Sjostrand et al. (hereinafter referred to as “Sjostrand et al.”).
Regarding claim 102, Huang et al. discloses that one or more machine learning modules comprise a full body lesion detection module that detects hotspots throughout an entire body (para. [0060]: “[e]xemplary embodiments of the invention as described herein generally include systems and methods for whole-body landmark detection, segmentation, and change quantification in single-mode and multi-mode medical images”), but does not explicitly disclose a prostate lesion module that detects hotspots within the prostate.
Sjostrand et al., in the same field of endeavor, discloses a prostate lesion module that detects and segments hotspots within the prostate (para. [0015]: “the image analysis approaches described herein utilize convolutional neural networks (CNNs) to accurately identify a prostate volume within the CT image that corresponds to the prostate of the subject. The identified prostate volume can be used to identify those voxels of the SPECT image that also correspond to the subject's prostate. Uptake metrics that provide a measure of uptake of the imaging agent (e.g., a labelled PSMA binding agent, e.g., .sup.99mTc-MIP-1404 or [.sup.18F]DCFPyL) in the prostate can thus be computed using the intensities of SPECT image voxels corresponding to the prostate of the subject”).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the present disclosure, to modify the systems and methods of Huang et al. to include as part of the whole-body detection and segmentation process of Huang et al. detection of the prostate as taught by Sjostrand et al. One of ordinary skill in the art would have been motivated to make the modification to provide the systems and methods of Huang et al. with the ability to identify cancerous lesions in the prostate to allow diagnosis and treatment. The modification could have been made by one of ordinary skill in the art before the effective filing date of the present disclosure with a reasonable expectation of success because making the modification merely involves extending the methodologies disclosed in Huang et al. for detecting cancerous lesions in the lungs, liver and kidneys to detecting cancerous lesions in the prostate (e.g., modifying the software executed by the CPU 192 of Huang et al. to implement the machine learning algorithms of Sjostrand et al.) to yield predictable results.
Regarding claim 103, the BRI for “merging the results”, based on para. [0253] of the present specification, is that it means combining the hotspot map information with other output such as detection/segmentation results obtained through other imaging modes. As indicated above in the rejection of claim 106, in Huang et al., the locations of the hotspots within the anatomy are determined because the anatomical image is aligned and registered with the functional image, as described in para. [0011]. The classified hotspots together with their locations comprise a hotspot map. Huang et al. discloses combining this hotspot map information with the anatomical whole-body information (Fig. 16, para. [0125]) to obtain a reshaped CT volume that includes the PET hotspot location and classification information. However, as indicated above, Huang et al. does not explicitly disclose that the hotspot location and classification information includes prostate lesion hotspot location and classification information.
As indicated above in the rejection of claim 102, Sjostrand et al. discloses a prostate lesion module that detects and segments hotspots within the prostate (para. [0015]: “the image analysis approaches described herein utilize convolutional neural networks (CNNs) to accurately identify a prostate volume within the CT image that corresponds to the prostate of the subject. The identified prostate volume can be used to identify those voxels of the SPECT image that also correspond to the subject's prostate. Uptake metrics that provide a measure of uptake of the imaging agent (e.g., a labelled PSMA binding agent, e.g., .sup.99mTc-MIP-1404 or [.sup.18F]DCFPyL) in the prostate can thus be computed using the intensities of SPECT image voxels corresponding to the prostate of the subject”).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the present disclosure, to modify the systems and methods of Huang et al. to include as part of the whole-body detection and segmentation process of Huang et al. detection and segmentation of the prostate as taught by Sjostrand et al. and to merge the corresponding hotspot location and classification information obtained for the prostate with the anatomical whole-body information as taught by Huang et al. One of ordinary skill in the art would have been motivated to make the modification to provide the systems and methods of Huang et al. with the ability to render a whole-body anatomical image that includes the prostate hotspot locations and classifications. Doing so would facilitate a healthcare provider in diagnosing prostate cancer and prescribing treatment. The modification could have been made by one of ordinary skill in the art before the effective filing date of the present disclosure with a reasonable expectation of success because making the modification merely involves extending the methodologies disclosed in Huang et al. for detecting cancerous lesions in the lungs, liver and kidneys to also detecting cancerous lesions in the prostate (e.g., modifying the software executed by the CPU 192 of Huang et al. to implement the machine learning algorithms of Sjostrand et al.) to yield predictable results.
