DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
Applicant’s Amendments filed on April 29, 2026, have been entered and made of record.
Currently pending Claim(s): 1-16
Independent Claim(s): 1, 14
New Claim(s): 16
Response to Arguments
This office action is responsive to the Applicant’s Arguments/Remarks made in an Amendment received April 29, 2026.
In view of the Amendments, Applicant has included a new claim, Claim 16. Originally, (in the claim set dated February 26, 2026), Claim 1 was rejected using Chen (CN 112842371), Ye (CN 102693540), and Li (Li, Zhuowei, et al., “Segmentation to Label: Automatic Coronary Artery Labeling from Mask Parcellation”, 2020).
Applicant’s arguments, see pg. 9 paragraphs 1-3, with respect to the rejection(s) of Claim(s) 1 and 14 under U.S.C. 103 have been fully considered and are persuasive. The Examiner agrees that Li fails to teach the limitation of “outputting a labelling result of the portal vein branch label for each image unit element”. Therefore, the rejection has been withdrawn. However, upon further consideration, a new ground(s) of rejection is made in view of Tian et al. (Tian, J., et al., “Automatic Couinaud Segmentation from CT Volumes on Liver Using GLC-UNet”, Machine Learning in Medical Imaging. MLMI 2019. Lecture Notes in Computer Science, Vol. 11861. Springer, Cham) in view of Chen, since Tian in view of Chen teaches a neural network which is trained to output hepatic segments.
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.
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 1, 6-8, 10, and 14-15 are rejected under 35 U.S.C. 103 as being unpatentable over Tian et al. (Tian, J., et al., “Automatic Couinaud Segmentation from CT Volumes on Liver Using GLC-UNet”, Machine Learning in Medical Imaging. MLMI 2019. Lecture Notes in Computer Science, Vol. 11861. Springer, Cham), hereinafter Tian, in view of Chen et al. (CN 112842371), hereinafter Chen.
As to Claim 1, Tian discloses a medical image processing apparatus comprising:
a processor; and a storage device that stores a program to be executed by the processor (see Section 3, pg. 279, “GLC-UNet is implemented and evaluated on an NVIDIA Tesla PlOO”, where the Tesla PIOO is a computing device having a processor that executes a stored program),
wherein the program includes a trained model generated by performing machine learning (see Abstract, pg. 274, "we propose a novel global and local contexts UNet (GLC-UNet) for Couinaud segmentation”, where UNet is a machine learning model)
using training data that includes first input data including a first image regarding a liver, (see Section 3, pg. 278, where for training datasets, “For liver, we annotate 443 CT volumes in total”),
and hepatic labeling data in which a hepatic label is attached to a region in the liver in the first image for each hepatic region corresponding to a hepatic segment (see pg. 275, Section 1, “We make annotation on both liver and Couinaud segments of this dataset”, and see pg. 275, where the model is trained on “our annotated 43,205 CT slices on liver and Couinaud segmentation), and
the trained model is a model obtained by updating parameters of a learning model (see Section 2, pg. 278, “Gti is dynamic maps that represent the relevant part of features at level t, where αti is a scalar weighting”, and it is well known that model weights are parameters of a model that are updated through training, and see Section 3, pg. 279, “We randomly select 50 cases for testing. Training is conducted on the remaining 143 ones”)
trained to output a labeling result of the hepatic label for each image unit element of a first image region of the first image by accepting an input of the first input data (see pg. 281, Figure 4, where a hepatic label has been produced for each pixel corresponding to the liver region),
and the processor executes a command of the program to accept second input data which is a same type of input data as the first input data and includes a second image regarding the liver (see Section 2, pg. 275, "Given a CT volume, a GLC-UNet is first utilized to obtain segmentation of the liver", where the given CT volume not used for training is interpreted as the second input data),
assign the hepatic label to each image unit element of a second image region of the second image using the trained model (see Section 2, pg. 276, : "GLC-UNet is used to obtain Couinaud segmentation of liver", see pg. 281, Figure 4, where a hepatic label has been produced for each pixel corresponding to the liver region),
and divide a liver region included in the second input data into a plurality of the hepatic segments on the basis of the hepatic label assigned to each image unit element of the second image region (see pg. 281, Figure 4, wherein a liver region has been divided into a plurality of hepatic segments).
