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 .
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.
Claim(s) 1-5, 10, 16-17, 19 and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Hsieh et al. (US 2023/0056923, hereinafter Hsieh), and further in view of Masumoto et al. (US 2011/0054295, hereinafter Masumoto).
Regarding claim 1, Hsieh teaches a computer-implemented method for reviewing medical images ([0021], Fig.1-3, 7: systems, computer-implemented methods, apparatus and/or computer program products that facilitate automatically detecting scan characteristics of a medical image series for workflow automation; see also example processes in [0085]-[0087], Fig.7-8.), the method comprising:
accessing a plurality of medical images of a subject, the medical images comprising at least one of a magnetic resonance imaging (MRI) medical image, a computed tomography (CT) medical image, or a positron emission tomography (PET) medical image ([0021], [0026]-[0028], Fig.1-3, 7, 102: e.g. CT exam series, MRI, PET), each medical image being a three-dimensional image comprising a plurality of two-dimensional image slices ([0033]: 2D scan images corresponding to slices of the 3D volumetric image from various perspective/orientations (e.g., relative to the axial plane, the coronal plane, the sagittal plane and other reformatted views); [0062], Fig.3: input data 302 can include the CT exam series (which consists of a series of 2D MRP images); [0071]: sub-volume dimensions);
determining values for a plurality of categories for each medical image ([0064], Fig.2: series characterization component 212 can apply one or more series characterization models (or algorithms), to automatically detect and characterize defined characteristics of the series. [0072]: The secondary series characterization models can include one or more contrast phase detection models adapted to detect whether contrast injection was performed and if so, the particular contrast phase reflected in the series);
determining a score for each medical image based on the values for the categories for each medical image ([0070]-[0071], Fig.2: Confidence scores to select possible type classifications for the series for further processing. [...] For example, the image generation component 210 can tailor the representative image generation parameters (e.g., the sub-volume dimensions, the directionality/orientation, the pixel weightings, the type of the projection (MIP vs. mIP) and so on) based on the potential type classifications to generate one or more new representative images that are tailored to the potential type classifications); and
identifying one of the medical images as an anchor medical image (representative image) based on the scores of the medical images ([0043]: The goal of the representative image generation process at 104 is to generate a single (or in some implementations two or more representative images) representative image 106 for each scan series included in the scan data 102 that captures the entire contents of the characteristics of the scan series; [0054]: if the scan data 102 consists of multiple images at each image location obtained at different scan cycles/times as in a perfusion scan, then the image generation component 210 can generate one or more representative images for each group of images within that scan where within each group there is only image at each location. In another example, the scan data 102 may include multiple series of different anatomical regions in the body, which is often the case the trauma workflow (e.g., sometimes up to a dozen). In accordance with this example, the image generation component 210 can identify the different series and generate a separate representative image (or images) for each series; [0070]: the series characterization models can include separate models for each of the different type classifications and the series characterization component 212 can be configured to apply each of the models to the representative image (or images) to determine which one provides a positive classification. In some implementations of these embodiments, in order to minimize errors attributed to false positives and/or false negatives, the different series type classification models can be adapted to generate a binary output that classifies the input images as either having or not having the specific type classification that the model is adapted to detect and a confidence score indicative of the degree of confidence in the accuracy of the model's output. For example, assume there are 15 different series type classification models each adapted to classify the representative image (or images) as either having (e.g., a positive classification) or not having (e.g., a negative classification) the specific series type classification the model is adapted to detect and provide a confidence score reflective of the degree of confidence in the result (e.g., as a percentage score with 0% being the lowest degree of confidence and 100% being the highest, or another suitable scoring scale). The series characterization component 212 can further evaluate the results of each of the models to select the most probable classification based on the binary classification and the confidence score. For example, in some implementations, if only one model provides a positive result, the series characterization component 212 can assume the series belongs to that series classification. However, if there is more than one positive result, the series classification component 212 can select the classification with the highest confidence score), the anchor medical image being used to fix the medical images (generating one or more new representative images with different orientations/perspectives when the confidence score is below a threshold level is functionally analogous to fixing or repairing one or more (defective) medical images; [0062], Fig.3: Using this input data, at 304 the image generation component 210 can generate one or more representative images for the scan series using the techniques described above; [0071]: The image generation component 210 can further be adapted to generate one or more new representative images with different orientations/perspectives for processing by the models based on the confidence score being below a threshold level of confidence) for creating a three-dimensional model of the subject ([0083], Fig.2: post-processing tasks; they can include automatic application of one or more image processing models/algorithms adapted to process medical images in the series. These image processing algorithms can include various medical image inferencing algorithms or models (e.g., AI models). For example, the image processing algorithms can include image synthesis algorithms used construct a three-dimensional image based on multiple two-dimensional images).