Regarding claims 104, Huang et al. discloses step (d) of claim 95 comprises:
segmenting and classifying the set of one or more hotspots to create a labeled 3D hotspot map that identifies, for each hotspot, a corresponding 3D hotspot volume within the 3D functional image and in which each hotspot volume is labeled as belonging to a particular hotspot class of a plurality of hotspot classes by:
using a machine learning module to segment a first initial set of one or more hotspots within the 3D functional image, thereby creating a first initial 3D hotspot map that identifies a first set of initial hotspot volumes, wherein the first machine learning module segments hotspots of the 3D functional image according to a single hotspot class (the BRI for this limitation, based on para. [0269] of the present specification, is that it means identifying image regions as corresponding to hotspots or not, with the regions that are identified as hotspots being of a single hotspot class; in Huang et al., a first initial set of candidate hotspots is generated by a trained machine learning module that segments hotspots and identifies them as either normal hotspots or pathological hotspots and excludes the normal hotspots from the set, para. [0099]);
a machine learning module to segment a second initial set of one or more hotspots within the 3D functional image, thereby creating a second initial 3D hotspot map that identifies a second set of initial hotspot volumes, wherein the second machine learning module segments the 3D functional image to according to the plurality of different hotspot classes, such that the second initial 3D hotspot map is a multi-class 3D hotspot map in which each hotspot volume is labeled as belonging to a particular one of the plurality of different hotspot classes (the BRI for this limitation, based on para. [0270] of the present specification, is that it means performing segmentation of hotspots, classifying the hotspots and labeling them as belonging to one of multiple classes, such as the classes of lymph, bone or prostate; Huang et al. discloses segmenting a second initial set of hotspots, classifying them as belonging to one of a plurality of classes (e.g., lungs, liver kidney), and labeling them accordingly, paras. [0111]-[0112]); and
merging, by the processor, the first initial 3D hotspot map and the second initial 3D hotspot map by, for at least a portion of the hotspot volumes identified by the first initial 3D hotspot map:
identifying a matching hotspot volume of the second initial 3D hotspot map, the matching hotspot volume of the second 3D hotspot map having been labeled as belonging to a particular hotspot class of the plurality of different hotspot classes (the BRI for this limitation, based on para. [0271] of the present specification, is that it means using some type of spatial correlation between the hotspot maps of the first and second initial sets to determine which hotspots of the first initial set correspond to hotspots of the second initial set; Huang et al. discloses performing a Competition-Diffusion (CD) algorithm that determines which hotspots of the first initial set correspond to hotspots of the second initial set by determining which pathological hotspots correspond to hotspots of particular classes, para. [0110]); and
labeling the particular hotspot volume of the first initial 3D hotspot map as belonging to the particular hotspot class, thereby creating a merged 3D hotspot map that includes segmented hotspot volumes of the first 3D hotspot map having been labeled according classes that matching hotspot volumes of the second 3D hotspot map are identified as belonging to (the CD algorithm of Huang et al. labels the particular hotspot volumes as belonging to one of the particular classes, e.g., lung, liver, kidney, paras. 0111]-[0112]; a merged 3D hotspot map is created in Huang et al., through performance of a region growing process that fine tunes the 3D hotspot map that detects SUV-maxima points corresponding to the primary locations of the “true” hotspots and uses them as seed points from which to grow the hotspots to delineate their boundaries, para. [0113]); and
step (f) comprises storing and/or providing, for display and/or further processing, the merged 3D hotspot map (para. [0149], Fig. 19, the system includes memory 197 and display 195; Fig. 15(a) shows a tumor hotspot displayed; para. [0077], Fig. 6, step 64, the segmentation results are provided to the user via a user interface to allow the user to make manual adjustments; see also para. [0078] discussing “providing a visualization of the segmented results. VOIs returned from different time points can be rendered as surfaces in different colors, or the segmented volumes can be rendered together with the fused PET/CT image for visual inspection and comparison”).
Huang et al. does not explicitly disclose that first and second machine learning modules are used for segmenting the first and second initial sets of hotspots. However, using multiple machine learning modules in combination is well known in the art.
Sjostrand et al. discloses using a combination of first and second machine learning modules, namely, convolutional neural networks (CNNs) to (1) segment a volume of interest (VOI) (2) identifies the prostate volume within the VOI, respectively, to save computational resources by reducing input size thereby allowing more resources to be allocated for improving accuracy and processing speed (para. [0021]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the present disclosure, to modify the system of Huang et al. to use first and second machine learning models to perform segmentation of the first and second hotspot sets, respectively, as taught by Sjostrand et al. One of ordinary skill in the art would have been motivated to make the modification to improve the accuracy and the processing speed of the system of Huang et al. The modification could have been made by one of ordinary skill in the art before the effective filing date of the present disclosure with a reasonable expectation of success because making the modification merely involves modifying the software executed by the CPU 192 of Huang et al. to implement the first and second machine learning modules instead of a single machine learning module to yield predictable results.