Tian teaches a trained convolutional model that assigns a per-pixel liver label but does not specifically teach that the label is a portal vein branch label attached to a portal vein region for each portal vein branch corresponding to a hepatic segment, or that the liver is divided into the hepatic segments on the basis of the portal vein branch label assigned to each image unit element. Instead, the model is trained on datasets labelled with Couinaud segments. The examiner notes that a Couinaud segment of the liver is a segment of a liver corresponding to a specific portal vein. Tian further teaches that the model uses portal vein data to determine the correct per-pixel label (see Section 2, pg. 275-276, “Next, the RPV and the LPV are detected. The slices contain these two veins are baseline for global context for Couinaud segmentation”, where LPV stands for ‘left portal vein’, and RPV stands for ‘right portal vein’), but does not explicitly teach a ‘portal vein branch label’ attached to a ‘portal vein region’.
However, in an analogous art, Chen teaches a neural network (see abstract) which is trained using training data that includes first input data including a first image regarding a liver
and portal vein branch labeling data in which a portal vein branch label is attached to a portal vein region in the liver in the first image for each portal vein branch corresponding to a hepatic segment (see paragraph [0117], “And a second training sample labeled with hepatic vein branches and/or portal vein branches can be used to train the second neural network for multiple rounds until the trained second neural network meets the preset conditions”),
the trained model is a model obtained by updating parameters of a learning model (see paragraph [0117], "And a second training sample labeled with hepatic vein branches and/or portal vein branches can be used to train the second neural network for multiple rounds until the trained second neural network meets the preset conditions. The preset conditions can be that the accuracy is greater than a set accuracy threshold; or, the preset conditions can be that the loss value is less than a set loss threshold, etc. The accuracy threshold or loss threshold can be set as needed", where the loss is a parameter generated by the learning model)
and assigning the branch label to each image unit element and dividing the liver on that basis (see paragraph [0271], “Furthermore, based on the second category corresponding to each second pixel to be classified, the image of the first target region can be hierarchically divided to obtain the target image, where the image of the first target region is the contour image corresponding to the portal vein segmentation region in the first detection image. For example, different colors can be used to label different second categories, and different categories of pixels on the portal vein can be distinguished by different colors, thus realizing the hierarchical division of the portal vein detection image”),
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply the learned per-pixel labeling technique of Tian to the portal vein branch labeling and liver division of Chen, so that a single trained model directly outputs the portal vein branch label for each image unit element and the liver is divided into the hepatic segments on the basis of that label. Chen teaches training a segmentation network on samples marked with the portal vein vessel, and Tian supplies the machine learning model (GLC-UNet) that performs the analogous per-pixel liver labeling. Although Chen employs trained networks to generate its liver-segment and vein detection images, Chen determines the branch category and performs the division by a geometric nearest-centerline algorithm rather than by a trained model (see Chen, paragraphs [0150-0153]); the combination of Chen and Tain provides the learned, direct per-pixel branch-label output that Chen lacks. The motivation would have been to improve accuracy and robustness against the anatomic variation of the portal vein and to reduce reliance on hand-tuned geometric thresholds (see Tian, pg. 275, Section 1, “Meanwhile, the liver is a common site of primary or secondary tumor development. Their heterogeneous and diffusive shape will make the situation even worse. Consequently, it is difficult to deduce segmental anatomy of the liver... To address the aforementioned problem, in this paper, we propose a frame work, GLC-UNet, where 3Dcontexts are effectively probed using attention mechanism for accurate Couinaud segmentation.” Thus, it would have been obvious to combine the teachings of Chen and Tian in order to obtain the invention as claimed in Claim 1.
As to Claim 6, Tian in view of Chen teaches wherein the first image region is an entire region of the first image, and the second image region is an entire region of the second image (see Chen, paragraph [0116], "For example, a three-dimensional liver image of the target object can be acquired, and the three-dimensional liver image can be input into a trained first neural network to obtain a three-dimensional liver segment detection image; and the three-dimensional liver image can be input into a trained second neural network to obtain a three-dimensional vein detection image", where both images input into the neural networks contain the entire liver area). Thus, it would have been obvious to combine the image region processing taught by Chen with the teachings of Tian. The motivation for doing so would be accuracy and robustness against the anatomic variation of the portal vein and to reduce reliance on hand-tuned geometric thresholds (see Tian, pg. 275, Section 1).