Hsieh does not explicitly teach each medical image comprising voxels.
Masumoto teaches each medical image comprising voxels ([0091]: In the first embodiment of the invention, modality 1 includes an MRI system. The MRI system is capable of performing dynamic imaging in which three-dimensional medical images (voxel data) V.sub.t0, V.sub.t1, - - - , V.sub.t6 (hereinafter, collectively referred to as three-dimensional dynamic images V.sub.tn) of an abdomen (including liver) are obtained before administering a contrast agent that includes gadoxetate sodium (Gd-EOB-DTPA) as an active ingredient (EOB contrast agent) and, for example, 20 sec, 1 min, 2 min, 5 min, 10 min, and 20 min after administration of the EOB contrast agent). 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 Masumoto’s knowledge of performing dynamic imaging in which voxel data represents three-dimensional MRI images as taught and modify the system of Hsieh because such a system obtains dynamic three-dimensional images corresponding to liver function angiographic images, thereby improving diagnostic efficiency ([0091], [0171]).
Claims 19-20 are similar in scope to claim 1, and therefore the examiner provides similar rationale to reject these claims. Moreover, Hsieh teaches corresponding systems, apparatus and/or computer program products (abstract, [0021], [0049], [0095]).
Regarding claim 2, Hsieh teaches the method of claim 1, wherein determining the values comprises: accessing data stored for each medical image to determine at least one of a resolution along an axis, a number of slices, a reconstruction kernel, or a reconstruction diameter for each medical image, wherein the axis is from front to back of the subject, from left to right of the subject, or from top to bottom of the subject (direction of one axis is oriented relative to the same anatomical plane is functionally analogous to the axis is from front to back of the subject, from left to right of the subject, or from top to bottom of the subject; [0041]: The metadata can also include information regarding the reconstruction parameters used to generate the 2D scan images included in the series (e.g., reconstruction kernel size, slice thickness, [...]). [0034]: slice thickness correlates with resolution; [0059]: In various embodiments, this can involve selecting/defining a sub-volume of the 3D volume image for the compression and the orientation of the dimensionality of the compression (e.g., the specific axis or orientation for the 2D image relative to the 3D volume). For example, there are three standard anatomic planes that are generally used to display data for CT scans: axial (or horizontal), coronal (cutting the body into slices that go from anterior to posterior (AP) or posterior to anterior (PA)), and sagittal or lateral (cutting the body into slices that go from right to left or left to right). However, 2D CT scan images can also be generated relative to other planes (e.g., oblique or even curved planes). As described herein, reference to scan images being generated relative to one axis (e.g., x, y or z) of a 3D coordinate system refers to the scan images being generated at different points along the direction of the one axis such that each scan image is oriented relative to the same anatomical plane; [0066]: This metadata can be used as input in parallel and/or in combination with the one or more representative images 106).
Regarding claim 3, Hsieh teaches the method of claim 1, wherein at least some of the values are determined by analyzing the medical images ([0070]: In some implementations of these embodiments, in order to minimize errors attributed to false positives and/or false negatives, the different series type classification models can be adapted to generate a binary output that classifies the input images as either having or not having the specific type classification that the model is adapted to detect and a confidence score indicative of the degree of confidence in the accuracy of the model's output. For example, assume there are 15 different series type classification models each adapted to classify the representative image (or images) as either having (e.g., a positive classification) or not having (e.g., a negative classification) the specific series type classification the model is adapted to detect and provide a confidence score reflective of the degree of confidence in the result (e.g., as a percentage score with 0% being the lowest degree of confidence and 100% being the highest, or another suitable scoring scale). The series characterization component 212 can further evaluate the results of each of the models to select the most probable classification based on the binary classification and the confidence score; [0072]-[0076]: contrast phase detection models; the image generation component 210 may generate a new representative image for a series based on its series type classification for input to a contrast phase detection model).
