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 .
DETAILED ACTION
The United States Patent & Trademark Office appreciates the application that is submitted by the inventor/assignee. The United States Patent & Trademark Office reviewed the following application and has made the following comments below.
Priority
This application claims benefit of foreign priority under 35 U.S.C. 119(a)-(d) of GB 2205748.3, filed in the United Kingdom on 04/20/2022, and PCT/EP/2023/060309, filed in European Patent Office on 04/20/2023.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 02/06/2025 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Preliminary Amendment
Applicant submitted a preliminary amendment on 10/02/2024. The Examiner acknowledges the amendment and has reviewed the claims accordingly.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claim 25 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter as follows.
Claim 25 is directed to “a computer-readable storage medium.” The broadest reasonable interpretation of a claim drawn to a “computer-readable storage medium” typically covers forms of non-transitory tangible media and transitory propagating signals per se in view of the ordinary and customary meaning of computer readable media, particularly when the specification is silent. See Subject Matter Eligibility of Computer Readable Media, 1351 OG 212 (26 Jan 2010). See MPEP 2111.01. Signals are nothing by the physical characteristics of a form of energy, and as such is non-statutory phenomena. See, e.g., In re Nuitjen’s is not a process, machine, manufacture, or composition of matter.”…Thus, such a signal cannot be patentable subject matter.”) Accordingly, Claim 25 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. It is suggested that Applicant amend the claim by inserting the term “non-transitory” before “computer-readable” in the preamble of the claim.
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 1-8, 11, 17, 20-21 and 27 are rejected under 35 U.S.C. 103(a) as being unpatentable over Lavdas et al. (NPL “Fully automatic, multiorgan segmentation in normal whole body magnetic resonance imaging (MRI), using classification forests (CFs), convolutional neural networks (CNNs), and a multi-atlas (MA) approach,” 2017, hereafter referred to as Lavdas) in view of Ceranka et al. (NPL “Multi‐atlas segmentation of the skeleton from whole‐body MRI—Impact of iterative background masking,” 2019, hereafter referred to as Ceranka) in view of Jeraj et al. (U.S. Patent Pub. No. 2018/0330495 A1, hereafter referred to as Jeraj) in further view of Sabuncu et al. (NPL “A Generative Model for Image Segmentation Based on Label Fusion,” 2010, hereafter referred to as Sabuncu).
Regarding Claim 1, Lavdas teaches (1 Introduction, 2.A. Healthy volunteers and imaging protocol, Lavdas teaches automatically segmenting major bones and organs in whole body diffusion-weighted MRI (DWI).), the method including: providing the diffusion weighted MRI 3D patient image of the human or animal body (2.A. Healthy volunteers and imaging protocol, Lavdas teaches fifty-one healthy volunteers (24 male—mean age = 37, range = 23–67 yr and 27 female—mean age = 39, range = 23–68 yr) were scanned with whole body MRI from February 2012 to May 2014. Axial slices were acquired during free breathing for diffusion-weighted MRI (DWI) (b = 0, 150, 400, 750, and 1000 s/mm2), while breath-holds were employed for the three first stations for anatomical imaging.), the patient image being formed of plural slices stacked along a direction of the body (2.A. Healthy volunteers and imaging protocol, Table 1, Lavdas teaches acquiring axial slices during free breathing for DWI. Table 1 shows the number of slices/thickness/distance (mm) for each imaging protocol. Specifically, see Column 2 of Table 1 for SS SE EPI imaging protocol.); analysing the patient image to identify different contiguous anatomical regions of the body (3 Results, Lavdas teaches automatic segmentation of the major organs (lungs, kidneys, liver, and spleen) and bones (spine and femurs) are segmented on the images. The Examiner interprets organs and bones to be “contiguous regions.”), the regions being distributed along said direction such that each slice of the patient image is allocated to a respective region (3 Results, Fig. 2, Lavdas teaches automatic segmentation in both the coronal (top row) and axial (bottom row) planes. See Fig. 2 below.);
PNG
media_image1.png
708
787
media_image1.png
Greyscale
providing an atlas of diffusion weighted MRI 3D ground truth images of plural other corresponding bodies containing the regions (2.D. Multi-atlas (MA) algorithm, Lavdas teaches using a set of atlases (images with corresponding segmentations) that represent the intersubject variability of the anatomy to be segmented. The anatomies of interest were manually segmented.), each ground truth image being formed of plural slices stacked along a corresponding direction of the respective body (2.A. Healthy volunteers and imaging protocol, 2.D. Multi-atlas (MA) algorithm, Lavdas teaches axial slices were acquired during free breathing for diffusion weighted imaging (DWI). The anatomies of interest were manually segmented on the T2-weighted volumes, which serve as the “ground truth” images.), the different regions of the body being pre-identified for each ground truth image such that each slice of that ground truth image is allocated to a respective region, and one or more parts of the body of each ground truth image being pre-delineated (2.D. Multi-atlas (MA) algorithm, Fig. 1, Lavdas teaches performing manual segmentation and annotation of the anatomies of interest on the T2-weighted volumes. The Examiner interprets “delineated” and “segmented” to be synonymous.);
PNG
media_image2.png
306
463
media_image2.png
Greyscale
registering each ground truth image to the patient image (2.D. Multi-atlas (MA) algorithm, Fig. 1, Lavdas teaches each atlas is registered to a new image to be segmented using a deformable image registration.)