As to Claim 7, Tian in view of Chen teaches wherein the portal vein branch label is a label for classifying the portal vein branch into eight classes corresponding to eight types of the hepatic segments from S1 to S8 (see Chen, paragraph [0098], "According to this method, the nearest principles of pressing that hepatic tissue is all are divided in the corresponding vessel branch”, and see paragraph [0044], "Through experimental verification, the above-mentioned method of segmentation of liver segments conforms to Couinaud's segmentation of liver segments", where Couinaud's segmentation is a term of the art used to describe the eight types of hepatic segments. Thus, it would have been obvious to combine the portal vein labelling taught by Chen with the per-pixel labelling taught by Tian. The motivation for doing so would be accuracy and robustness against the anatomic variation of the portal vein and to reduce reliance on hand-tuned geometric thresholds (see Tian, pg. 275, Section 1).
As to Claim 8, Tian in view of Chen teaches the trained model is configured using a convolution neural network (see Tian, pg. 274, Abstract, “In this paper, we propose a novel global and local contexts UNet (GLC UNet) for Couinaud segmentation”, where UNet is a well-known convolutional neural network).
As to Claim 10, Tian in view of Chen teaches wherein each of the first image and the second image is a three-dimensional image (see Tian pg. 275, Section 2, “Given a CT volume, a GLC-UNet is first utilized”, where a CT volume is a 3D image, and see pg. 278, Section 3, “For liver, we annotate 443 CT volumes in total”, thus showing that the first training image, and the second image input into the model are three dimensional).
As to Claim 14, Claim 14 claims a hepatic segment division method (see Chen, paragraph [0004], "this disclosure provides at least one image processing method" the method comprising that performs the same steps executed by the medical image processing apparatus claimed in Claim 1. Therefore, the rejection and rationale are analogous to that made in Claim 1.
As to Claim 15, Claim 15 claims a non-transitory, computer-readable tangible recording medium (see Chen, paragraph [0095], "this disclosure provides a computer readable storage medium storing a computer program, which, when run by a processor, performs the steps of the image processing method as described in the first aspect or any embodiment above"), which records thereon a program that causes a computer to operate as a medical image processing apparatus, which is the same as the medical image processing apparatus as claimed in Claim 1. Therefore, the rejection and rationale are analogous to that made in Claim 1.
Claims 2-5, 11, and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Tian et al. (Tian, J., et al., “Automatic Couinaud Segmentation from CT Volumes on Liver Using GLC-UNet”, Machine Learning in Medical Imaging. MLMI 2019. Lecture Notes in Computer Science, Vol. 11861. Springer, Cham), hereinafter Tian, in view of Chen (CN 112842371), hereinafter Chen, and further in view of Li et al. (CN 111161241), hereinafter Li.
As to Claim 2, Tian in view of Chen teaches that the first input data includes at least one of a computed tomography (CT) image in which a region including the liver is imaged (see Tian, Section 1, pg. 275, " we apply this system to a publicly available liver CT dataset").
Tian in view of Chen also teaches obtaining a portal vein detection image (see Tian, “see Section 2, pg. 275-276, “Next, the RPV and the LPV are detected. The slices contain these two veins are baseline for global context for Couinaud segmentation”, and see Chen, paragraph [0114], "The first detection image can include: the hepatic vein detection image representing the hepatic vein contour information, and/or, the portal vein detection image representing the portal vein contour information"), but fails to teach a mask image of the portal veins is created.
However, Li teaches a hepatic vessel segmentation method (see abstract), that can be used to obtain a portal vein mask image sequence (see paragraph [0080], "Based on the liver region mask image sequence and the liver vascular mask image sequence, determine the portal vein image sequence and the hepatic vein image sequence"). Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed inventio to combine the portal mask image sequence taught by Li with the teachings of Chen and Tian. The motivation for doing so would be to would be to increase the segmentation accuracy and remove unnecessary body tissue. Li teaches in paragraph [0079], "In this embodiment of the application, after obtaining the medical image sequence, since the medical image sequence also contains image content of other body tissues besides liver tissue, in order to improve the segmentation accuracy of the liver to be segmented, it is necessary to extract and identify the image of the liver tissue in the obtained medical image sequence to obtain the image of the liver region, that is, the liver region mask image sequence. In addition, in order to further improve the segmentation accuracy of the liver to be segmented, the blood vessels related to the liver in the medical image sequence are also extracted and identified to obtain the liver blood vessel mask image sequence"). Thus, it would have been obvious to combine the teachings of Li with the teachings of Tian and Chen in order to obtain the invention as claimed in Claim 2.