Regarding claim 4, Hsieh teaches the method of claim 1, wherein determining the values comprises: analyzing each medical image to assign a binary value as a contrast value for each medical image ([0070]: In some implementations of these embodiments, in order to minimize errors attributed to false positives and/or false negatives, the different series type classification models can be adapted to generate a binary output that classifies the input images as either having or not having the specific type classification that the model is adapted to detect and a confidence score indicative of the degree of confidence in the accuracy of the model's output. For example, assume there are 15 different series type classification models each adapted to classify the representative image (or images) as either having (e.g., a positive classification) or not having (e.g., a negative classification) the specific series type classification the model is adapted to detect and provide a confidence score reflective of the degree of confidence in the result (e.g., as a percentage score with 0% being the lowest degree of confidence and 100% being the highest, or another suitable scoring scale). The series characterization component 212 can further evaluate the results of each of the models to select the most probable classification based on the binary classification and the confidence score; [0072]-[0076]: contrast phase detection models; the image generation component 210 may generate a new representative image for a series based on its series type classification for input to a contrast phase detection model).
Regarding claim 5, Hsieh teaches the method of claim 1, wherein determining the values comprises: accessing data stored for each medical image to determine a reconstruction kernel for each medical image, and assigning a reconstruction kernel value for each medical image based on the reconstruction kernel for each medical image ([0041]: The scan data 102 can also include or otherwise be associated with metadata describing characteristics of the medical image series associated therewith. For example, the metadata can include identifying or indicating the type of the scan, the ROI scanned, the acquisition protocols used, the scan mode used, the acquisition plane, and the scanner device/system used. The metadata can also include information regarding the reconstruction parameters used to generate the 2D scan images included in the series (e.g., reconstruction kernel size, slice thickness, the modulation transfer function (MTF) value, slice sensitivity profile (SSP), point spread function (PSF) characteristics, etc.). The metadata may also include information identifying or indicating the relative orientation of the medical image series (e.g., axial, coronal, sagittal, or another orientation) and/or the relative location/position of each (or in some implementation one or more) scan image included in the series in 3D. In some implementations in which the scan was performed with perfusion, the metadata may also include time stamp information that identifies or indicates the relative timing of capture of each scan relative to perfusion process and/or the relative phase of the perfusion process at which time each scan image was captured/generated. The metadata may also include patient information, such as information identifying or indicating (e.g., in an anonymized manner as appropriate) the identity of the patient (e.g., a unique patient identifier such a name or anonymized identification number), demographic information for the patient (e.g., age, gender, body mass index (BMI), and relevant medical history information for the patient (e.g., comorbidities, current condition/diagnosis, current pathology, reason for performance of the scan, etc.). In some embodiments, this metadata and/or patient information can be used as input in parallel and/or in combination with the one or more representative images 106 at 108 into one or more series characterization models to facilitate inferring the series characteristics; [0066]: This metadata and/or patient information can be used as input in parallel and/or in combination with the one or more representative images 106).
Regarding claim 10, Hsieh teaches the method of claim 1, wherein the score for each medical image is based on a weighted sum of values corresponding to the values for the categories for each medical image ([0061], [0063], [0071], [0076]: transformation of the series of images into single image weighted by the centroid values of pixels in the x and y directions so that the MPR-MIP image has the most representative view of the content of the human anatomy in the scanned series; pixel weightings).
Regarding claim 16, Hsieh teaches the method of claim 1, wherein identifying one of the medical images as the anchor medical image comprises selecting the medical image having a highest score as the anchor medical image ([0071]: the series classification component 212 can select the top N positive results with the top N highest confidence scores as possible type classifications for the series for further processing with additional more refined type classification models tailored to those categories to further drill down on the correct series type classification).
Regarding claim 17, Hsieh teaches the method of claim 1, wherein identifying one of the medical images as the anchor medical image comprises: selecting at least two of the medical images as recommended medical images based on the scores of the medical images (the series classification component 212 can select the top N positive results with the top N highest confidence scores as possible type classifications for the series for further processing with additional more refined type classification models tailored to those categories to further drill down on the correct series type classification); presenting on a display an indication of the recommended medical images ([0079]-[0080]: system 200 can include visualization component 222 to facilitate reviewing the medical image series via a medical imaging application accessed via the user device (Fig.5: exemplary visualization layout 500 for a brain scan series)); and receiving user input selecting one of the recommended medical images as the anchor medical image ([0079]-[0080]: system 200 can include visualization component 222 to facilitate reviewing the medical image series via a medical imaging application accessed via the user device (Fig.5: exemplary visualization layout 500 for a brain scan series); [0081]: feedback component 224 to allow the technician to review the automatically detected characteristics and either accept and apply the annotation to the medical image series, edit the detected characteristics or reject the characteristics).