PNG
media_image3.png
182
322
media_image3.png
Greyscale
(2.D. Multi-atlas (MA) algorithm, Lavdas teaches using an efficient 3D-3D intensity-based image registration with free-form deformations as the transform model. Majority voting is used to derive the final tissue label at each voxel. T1w and DWI volumes were registered to the T2w volumes using an affine transformation.) (3 Results, Fig. 2, 2.C. Convolutional neural networks (CNNs) algorithm, Lavdas teaches automatically segmenting the patient image using machine learning algorithms (CF, CNN, MA). The Examiner interprets the CNN segmented image of major organs and bones to be a “probability image” for the corresponding parts of the body of the patient image since the CNN segmentation output represents a prediction for the most likely class label for each voxel in the image.),
Lavdas does not explicitly disclose the following:
A computer-implemented method; translating and stretching each ground truth image in its corresponding direction to align the identified regions of that ground truth image with the corresponding identified regions of the patient image, identifying a transformation of each registered ground truth image that matches the pre-delineated parts of the body of that registered ground truth image to the corresponding parts of the body of the patient image by minimising a cost function and wherein the probability image combines the pre- delineated parts of the bodies of the ground truth images transformed according to their respective non-linear deformable transformations and weighted according to their respective cost-functions.
Ceranka is in the same field of art of performing segmentation to automatically identify one or more parts of a body within an MRI image. Further, Ceranka teaches identifying a transformation of each registered ground truth image that matches the pre-delineated parts of the body of that registered ground truth image to the corresponding parts of the body of the patient image (2.4 Registration, Ceranka teaches obtaining transformation maps to deform each atlas label image onto a target whole-body image using the inverse transformation.
T
μ
n
-
1
.) by minimising a cost function (2.4 Registration, Ceranka teaches the registration process of multi-atlas segmentation is typically defined as the optimization problem over the parameters
μ
n
of the spatial transformation T, guided by the cost function C. The transformation was redefined from atlases an to the target t, see the cost function below. “argmin” indicates that the cost function is minimized.)
PNG
media_image4.png
49
264
media_image4.png
Greyscale
and wherein the probability image combines the pre- delineated parts of the bodies of the ground truth images transformed according to their respective non-linear deformable transformations (1 Introduction, 2.5 Label Fusion Strategies, Ceranka teaches a multi-stage registration in which target patient and atlas are first roughly aligned using standard rigid and deformable registration (b-spline transform/stage). Then, labels are combined using label fusion, which aims to maximize the final skeleton segmentation accuracy.) and weighted according to their respective (2.5 Label fusion strategies, Ceranka teaches determining the optimal combination of the segmentations is obtained by facilitation of the local and global weighting of each deformed atlas label (using normalized cross-correlation metric (NCC)) NCC is used to assess the similarity of the registered image with the reference. NCC provides a quantitative value representing the registration accuracy for each image pair.).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas by identifying a transformation by minimizing a cost function, aligning the pre-segmented portions of the atlases with the patient image according to a non-linear deformable transformation, and weighting them according to a similarity metric that is taught by Ceranka, to make the invention that minimizes the error between registrations of the ground truth and patient images and is able to perform registrations on more complex objects, such as images of the human body, which cannot be properly aligned using solely affine/linear transformations; thus, one of ordinary skilled in the art would be motivated to combine the references to minimize the difference between the atlas and the target image registration to provide accurate registration approximations (Ceranka, 2.4 Registration). Additionally, by weighting the registration accuracy (using a similarity measure such as NCC) for each image pair, only the best-ranked atlases can be used in the majority voting scheme to maximize the final skeleton segmentation accuracy (Ceranka, 2.5 Label fusion strategies).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Lavdas in view of Ceranka does not explicitly disclose the following:
A computer-implemented method and translating and stretching each ground truth image in its corresponding direction to align the identified regions of that ground truth image with the corresponding identified regions of the patient image and cost-functions.
Jeraj is in the same field of art of performing registration between skeleton template data and patient image data to identify and detect bone disease. Further, Jeraj teaches a computer-implemented method (Paragraph [0008], Jeraj teaches an electronic computer executes a stored program.) and translating and stretching each ground truth image in its corresponding direction to align the identified regions of that ground truth image with the corresponding identified regions of the patient image (Paragraphs [0041], [0044], Fig. 2, Jeraj teaches the bony regions, for example, represented by region (42) of the population scans (16b) are moved in translation to match the location and orientation of corresponding bony regions (46) in the skeleton template (34). The registration of skeleton template may also include a deforming registration such as expansion of each region of the reflected skeleton template to match the skeleton template. The Examiner interprets “expansion” and “stretching” to be synonymous.).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas in view of Ceranka by implementing the method on a computer and by translating and expanding regions of the template image data (or patient image data) to register the corresponding body regions in the sets of images that is taught by Jeraj, to make the invention that both automates the image analysis steps using a computer and improves the accuracy of registration by performing specific transformations (translation, expansion/stretching) between the images; thus, one of ordinary skilled in the art would be motivated to combine the references since computer-aided image analysis may alleviate workflow and automatic algorithms could reduce the reading time and improve the diagnostic accuracy of whole-body MRI (Ceranka, 1 Introduction). In addition, one of ordinary skilled in the art would be motivated to combine the references to maximize the matching between corresponding regions of the ground truth image and the patient image by performing translations and expansions to align the regions (Jeraj, Paragraph [0042]).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Lavdas in view of Ceranka in further view of Jeraj does not explicitly disclose cost-functions.