Regarding claim 105, Huang et al. discloses that the different hotspot classes include bone hotspots, determined to represent lesions located in bone (para. [0014]), and lymph hotspots, determined to represent lesions located in lymph nodes (para. [0071]), but does not explicitly disclose that one of the hotspot classes can be prostate. Sjorstrand et al discloses that one of the hotspot classes can be prostate (para. [0012]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the present disclosure, to modify the systems and methods of Huang et al. to include as part of the whole-body detection and segmentation process of Huang et al. detecting, segmenting, classifying and labeling hotspots of the prostate class as taught by Sjostrand et al. One of ordinary skill in the art would have been motivated to make the modification to provide the systems and methods of Huang et al. with the ability to render a whole-body anatomical image that includes the prostate hotspot locations and classifications. Doing so would facilitate a healthcare provider in diagnosing prostate cancer and prescribing treatment. The modification could have been made by one of ordinary skill in the art before the effective filing date of the present disclosure with a reasonable expectation of success because making the modification merely involves extending the methodologies disclosed in Huang et al. for detecting cancerous lesions in the lungs, liver and kidneys to also detecting cancerous lesions in the prostate (e.g., modifying the software executed by the CPU 192 of Huang et al. to implement the machine learning algorithms of Sjostrand et al.) to yield predictable results.
Claim 109 is rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. in view of U.S. Publ. Appl. No. 2016/0350947 A1 to Kelly (hereinafter referred to as “Kelly”).
With regard to claim 109, as indicated above in the rejection of claim 108, Huang et al. discloses determining hotspot-specific threshold values based on the corresponding hotspot intensity and based on a Gaussian curve to which a cross-section of the hotspot is fitted. The selected Gaussian curve is selected from a plurality of Gaussian curves of different scales (para. [0118]). However, Huang et al. does not explicitly disclose that the threshold function is selected based on a comparison of the corresponding hotspot intensity with the regions-specific threshold reference value. The BRI for this limitation, based on paras. [0285]-[0286] of the present specification, is that the hotspot intensity value, such as the maximum SUV value of the hotspot, is compared to the reference threshold value for the corresponding region, and based on the comparison, a hotspot-specific threshold is set equal to some mathematic function of the reference threshold value, such as a percentage multiplied by the reference threshold value.
Kelly, in the same field of endeavor, discloses calculating an adaptive threshold value by, for example, multiplying a percentage value by a difference between the maximum SUV value of the hotspot and the mean SUV of a region of interest (para. [0047]). As indicated in Huang et al., the mean SUV value for a region, e.g., the liver region, the lung region, etc., can be used as the region-specific threshold reference value (para. [0114]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the present disclosure, to modify the systems and methods of Huang et al. to obtain the hotspot-specific threshold value by comparing the maximum SUV value to the region-specific mean SUV reference value and multiplying the difference by a percentage as taught by Kelly. One of ordinary skill in the art would have been motivated to make the modification to improve the accuracy of hotspot segmentation, classification and labeling. Doing so would facilitate a healthcare provider in diagnosing cancer and prescribing treatment. The modification could have been made by one of ordinary skill in the art before the effective filing date of the present disclosure with a reasonable expectation of success because making the modification merely involves modifying the software executed by the CPU 192 of Huang et al. to perform adaptive thresholding in this manner to yield predictable results.
Allowable Subject Matter
Claim 110 is objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Regarding claim 110, none of the prior art of record teaches or suggests the following limitation in combination with the other limitations recited in claims 95, 107 and 108: an adaptive thresholding method that uses region-specific threshold values and hotspot-specific threshold values wherein the hotspot-specific threshold value is determined as a variable percentage of the corresponding hotspot intensity, wherein the variable percentage decreases with increasing hotspot intensity.
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 DANIEL J SANTOS whose telephone number is (571)272-2867. The examiner can normally be reached M-F 9-5.
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, Matt Bella can be reached at (571)272-7778. 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.
/DANIEL J. SANTOS/Examiner, Art Unit 2667
/MATTHEW C BELLA/Supervisory Patent Examiner, Art Unit 2667
Prosecution Insights