As to Claim 3, Tian in view of Chen teaches a CT image but fails to explicitly teach a portal vein mask image. However, Li teaches a hepatic vessel segmentation method that can be used to obtain a portal vein mask image sequence (see paragraph [0080], "Step 103: Based on the liver region mask image sequence and the liver vascular mask image sequence, determine the portal vein image sequence and the hepatic vein image sequence"). Thus, it would have been obvious to combine the teachings of Li with the teachings of Tian and Chen in order to improve segmentation accuracy as earlier discussed in Claim 2.
As to Claim 4, Chen in view of fails to teach that the first input data further includes liver mask at least one of a liver mask image in which a liver region is specified, a vein mask image in which a vein region is specified, or an inferior vena cava mask image in which an inferior vena cava region is specified. However, Li teaches a liver mask image (see paragraph [0094], "When obtaining the liver region mask image sequence and the liver blood vessel mask image sequence, the pixel values belonging to the liver in each image of the medical image sequence are set to 1, and the corresponding pixels other than the liver are set as background. The corresponding background pixels can be set to 0, thus obtaining the liver region mask image sequence"). Thus, it would have been obvious to combine the teachings of Li with the teachings of Tian and Chen in order to improve segmentation accuracy as earlier discussed in Claim 2.
As to Claim 5, Tian in view of Chen fails to teach the first input data includes the portal vein mask image, the liver mask image, and the vein mask image. However, Li teaches a portal vein mask image sequence (see paragraph [0080], "portal vein image sequence" ),
a liver mask image (see paragraph [0080], "liver region mask image sequence"),
and a vein mask image sequence (see paragraph [0080], "liver vascular mask image sequence").
Thus, it would have been obvious to combine the multiple masks taught by Li with the teachings of Chen in order to obtain the invention as claimed in Claim 5. The motivation for doing so would be to increase the segmentation accuracy, as earlier discussed in Claim 2.
As to Claim 11, Tian in view of Chen fails to explicitly the processor performs labeling of a hepatic segment label indicating the hepatic segment on the basis of the portal vein branch label assigned to each image unit element of the second image region. Tian teaches a hepatic label can be assigned to each pixel (see Tian, pg. 281), and Chen teaches assigning a portal branch label to teach image unit element (see Chen, paragraph [0018], “and the portal vein detection image can be hierarchically divided to obtain the target image corresponding to the portal vein of the target object”), but both fail to teach labelling a hepatic segment based on a portal vein branch label.
However Li teaches that another label corresponding to hepatic segments can be generated on the basis of the location of a portal vein (see paragraph [0183] and paragraph [0187], "Based on the upper and lower segment image sequences of the right portal vein, the right anterior region image sequence and the right posterior region image sequence are segmented sequentially to obtain the upper segment image sequence of the right anterior region, the lower segment image sequence of the right anterior region, the upper segment image sequence of the right posterior region, and the lower segment image sequence of the right posterior region in the target liver segment. Mark the upper segment image sequence of the left lateral region as segment 2, the lower segment image sequence of the left lateral region as segment 3, the left medial region image sequence as segment 4, the upper segment image sequence of the right anterior region as segment 5, the upper segment image sequence of the right posterior region as segment 6, the lower segment image sequence of the right posterior region as segment 7, and the lower segment image sequence of the right anterior region as segment 8", where segments 2-8 labels for the hepatic segments").
Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the hepatic segment labels taught by Li with the teachings of Tian and Chen. The motivation for doing so would be to create accurate hepatic segment labels. Li teaches in paragraph [0005], "In order to solve the above-mentioned technical problems, the embodiments of this application expect to provide a liver image recognition method, electronic device and storage medium, which solves the problem that the liver part cannot be segmented automatically in the prior art, and improves the accuracy of the electronic device in segmenting the liver part." Thus, it would have been obvious to one of ordinary skill in the art to combine the teaching of Li with the teachings of Chen and Tian in order to obtain the invention as claimed in Claim 11.