Sabuncu is in the same field of art of performing automatic segmentation of medical images. Further, Sabuncu teaches (weighted according to their respective cost-functions (VI. Discussion, Sabuncu teaches the registration cost function is a sum of squared intensity differences. In weighted label fusion, the weights are a function of sum of squared differences, i.e., anatomical similarity is measured based on squared differences of intensity values.).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas in view of Ceranka in view of Jeraj by replacing the similarity metric used to influence the atlas weighting for the final segmentation (taught by Ceranka) with a cost function such as a sum of squared intensity differences that is taught by Sabuncu, to make the invention that minimizes the error between the ground truth image and predicted image to obtain an accurate segmentation; thus, one of ordinary skilled in the art would be motivated to combine the references to improve segmentation quality since training subjects (patient images) more similar to the test subject should carry more weight during label fusion (Sabuncu, I. Introduction).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
In regards to Claim 2, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses the computer-implemented method according to claim 1, wherein the diffusion weighted MRI 3D patient image of a human or animal body contains a skeleton (2.A. Healthy volunteers and imaging protocol, Lavdas teaches 51 volunteers were scanned (24 male—mean age = 37, range = 23–67 yr and 27 female—mean age = 39, range = 23–68 yr) were scanned with whole body MRI. Axial slices were acquired during free breathing for DWI (b = 0, 150, 400, 750, and 1000 s/mm2), while breath-holds were employed for the three first stations for anatomical imaging.), the diffusion weighted MRI 3D ground truth images of other corresponding bodies contain skeletons (2.D. Multi-atlas (MA) algorithm, Lavdas teaches the multi-atlas segmentation uses a set of atlases (images with corresponding segmentations) that represent the intersubject variability of the anatomy to be segmented.), the patient image is analysed to identify different skeletal regions of the body (3 Results, Lavdas teaches automatic segmentation in both the coronal and axial planes, as shown in Fig. 2. Major organs (lungs, heart, kidneys, liver, and spleen) and bones (spine and femurs) are segmented. The Examiner interprets organs and/or bones to be contiguous regions since the claim is silent to the meaning of contiguous regions.), and each ground truth image is registered to the patient image (2.D. Multi-atlas (MA) algorithm, Lavdas teaches each atlas is registered to the new image to be segmented using a deformable image registration.) by translating and stretching each ground truth image in its corresponding direction to align the identified skeletal regions of that ground truth image with the corresponding identified skeletal regions of the patient image (Paragraphs [0041], [0044], Fig. 2, Jeraj teaches the bony regions, for example, represented by region (42) of the population scans (16b) are moved in translation to match the location and orientation of corresponding bony regions (46) in the skeleton template (34). The registration of skeleton template may also include a deforming registration such as expansion of each region of the reflected skeleton template to match the skeleton template.).
In regards to Claim 3, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses the computer-implemented method according to claim 1, wherein the pre-delineated parts of the body include the skeleton and/or one or more soft tissues (2.1 Atlas-based segmentation, 2.2 MRI and atlas skeleton label images, Ceranka teaches an atlas, which is a template intensity image for which the segmentation (the atlas labels) are available. The skeleton segmentation for the skeleton atlases was performed. The segmented bones were those most relevant to metastatic bone disease involvement, including the clavicle, vertebra from the second cervical up to the sacrum, pelvis, and femur bones. The Examiner interprets “and/or” to mean only one of either the skeleton or the soft tissues are required to meet the limitation.).
In regards to Claim 4, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses the computer-implemented method according to claim 1, wherein the analysing of the patient image to identify different regions of the body is performed using a neural network (1 Introduction, 2.C. Convolutional neural networks (CNNs) algorithm, Lavdas teaches using a convolutional neural network (CNN) to perform multiorgan segmentation in whole body MRI. Specifically, a CNN called DeepMedic is used.).
In regards to Claim 5, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses the computer-implemented method according to claim 1, wherein said direction of the body of the patient image, and the corresponding directions of the ground truth images are craniocaudal directions (2.3 Data preprocessing, Ceranka teaches aligning the whole-body intensity profile along the craniocaudal direction.), the plural slices of the patient image and the plural slices of the ground truth images being axial slices (2.A. Healthy volunteers and imaging protocol, 2.D. Multi-atlas (MA) algorithm, Lavdas teaches axial slices were acquired during free breathing for diffusion weighted imaging (DWI). The anatomies of interest were manually segmented on the T2-weighted volumes, which serve as the “ground truth” images.).
In regards to Claim 6, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses the computer-implemented method according to claim 4, wherein the identified different regions of the body include cervical, thoracic, lumbar and pelvic regions (2.2 MRI and atlas skeleton labels, Fig. 1, Ceranka teaches the segmented bones were those most relevant to metastatic bone disease involvement and included the clavicle, vertebra from the second cervical up to sacrum, pelvis, and femur bones (see Fig. 1).).
In regards to Claim 7, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses the computer-implemented method according to claim 1, further including: analysing the patient image to delineate a reference feature extending in said direction (4 Discussion and Conclusions, Ceranka teaches the pelvic region was usually aligned with high accuracy, providing a precise atlas-based segmentation. The Examiner interprets the pelvic region to be a reference feature since the claim is silent to the specific meaning of reference feature.); wherein the ground truth images are images of plural other corresponding bodies having the reference feature (2.2 MRI and atlas skeleton label images, Ceranka teaches performing skeleton segmentation to the skeleton atlases. The skeleton atlases consisted of 10 whole-body images. The segmented bones were those most relevant for metastatic bone disease involvement and included the clavicle, vertebra from the second cervical up to sacrum, pelvis, and femur bones (see Figure 1). Under BRI, the Examiner interprets any of the listed “most relevant” bones to be “reference features” since the claim is silent to the specific meaning of reference feature.); and wherein the registering of each ground truth image to the patient image also includes: translating in-plane each slice of each translated and stretched ground truth image (Paragraphs [0041], [0044], Fig. 2, Jeraj teaches the bony regions, for example, represented by region (42) of the population scans (16b) are moved in translation to match the location and orientation of corresponding bony regions (46) in the skeleton template (34).) to align the reference feature of the ground truth image in that slice with the reference feature of the patient image in a corresponding slice (Paragraphs [0041-42], Jeraj teaches aligning the bony regions of the population scans with the corresponding bony regions in the skeleton template. For example, the center of mass of the regions may be aligned. The Examiner interprets the “bony region” to be the reference feature since the claim is silent to the meaning of reference feature.).