Claim(s) 7-9 is/are rejected under 35 U.S.C. 103 as being unpatentable over Hsieh, in view of Masumoto, and further in view of Ping et al. (CN 113506250, hereinafter Ping).
Regarding claim 7, Hsieh teaches the method of claim 1, wherein the categories of the values comprise contrast ([0040]: For dual energy or photon counting CT, an iodine map can be used to characterize the contrast phase of the scan. The 2D scan images may include native scan images that are generated relative to their native acquisition plane and/or reconstructed images that are regenerated from the 3D scan image volume data relative to a different acquisition plane. In some implementations, the scan data 102 may include two or more different medical image series. For example, the two or more different medical image series can correspond to different perspectives/orientations of the same anatomical region of interest (ROI) captured, different acquisition cycles, different contrast phases of a perfusion scan, and/or different ROIs; [0072]: the secondary series characterization models can include one or more contrast phase detection models adapted to detect whether contrast injection was performed and if so, the particular contrast phase reflected in the series. The contrast phases can be classified as one of a defined set of possible contrast phases. For example, in some implementations, the defined set can include either non-contrast, arterial, portal/venous, and delayed. In other implementations, the defined set can include two or more of the following phases: pre-contrast phase (or unenhanced phase), intravenous (IV) phase (IVP), arterial phase (AP), early AP, late AP, extracellular phase (ECP), portal venous phase (PVP), delayed phase (DP), transitional phase (TP), hepatobiliary phase (HBP), and variants thereof. As noted above, these contrast phase detection models can include different models tailored to different series type classification. In this regard, the series characterization component 212 can select the appropriate contrast phase detection model (or models) for applying to the representative image (or images) based the series type classification. For example, the contrast phase detection models can be optimized by anatomy. Accordingly, if a first series characterization model detects a particular anatomical ROI and/or landmark present, the series characterization component 212 can invoke the contrast phase detection model for that anatomical ROI and/or landmark (e.g., cardiac, head, abdomen, etc.). The number of clinically relevant contrast phases present in a series often vary by anatomical region. Thus, by tailoring the different contrast phase detection models to specific anatomies, the respective models will be adapted to detect the appropriate clinically relevant contrast phases) and field of view (scan field is functionally analogous to scanning field of view in medical imagining likes CT scans; [0034]: the nominal slice thickness in CT is defined as the full width at half maximum (FWHM) of the sensitivity profile, in the center of the scan field).
The combination of Hsieh and Masumoto does not explicitly teach the categories of the values comprise z-axis length, wherein a z-axis is from top to bottom of the subject.
Ping teaches the categories of the values comprise z-axis length, wherein a z-axis is from outside to inside of the subject (page 53 paragraph 7: in order to explain the position relationship between the longitudinal partition image and each connecting component, constructing a three-dimensional coordinate system in the CTPA image, comprising an X axis, a Y axis; Z axis, wherein the X axis is a positive view CTPA image from left to right coordinate axis; the Y axis is a coordinate axis of positive view CTPA image from top to bottom; the Z axis is a coordinate axis of positive view CTPA image from outside to inside. In another implementation, also can be the shooting CTPA image of the device of the world coordinate system as a reference, to describe the position relationship between the longitudinal image and each connecting component, the present disclosure is not specifically limited. Further, the position coordinate of the candidate communication component can be divided into 6 dimensions: the component-region; (0) represents the minimum coordinate of the candidate communication component on the X axis; the component-region; (1) represents the minimum coordinate of the candidate communication component on the Y-axis; the component-region; (2) represents the minimum coordinate of the candidate communication component on the Z-axis; the component-region; (3) represents the length of the candidate communication component on the X axis; the component-region_region( 4) represents the length of the candidate communication component on the Y-axis; the component-region_region( 5) represents the length of the candidate communication component on the Z-axis. Similarly, the position coordinate of the longitudinal image can be divided into 6 dimensions: MD-region (0) represents the minimum coordinate of middle-longitudinal-partition image on the X axis; MD-region (1) represents the minimum coordinate of the longitudinal-partition image on the Y-axis; MD-region (2) represents the minimum coordinate of the longitudinal-partition image on the Z-axis; MD-region( 3) represents the length of the longitudinal partition image on the X axis, MD-region( 4) represents the length of the longitudinal partition image on the Y-axis, MD-region( 5) represents the length of the longitudinal partition image on the Z-axis). 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 Ping’s knowledge of using z-axis length as a value for categories as taught and modify the system of Hsieh and Masumoto because such a system ensures the accuracy of lung parenchyma binary image (page 50 paragraph 6).