As to Claim 13, Tian in view of Chen teaches a hepatic segment division image in which a region is divided into the hepatic segments (see Tian, pg. 281, Figure 4, “An example of generating Couinaud segmentation (more results are available in Sect. 4 in the Supplementary). Segments I to VIII are marked with different colors). However, Tian in view of Chen fails to explicitly teach that the hepatic segments are generated by converting the portal vein branch label assigned to each image unit element of the second image region into the hepatic segment label. However Li teaches that another label corresponding to hepatic segments can be generated on the basis of the location (see paragraph [0183] and paragraph [0187], "Based on the upper and lower segment image sequences of the right portal vein, the right anterior region image sequence and the right posterior region image sequence are segmented sequentially to obtain the upper segment image sequence of the right anterior region, the lower segment image sequence of the right anterior region, the upper segment image sequence of the right posterior region, and the lower segment image sequence of the right posterior region in the target liver segment. Mark the upper segment image sequence of the left lateral region as segment 2, the lower segment image sequence of the left lateral region as segment 3, the left medial region image sequence as segment 4, the upper segment image sequence of the right anterior region as segment 5, the upper segment image sequence of the right posterior region as segment 6, the lower segment image sequence of the right posterior region as segment 7, and the lower segment image sequence of the right anterior region as segment 8", where segments 2-8 labels for the hepatic segments"). Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to convert the portal vein labels taught by Chen to the hepatic segment labels taught by Li. The motivation for doing so would be to increase the accuracy of hepatic segmentation. Li teaches in paragraph [0005], "In order to solve the above-mentioned technical problems, the embodiments of this application expect to provide a liver image recognition method, electronic device and storage medium, which solves the problem that the liver part cannot be segmented automatically in the prior art, and improves the accuracy of the electronic device in segmenting the liver part." Thus, it would have been obvious to combine the teachings of Li with the teachings of Tian and Chen in order to obtain the invention as claimed in Claim 13.
Claims 9 is rejected under 35 U.S.C. 103 as being unpatentable over Tian et al. (Tian, J., et al., “Automatic Couinaud Segmentation from CT Volumes on Liver Using GLC-UNet”, Machine Learning in Medical Imaging. MLMI 2019. Lecture Notes in Computer Science, Vol. 11861. Springer, Cham), hereinafter Tian, in view of Chen (CN 112842371), hereinafter Chen, and further in view of Ke-Feng et al. (CN 112733708), hereinafter Ke-Feng.
As to Claim 9, Tian in view of Chen teaches calculating a loss only for a portal vein region in which the portal vein branch label is attached, in the portal vein branch labeling data corresponding to the first input data (see [0117] of Chen, "And a second training sample labeled with hepatic vein branches and/or portal vein branches can be used to train the second neural network for multiple rounds until the trained second neural network meets the preset conditions. The preset conditions can be that the accuracy is greater than a set accuracy threshold; or, the preset conditions can be that the loss value is less than a set loss threshold, etc. The accuracy threshold or loss threshold can be set as needed. ")
Tian in view of Chen fails to teach a score map indicating a probability of the portal vein branch label output from the learning model. However, Ke-Feng teaches that a convolutional neural network that can calculate the probability that a feature is a portal vein (see paragraph [0069], "the data is processed by a convolutional neural network with added prior knowledge, extracting features through convolution operations, and using the softmax function to calculate the probability of whether the extracted feature is the portal vein during prediction", and multiple features can contain scores, thus creating a 'score map' for an image containing multiple features). Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the probability scores taught by Ke-Feng with the teachings of Tian and Chen. The motivation for doing so would be to use the probabilities output to train the model. Ke-Feng teaches in paragraphs [0088-0089], “After CT image feature extraction, logistic regression is performed using a softmax classifier to detect the target and coordinate regression to locate the target, thereby identifying the portal vein. Focal loss is selected as the loss function of the classifier, allowing the neural network to train normally even under imbalanced positive and negative samples. By using a semi-supervised learning model for training, the model can be trained using a small number of labeled samples and a large number of unlabeled samples, thus greatly reducing the cost of manual annotation”. Thus, it would have been obvious to combine the probability scores taught by Ke-Feng with the teachings of Tian and Chen in order to obtain the invention as claimed in Claim 9.