In regards to Claim 8, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses the computer-implemented method according to claim 7, wherein the analysing of the patient image to delineate a reference feature is performed using a machine learning classifier (Classification forests (CFs) algorithm, Lavdas teaches a classification forests (CFs) algorithm. CFs are powerful multilayer classifiers that facilitate simultaneous segmentation of multiple organs.).
In regards to Claim 11, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses the computer-implemented method according to claim 1, wherein the identified transformation of each registered ground truth image that matches the pre-delineated parts of the body of that registered ground truth image to the corresponding parts of the body of the patient image is a non-linear deformable transformation (2.4 Registration, Ceranka teaches the obtained set of transformation maps were used to deform each atlas label image ln onto a target whole-body image t using the inverted transformation. An additional B-spline stage was also performed for registration.).
In regards to Claim 17, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu teaches a computer-implemented procedure (Paragraph [0026], Jeraj teaches an electronic computer which executes a stored program.) for identifying possible regions of disease in a human or animal body (Paragraphs [0053-54], [0027], Fig. 4, Jeraj teaches preparing population templates for individuals identified as having a particular medical condition whose diagnosis is of interest. Each population template may provide standard normal measured values of individuals having a particular disease diagnosis and may be tagged by that disease with a disease diagnosis tag. The population templates may then be used to provide disease diagnosis. A patient scan may then be obtained and registered to the population templates. The population template providing the best match with the registered patient scan may be used to indicate a possible diagnosis of the patient or a ranking of possible diagnoses.), the procedure including: performing the method of claim 1 (the Examiner states see rejection for Claim 1, method of combining references); and using the probability image to identify possible regions of disease (Paragraph [0026], Jeraj teaches outputting a disease diagnosis associated with at least one template image based on the match value. Each region of the composite template may provide a parameter such as probability of disease occurrence. The Examiner interprets the composite image to be a probability image since it indicates the probability of disease occurrence.).
In regards to Claim 20, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses the computer-implemented procedure according to claim 17, wherein the disease is metastatic bone disease (2.2 MRI and atlas skeleton label images, Ceranka teaches anatomical whole-body 3D-T1 15 spin-echo images (Philips Ingenia 3T) were acquired as a follow-up examination of male patients with both focal and multi-focal bone metastases.).
In regards to Claim 21, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses a computer system programmed (Paragraph [0038], Jeraj teaches an electronic computer (22) includes one or more processing units (24) communicating with memory (26) holding data and a stored program (28) for effecting portions of the invention.) to perform the method of claim 1 (the Examiner states see rejection for Claim 1, method of combining references).
In regards to Claim 27, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu teaches a method of training a computer-implemented machine learning model (2.D. Multi-atlas (MA) algorithm, Lavdas teaches training of CFs and CNNs. The CFs implementation uses all available CPUs, while the CNN implementation runs mostly on the GPU. Training was based on manual annotation of the anatomies of interest on the T2-weighted volumes.) which receives an input which is a diffusion weighted MRI 3D patient image of a human or animal body (Methods, Lavdas teaches MRI data were used as input data to the algorithms. All imaging combinations were provided to the algorithms (T2w + T1w + DWI). DWI is “diffusion-weighted imaging.”), and produces in response thereto an output which is a probability image corresponding to the diffusion weighted MRI 3D patient image for one or more parts of the body (2.C. Convolutional neural networks (CNNs) algorithm, Fig. 2, Lavdas teaches the last layer of the CNN combines all the outputs to make a prediction about the most likely class label for each voxel in an image. Automatic segmentation of the major organs (lungs, heart, kidneys, liver, and spleen) and bones (spine and femurs) is performed by each machine learning algorithm (CF, CNN, MA).), the method including: performing the method of claim 1 (the Examiner states see rejection for Claim 1, method of combining references) for plural diffusion weighted MRI 3D patient images to form a training data set (Fig. 1 caption, 2.D. Multi-atlas (MA) algorithm, Lavdas teaches manual segmentation of the anatomies of interest is performed to generate training data for the machine learning algorithms. Training of CFs and CNNs was performed using 27 images.) in which each of the diffusion weighted MRI 3D patient images is paired with a corresponding probability image for one or more parts of the body (2.1 Atlas-based segmentation, 2.5 Label fusion strategies, Ceranka teaches multiple pairwise image registrations, performed between each atlas and the target image. In addition, for each pairwise registration, a similarity metric can be calculated between the reference image and the transformed atlas image.); and minimize(ing) a cost function (2.4 Registration, Ceranka teaches an optimization problem over the parameters of the spatial transformation guided by a cost function. Further, “arg min” indicates finding the minimum of the cost function.); and training the machine learning model using the training data set (2.D. Multi-atlas (MA) algorithm, Lavdas teaches training of CFs and CNNs using the manual annotation of anatomies of interest on the T2w volumes.) to minimise a cost function (2.4 Registration, Ceranka teaches an optimization problem over the parameters of the spatial transformation guided by a cost function. Further, “arg min” indicates finding the minimum of the cost function.) that, when each diffusion weighted MRI 3D patient image from the training data set is inputted into the model (2.D. Multi-atlas (MA) algorithm, 5 Conclusions, Lavdas teaches training was based on manual annotation of the anatomies of interest on the T2-weighted volumes. CNNs were trained using T2w only images as input, CFs or MA algorithm were trained with either T2w only images or a combination of imaging inputs (T2w + T1w + DWI).), measures similarity between the output of the model and the corresponding probability image (2.D. Multi-atlas (MA) algorithm, Lavdas teaches measuring correlation coefficient as the similarity measure.).
Claims 9, 10, 22 and 25 are rejected under 35 U.S.C. 103(a) as being unpatentable over Lavdas et al. (NPL “Fully automatic, multiorgan segmentation in normal whole body magnetic resonance imaging (MRI), using classification forests (CFs), convolutional neural networks (CNNs), and a multi-atlas (MA) approach,” 2017, hereafter referred to as Lavdas) in view of Ceranka (NPL “Multi‐atlas segmentation of the skeleton from whole‐body MRI—Impact of iterative background masking,” 2019, hereafter referred to as Ceranka) in view of Jeraj (U.S. Patent Pub. No. 2018/0330495 A1, hereafter referred to as Jeraj) in view of Sabuncu et al. (NPL “A Generative Model for Image Segmentation Based on Label Fusion,” 2010, hereafter referred to as Sabuncu) in further view of Weistrand (U.S. Patent No. 9,373,173, hereafter referred to as Weistrand).
Regarding Claim 9, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses the computer-implemented method according to claim 7.
Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu does not explicitly disclose wherein the aligning of the reference feature of the ground truth image with the reference feature of the patient image in a corresponding slice is performed by matching respective centroids of the reference features.
Weistrand is in the same field of art of segmenting a patient image based on an atlas by performing registration. Further, Weistrand teaches wherein the aligning of the reference feature of the ground truth image with the reference feature of the patient image in a corresponding slice is performed by matching respective centroids of the reference features (Col. 11, lines 61-67 and Col. 12, lines 1-18, Fig. 7, Weistrand teaches determining the center points, i.e. points corresponding to the centroids of the ROIs in atlases. Then the atlases are registered with the patient image using point-based registration where the ROI center points of the registered atlas are matched with corresponding center points of ROIs in atlases. Under BRI, the Examiner interprets the region of interest “ROI” to be a reference feature since the claim is silent to the type of “reference feature.”).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu by aligning the centroids of the regions of interest between the atlases and patient images that is taught by Weistrand, to make the invention that performs registration between the atlas(es) and the patient image(s) using a point-based registration; thus, one of ordinary skilled in the art would be motivated to combine the references since performing point-based registration by matching the center points (centroids) of the atlas and patient image eliminates the need for performing image-based registration (Weistrand, Col. 12, lines 3-18). By using point-based registration, the computation time of the rigid registrations can be reduced since the point-based rigid registration is less computationally demanding (Weistrand, Col. 4, lines 31-39).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
In regards to Claim 10, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses the computer-implemented method according to claim 7.
Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu does not explicitly disclose wherein the reference feature is a spinal cord.
Weistrand is in the same field of art of segmenting a patient image based on an atlas by performing registration. Further, Weistrand teaches wherein the reference feature is a spinal cord (Col. 10, lines 58-67 – Col. 11, lines 1-4, Weistrand teaches the spinal cord may be segmented using sub-atlas registration.).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu by using the spinal cord as a region of interest that is taught by Weistrand, to make the invention that segments the spinal cord as the region of interest; thus, one of ordinary skilled in the art would be motivated to combine the references since the spinal cord is an easily discernable structure to segment (Weistrand, Col. 10, lines 58-67).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
In regards to Claim 22, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses (the Examiner states see rejection for Claim 1, method of combining references).
Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu does not explicitly disclose a computer program comprising code which, when the code is executed on a computer, causes the computer to perform the method.
Weistrand is in the same field of art of segmenting a patient image based on an atlas by performing registration. Further, Weistrand discloses a computer program comprising code which, when the code is executed on a computer, (Col. 13, lines 4-11, Weistrand teaches a computer system comprising a processor, coupled to a memory. The memory has a computer program stored thereon. The computer program comprises computer-readable instructions for performing atlas-based segmentation, where the computer-readable instructions can be transferred to, and executed by, a processor. When executed by the processor, the computer-readable instructions will perform a method.).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu by performing the method of claim 1 with a computer program stored on a computer taught by Weistrand, to make the invention that automates the method by executing the computer-readable instructions by a processor; thus, one of ordinary skilled in the art would be motivated to combine the references a large number of computations are necessary, and automating the method enables automatic segmentation for ROIs in medical images (Weistrand, Col. 13, lines 23-29 and Col. 1, lines 39-42).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
In regards to Claim 25, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses (the Examiner states see rejection for Claim 22 method of combining references).
Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu does not explicitly disclose a computer readable medium.