Claims 16 is rejected under 35 U.S.C. 103 as being unpatentable over Tian et al. (Tian, J., et al., “Automatic Couinaud Segmentation from CT Volumes on Liver Using GLC-UNet”, Machine Learning in Medical Imaging. MLMI 2019. Lecture Notes in Computer Science, Vol. 11861. Springer, Cham), hereinafter Tian, in view of Chen (CN 112842371), hereinafter Chen, further in view of Ke-Feng et al. (CN 112733708), hereinafter Ke-Feng, and further in view of Birenbaum et al. (US Pub No 20240303927), hereinafter Birenbaum.
As to Claim 16, Tian in view of Chen Ke-Feng fails to teach wherein the trained model outputs a prediction map, the prediction map being a score map of portal vein branch labels in which a score indicating a probability of the portal vein branch label is attached to each image unit element for the entire first image region.
However, in an analogous art of medical image analysis, Birenbaum teaches outputting a prediction map, the prediction map being a score map of vein branch labels in which a score indicating a probability of the label is attached to each image unit element for the entire first image region (see paragraph [0068], “ The classification layer 340 generates a classification map 345, which includes, for each voxel, a probability that the voxel is an artery or vein”.
Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the prediction map taught by Birenbaum with the teachings of with the probability scores output by Ke-Feng. The motivation for doing so would be to use the prediction map to correct the final model’s output. Birenbaum teaches in paragraph [0069], “Or, conversely, the classification layer 340 may indicate that two points belong to different blood vessels, while the topological layer 320 may indicate that the two points belong to the same blood vessel. The consistency loss 344 smooths the inconsistencies between classification layer 340 and the topological layer 320.” Thus, it would have been obvious to combine the prediction map with the teachings of Tian, Chen, and Ke-Feng in order to produce a prediction map indicating a probability of the portal vein branch label is attached to each image unit element for the entire first image region.
Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over Tian et al. (Tian, J., et al., “Automatic Couinaud Segmentation from CT Volumes on Liver Using GLC-UNet”, Machine Learning in Medical Imaging. MLMI 2019. Lecture Notes in Computer Science, Vol. 11861. Springer, Cham), hereinafter Tian, in view of Chen (CN 112842371), hereinafter Chen, further in view of Li et al. (CN111161241), hereinafter Li, and further in view of Masahiko et al. (JP 2003070782), hereinafter Masahiko.
As to Claim 12, Tian in view of Chen teaches extracting a liver region from the CT image included in the second input data, (see Tian, pg. 275, “Given a CT volume, a GLC-UNet is first utilized to obtain segmentation of the liver, which serves as the region of interest for Couinaud segmentation”). However, Tian in view of Chen fails to teach that label information labeled for a region other than the extracted liver region may be in validated. However, Masahiko teaches an image processor for computer tomography images (see abstract) that can remove unnecessary labels from regions labeled outside of a target region (see paragraph [0009], "Therefore, the tomographic image is converted into a binary image by the binary image creating means, and then the binary image is labeled by the labeling means. The labeling process is a process in which all connected pixels (connected components) are given the same value, and different connected components are given different values. Then, the maximum connected component extracting means extracts the maximum connected component (area having the maximum area) among the connected components, and the contour creating means creates the outermost contour of the maximum connected component. The deleting means deletes the pixels existing outside the outermost contour of the maximum connected component from the tomographic image. As a result, only the tomographic image of the region of the subject necessary for diagnosis remains"). Masahiko is combinable with Tian and Chen since all three are from the analogous field of medical image analysis. Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the label removal process taught by Masahiko with the teachings of Tian and Chen. The motivation for doing so would be to only include relevant information within the image, as taught by Masahiko in paragraph [0009].
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant’s disclosure.
Zhang et al. (Zhang, Qin, et al., “An Efficient and Clinical-Oriented 3D Liver Segmentation
Method”, 2017) teaches an algorithm for extracting hepatic portal veins, labelling the portal veins
according to their respective hepatic segment, and then segmenting the liver according to the portal
veins.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SOUMYA THOMAS whose telephone number is (571)272-8639. The examiner can normally be reached M-F 8:30-5:00.
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, Jennifer Mehmood can be reached at (571) 272-2976. 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.
/S.T./Examiner, Art Unit 2664
/JENNIFER MEHMOOD/Supervisory Patent Examiner, Art Unit 2664
Prosecution Insights