Weistrand is in the same field of art of segmenting a patient image based on an atlas by performing registration. Further, Weistrand discloses a computer readable medium (Col. 13, lines 13-18, Weistrand teaches the computer program can be stored on a non-transitory computer readable medium.).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu by storing the computer program of claim 22 on a non-transitory computer readable medium taught by Weistrand, to make the invention that stores the computer program on the non-transitory computer-readable medium; thus, one of ordinary skilled in the art would be motivated to combine the references since storing the computer program on a non-transitory computer-readable memory enables the computer program to be loaded onto memory or transferred to different computing systems (Weistrand, Col. 13, lines 11-22).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Claims 12, 18 and 19 are rejected under 35 U.S.C. 103(a) as being unpatentable over Lavdas et al. (NPL “Fully automatic, multiorgan segmentation in normal whole body magnetic resonance imaging (MRI), using classification forests (CFs), convolutional neural networks (CNNs), and a multi-atlas (MA) approach,” 2017, hereafter referred to as Lavdas) in view of Ceranka et al. (NPL “Multi‐atlas segmentation of the skeleton from whole‐body MRI—Impact of iterative background masking,” 2019, hereafter referred to as Ceranka) in view of Jeraj et al. (U.S. Patent Pub. No. 2018/0330495 A1, hereafter referred to as Jeraj) in view of Sabuncu et al. (NPL “A Generative Model for Image Segmentation Based on Label Fusion,” 2010, hereafter referred to as Sabuncu) in further view of Blackledge et al. (NPL “Assessment of Treatment Response by Total Tumor Volume and Global Apparent Diffusion Coefficient Using Diffusion-Weighted MRI in Patients with Metastatic Bone Disease: A Feasibility Study,” 2014, hereafter referred to as Blackledge).
In regards to Claim 18, Lavdas in view of Ceranka in view of Jeraj in view of Sabuncu discloses the computer-implemented method according to claim 1.
Lavdas in view of Ceranka in view of Jeraj in view of Sabuncu does not explicitly disclose wherein the cost function is a mean square error cost function.
Blackledge is in the same field of art of performing segmentation of whole-body diffusion-weighted MRI for monitoring metastatic bone disease in patients. Further, Blackledge discloses wherein the cost function is a mean square error cost function (Image processing and disease segmentation, Blackledge teaches images are registered according to the minimum of a mean-square-difference of image intensities.)
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu by substituting the cost function with a mean square error cost function, to make the invention that accurately registers the patient image to the ground truth image by minimizing a difference between the two images; thus, one of ordinary skilled in the art would be motivated to combine the references to provide improved transformations and registration between the patient and ground truth images to obtain accurate final segmentations of the patient images (Ceranka, 2.4 Registration).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
In regards to Claim 18, Lavdas in view of Ceranka in view of Jeraj in view of Sabuncu discloses the computer-implemented procedure according to claim 17.
Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu does not explicitly disclose wherein the probability image is used by combining the probability image with other information derived from the diffusion weighted MRI 3D patient image of the human or animal body.
Blackledge is in the same field of art of performing segmentation of whole-body diffusion-weighted MRI for monitoring metastatic bone disease in patients. Further, Blackledge discloses wherein the probability image is used by combining the probability image with other information derived from the diffusion weighted MRI 3D patient image of the human or animal body (Introduction, Image processing and disease segmentation, Blackledge teaches the segmentation process generates multiple regions of interest (ROIs) that are summed to produce volumes of interest (VOIs) of disease across the body. The generated VOIs are used to calculate an estimate for the total body tumor burden. The VOIs are transferred onto corresponding ADC maps to derive global median Apparent Diffusion Coefficient (gADC) values. Changes in the gADC values may be used to assess the treatment response of metastatic bone disease. The distribution of ADC values across all lesions is visualized by histogram analysis. The Examiner interprets the segmentation images to be “probability images” since they provide estimates of total tumor diffusion.).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu by transferring the volumes of interest onto corresponding ADC maps to derive global median Apparent Diffusion Coefficient values that is taught by Blackledge, to make the invention that assesses treatment response of patients with metastatic bone disease; thus, one of ordinary skilled in the art would be motivated to combine the references to assess tumor response to treatment in patients with bone metastases and provide quantitative metrics from a single radiologic investigation (Blackledge, Introduction).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
In regards to Claim 19, Lavdas in view of Ceranka in view of Jeraj in view of Sabuncu in further view of Blackledge discloses the computer-implemented procedure according to claim 18, wherein the other information is apparent diffusion coefficient information or high b-value image signal information (2.A. Health volunteers and imaging protocol, Lavdas teaches acquiring whole body MRI images. Axial slices were acquired during free breathing for diffusion weighted imaging (DWI) (b = 0, 150, 400, 750, and 1000 s/mm2. Apparent diffusion coefficient (ADC) maps were generated online using a monoexponential fit to the equation: S = S0·e−b·ADC.).
Claims 13-16 and 23 are rejected under 35 U.S.C. 103(a) as being unpatentable over Lavdas et al. (NPL “Fully automatic, multiorgan segmentation in normal whole body magnetic resonance imaging (MRI), using classification forests (CFs), convolutional neural networks (CNNs), and a multi-atlas (MA) approach,” 2017, hereafter referred to as Lavdas) in view of Jeraj (U.S. Patent Pub. No. 2018/0330495 A1, hereafter referred to as Jeraj).
In regards to Claim 13, Lavdas discloses a of delineating one or more parts of a body within a diffusion weighted MRI 3D patient image of a human or animal body (1 Introduction, Lavdas teaches algorithms for automatic, multiorgan segmentation in whole body MRI using machine learning approaches. Diffusion-weighted images (DWI) are captured.), the method including: providing a processor adapted to provide a machine learning model (2.D. Multi-atlas (MA) algorithm, Lavdas teaches the CNN implementation runs on the GPU.) which receives an input which is a diffusion weighted MRI 3D patient image of a human or animal body (Methods, Lavdas teaches MRI data were used as input data to the algorithms. All imaging combinations were provided to the algorithms (T2w + T1w + DWI). DWI is “diffusion-weighted imaging.”), and produces in response thereto an output which is a probability image corresponding to the diffusion weighted MRI 3D patient image for one or more parts of the body (2.C. Convolutional neural networks (CNNs) algorithm, Fig. 2, Lavdas teaches the last layer of the CNN combines all the outputs to make a prediction about the most likely class label for each voxel in an image. Automatic segmentation of the major organs (lungs, heart, kidneys, liver, and spleen) and bones (spine and femurs) is performed by each machine learning algorithm (CF, CNN, MA).); and inputting the diffusion weighted MRI 3D patient image of the human or animal body into the machine learning model (3 Results, Lavdas teaches using T2w volumes and all imaging combinations (T2w + T1w + DWI) as input to the CFs and CNNs. DWI is diffusion weighted image. CF is classification forests algorithm and CNN is convolutional neural network. Both are machine learning algorithms.) to produce the corresponding probability image (3 Results, Lavdas teaches automatic segmentations of major organs (lungs, heart, kidneys, liver, and spleen) and bones (spine and femurs) from the three machine learning algorithms. The Examiner interprets the segmentations to be “probability images” since in the case of the CNN, the last layer combines all the outputs to make a prediction about the most likely class label for each voxel in an image.”).
Lavdas does not explicitly disclose a computer-implemented method.
Jeraj is in the same field of art of performing registration between template “ground truth” image data and patient image data to identify and detect disease, specifically bone disease. Further, Lavdas teaches a computer-implemented method (Paragraph [0008], Jeraj teaches an electronic computer executes a stored program.).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas by implementing the automatic segmentation method on a computer via a computer program that is taught by Jeraj, to make the invention that automates the registration and segmentation methods; thus, one of ordinary skilled in the art would be motivated to combine the references to automate the registration step and modification steps and further identify and display differences to detect the presence of disease in patient images (Jeraj, Paragraph [0008]). Further, automating the method via a computer improves scalability and efficiency of the method.
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
In regards to Claim 14, Lavdas in view of Jeraj discloses the computer-implemented method according to claim 13, wherein the machine learning model receives an input which is a diffusion weighted MRI 3D ground truth image containing a skeleton (2.D. Multi-atlas (MA) algorithm, Lavdas teaches multimodal MRI data were used as input to CFs and CNNs (e.g., T2w + T1w + DWI data, where T1w refers to T1w in- and opposed-phase images from the DIXON acquisitions, and DWI refers to b = 1000 s/mm2 images and ADC maps).).
In regards to Claim 15, Lavdas in view of Jeraj discloses the computer-implemented method according to claim 13, wherein the one or more parts of the body include the skeleton and/or one or more soft tissues (Abstract, Figs. 1 & 2, Lavdas teaches automatic segmentation of major organs and bones in whole body MRI. See Figs. 1 & 2 which show the imaged skeleton with soft tissues (organs, muscles, etc.).).
In regards to Claim 16, Lavdas in view of Jeraj discloses the computer-implemented method according to claim 13, wherein the machine learning model is a neural network (1 Introduction, 2.C. Convolutional neural networks (CNNs) algorithm, Lavdas teaches using a convolutional neural network (CNN) to perform multiorgan segmentation in whole body MRI. Specifically, a CNN called DeepMedic is used.).
In regards to Claim 23, Lavdas discloses (2.D. Multi-atlas (MA) algorithm, Lavdas teaches the CNN implementation runs on the GPU.) which receives an input which is a diffusion weighted MRI 3D patient image of a human or animal body (Methods, Lavdas teaches MRI data were used as input data to the algorithms. All imaging combinations were provided to the algorithms (T2w + T1w + DWI). DWI is “diffusion-weighted imaging.”), and produces in response thereto an output which is a probability image corresponding to the diffusion weighted MRI 3D patient image for one or more parts of the body (2.C. Convolutional neural networks (CNNs) algorithm, Fig. 2, Lavdas teaches the last layer of the CNN combines all the outputs to make a prediction about the most likely class label for each voxel in an image. Automatic segmentation of the major organs (lungs, heart, kidneys, liver, and spleen) and bones (spine and femurs) is performed by each machine learning algorithm (CF, CNN, MA).).
Lavdas does not explicitly disclose a computer system.
Jeraj is in the same field of art of performing registration between “template image data” (ground truth) and patient image data to identify and detect disease, specifically bone disease. Further, Jeraj discloses a computer system (Paragraph [0038], Jeraj teaches an electronic computer (22).).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas by implementing the machine learning algorithms for automatic segmentation on a computer system that is taught by Jeraj, to make the invention that automatically segments the patient images using a stored program (machine learning algorithm) on a computer system; thus, one of ordinary skilled in the art would be motivated to combine the references since it enables automated automatic detection and segmentation of lesions/abnormalities in whole body MRI scans for cancer patients (Lavdas, 5 Conclusions).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Claim 24 is rejected under 35 U.S.C. 103(a) as being unpatentable over Lavdas et al. (NPL “Fully automatic, multiorgan segmentation in normal whole body magnetic resonance imaging (MRI), using classification forests (CFs), convolutional neural networks (CNNs), and a multi-atlas (MA) approach,” 2017, hereafter referred to as Lavdas) in view of Weistrand (U.S. Patent No. 9,373,173, hereafter referred to as Wesitrand).
Regarding Claim 24, Lavdas discloses (Methods, Lavdas teaches MRI data were used as input data to the algorithms. All imaging combinations were provided to the algorithms (T2w + T1w + DWI). DWI is “diffusion-weighted imaging.”), and produces in response thereto an output which is a probability image corresponding to the diffusion weighted MRI 3D patient image for one or more parts of the body (2.C. Convolutional neural networks (CNNs) algorithm, Fig. 2, Lavdas teaches the last layer of the CNN combines all the outputs to make a prediction about the most likely class label for each voxel in an image. Automatic segmentation of the major organs (lungs, heart, kidneys, liver, and spleen) and bones (spine and femurs) is performed by each machine learning algorithm (CF, CNN, MA).).
Lavdas does not explicitly disclose a computer program comprising code which, when the code is executed on a computer, causes the computer to execute
Weistrand is in the same field of art of segmenting a patient image based on an atlas by performing registration. Further, Wesitrand teaches a computer program comprising code which, when the code is executed on a computer, causes the computer to execute (Col. 13, lines 4-11, Weistrand teaches a computer system comprising a processor, coupled to a memory. The memory has a computer program stored thereon. The computer program comprises computer-readable instructions for performing atlas-based segmentation, where the computer-readable instructions can be transferred to, and executed by, a processor. When executed by the processor, the computer-readable instructions will perform a method.).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas by executing the machine learning algorithm on a computer program comprising computer-readable instructions which, when executed, will cause the computer to perform the method that is taught by Wesitrand, to make the invention that performs the machine learning algorithm on a computer; thus, one of ordinary skilled in the art would be motivated to combine the references since these algorithms could facilitate the process of reading whole body scans in cancer patients by reducing the reading time, and possibly also improving the diagnostic accuracy of whole body MRI (Lavdas, 4 Discussion).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Claims 26 and 28 are rejected under 35 U.S.C. 103(a) as being unpatentable over Lavdas et al. (NPL “Fully automatic, multiorgan segmentation in normal whole body magnetic resonance imaging (MRI), using classification forests (CFs), convolutional neural networks (CNNs), and a multi-atlas (MA) approach,” 2017, hereafter referred to as Lavdas) in view of Ceranka et al. (NPL “Multi‐atlas segmentation of the skeleton from whole‐body MRI—Impact of iterative background masking,” 2019, hereafter referred to as Ceranka) in view of Jeraj et al. (U.S. Patent Pub. No. 2018/0330495 A1, hereafter referred to as Jeraj) in view of Sabuncu et al. (NPL “A Generative Model for Image Segmentation Based on Label Fusion,” 2010, hereafter referred to as Sabuncu) in further view of Yankeelov (U.S. Patent Pub. No. 2025/0272828 A1, hereafter referred to as Yankeelov).
Regarding Claim 26, Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu discloses (the Examiner states see rejection for Claim 21, method of combining references.)
Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu does not explicitly disclose an imaging system for performing diffusion-weighted MRI, the system including: a magnetic resonance imaging scanner for acquiring a diffusion-weighted MRI 3D patient image of a human or animal body; the computer system being configured to communicate with the scanner such that the computer system is provided with the acquired image.
Yankeelov is in the same field of art of analyzing regions of interest in MRI data such as diffusion-weighted MRI scans of the ROI. Further, Yankeelov teaches an imaging system for performing diffusion-weighted MRI, (Paragraph [0039], Yankeelov teaches an imaging system that may include a sequential imaging data source that acquires data based on sequential images (such as diffusion-weighted MRI (DW-MRI) data.) the system including: a magnetic resonance imaging scanner for acquiring a diffusion-weighted MRI 3D patient image of a human or animal body (Paragraph [0039], Yankeelov teaches the sequential imaging source may be part of one MRI system, or may be connected to one or more MRI scanners. The sequential imaging data source acquires diffusion-weighted MRI data.); the computer system being configured to communicate with the scanner such that the computer system is provided with the acquired image (Paragraph [0034], Yankeelov teaches a system that includes a computing device (or multiple computing devices) capable of communicating with an imaging system, such as an MRI scanner. The computing device may include a controller that exchanges control signals with the imaging system, allowing the computing device to control the capture of images via the imaging system, retrieve imaging data, etc.).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Lavdas in view of Ceranka in view of Jeraj in further view of Sabuncu by enabling communication between a diffusion-weighted MRI scanner/imaging system and the computer system that is taught by Yankeelov, to make the invention that directly acquires and provides images to the computer system/machine learning algorithms for analysis; thus, one of ordinary skilled in the art would be motivated to combine the references to retrieve images and perform direct analysis of the imaging data as well as output analysis results (Yankeelov, Paragraph [0034]).
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
In regards to Claim 28, Lavdas in view of Ceranka in view of Jeraj in view of Sabuncu in further view of Yankeelov discloses the computer-implemented method according to claim 26 (the Examiner states see rejection for Claim 26, method of combining references) wherein the machine learning model is a neural network (1 Introduction, 2.C. Convolutional neural networks (CNNs) algorithm, Lavdas teaches using a convolutional neural network (CNN) to perform multiorgan segmentation in whole body MRI. Specifically, a CNN called DeepMedic is used.).
Pertinent Prior Art
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Wang et al. (NPL “Multi-Atlas Segmentation with Joint Label Fusion,” 2013) discloses a weighted voting strategy which minimizes the total expectation of labeling error and in pairwise dependence between atlases is explicitly modeled as the joint probability of two atlases making a segmentation error at a voxel.
Sjöberg et al. (NPL “Multi-atlas based segmentation using probabilistic label fusion with adaptive weighting of image similarity measures,” 2013) discloses a multi-atlas label fusion strategy based on probabilistic weighting of distance maps. Relationships between image similarities and segmentation similarities are estimated in a learning phase and used to derive fusion weights that are proportional to the probability for each atlas to improve the segmentation result. Mean square difference is one of the similarity measures used for scoring registrations.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SYDNEY L BLACKSTEN whose telephone number is (571)272-7120. The examiner can normally be reached 8:30am-4:30pm.
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, Oneal Mistry can be reached at 313-446-4912. 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.
/SYDNEY L BLACKSTEN/Examiner, Art Unit 2674
/ONEAL R MISTRY/Supervisory Patent Examiner, Art Unit 2674
Prosecution Insights