Although Ping teaches z-axis is from outside to inside of the subject, Ping does not explicitly teach the z-axis is from top to bottom of the subject. However, Ping teaches the y-axis is from top to bottom of the subject (page 53 paragraph 7). It would have been prima facie obvious for the z-axis to be from top to bottom of the subject and the y-axis to be from outside to inside. Whether the z-axis is from top to bottom or y-axis is from top to bottom of the subject is solely a matter of aesthetic design choice, and would not be sufficient to distinguish over the prior art. See MPEP 2144.04.
Regarding claim 8, Hsieh teaches the method of claim 1, wherein the categories of the values comprise contrast ([0040]: For dual energy or photon counting CT, an iodine map can be used to characterize the contrast phase of the scan. The 2D scan images may include native scan images that are generated relative to their native acquisition plane and/or reconstructed images that are regenerated from the 3D scan image volume data relative to a different acquisition plane. In some implementations, the scan data 102 may include two or more different medical image series. For example, the two or more different medical image series can correspond to different perspectives/orientations of the same anatomical region of interest (ROI) captured, different acquisition cycles, different contrast phases of a perfusion scan, and/or different ROIs; [0072]: the secondary series characterization models can include one or more contrast phase detection models adapted to detect whether contrast injection was performed and if so, the particular contrast phase reflected in the series. The contrast phases can be classified as one of a defined set of possible contrast phases. For example, in some implementations, the defined set can include either non-contrast, arterial, portal/venous, and delayed. In other implementations, the defined set can include two or more of the following phases: pre-contrast phase (or unenhanced phase), intravenous (IV) phase (IVP), arterial phase (AP), early AP, late AP, extracellular phase (ECP), portal venous phase (PVP), delayed phase (DP), transitional phase (TP), hepatobiliary phase (HBP), and variants thereof. As noted above, these contrast phase detection models can include different models tailored to different series type classification. In this regard, the series characterization component 212 can select the appropriate contrast phase detection model (or models) for applying to the representative image (or images) based the series type classification. For example, the contrast phase detection models can be optimized by anatomy. Accordingly, if a first series characterization model detects a particular anatomical ROI and/or landmark present, the series characterization component 212 can invoke the contrast phase detection model for that anatomical ROI and/or landmark (e.g., cardiac, head, abdomen, etc.). The number of clinically relevant contrast phases present in a series often vary by anatomical region. Thus, by tailoring the different contrast phase detection models to specific anatomies, the respective models will be adapted to detect the appropriate clinically relevant contrast phases), z-axis resolution (thickness is inherently on z-axis and therefore resolution based on slice thickness is functionally analogous to z-axis resolution; [0034]: It should be appreciated that a thin slice has a smaller thickness than a thick slice. In accordance with most 3D medical imaging modalities (e.g., CT, MRI, PET, etc.), the native resolution of thin scan images (e.g., obtained with thin detectors) is higher than the native resolution of thicker scan images (e.g., obtained with thicker detectors). For example, the nominal slice thickness in CT is defined as the full width at half maximum (FWHM) of the sensitivity profile, in the center of the scan field; its value can be selected by the operator according to the clinical requirement and generally lies in the range between 1 millimeter (mm) and 10 mm In general, the larger the slice thickness, the greater the low contrast resolution in the image, while the smaller the slice thickness, the greater the spatial resolution), field of view (scan field is functionally analogous to scanning field of view in medical imagining likes CT scans; [0034]: the nominal slice thickness in CT is defined as the full width at half maximum (FWHM) of the sensitivity profile, in the center of the scan field), slice thickness ([0034]: the nominal slice thickness in CT is defined as the full width at half maximum (FWHM) of the sensitivity profile, in the center of the scan field; [0041]: The metadata can also include information regarding the reconstruction parameters used to generate the 2D scan images included in the series (e.g., reconstruction kernel size, slice thickness, the modulation transfer function (MTF) value, slice sensitivity profile (SSP), point spread function (PSF) characteristics, etc.)), and reconstruction kernel ([0041]: The metadata can also include information regarding the reconstruction parameters used to generate the 2D scan images included in the series (e.g., reconstruction kernel size, slice thickness, the modulation transfer function (MTF) value, slice sensitivity profile (SSP), point spread function (PSF) characteristics, etc.)).
The combination of Hsieh and Masumoto does not explicitly teach the categories of the values comprise z-axis length, wherein a z-axis is from top to bottom of the subject.
Ping teaches the categories of the values comprise z-axis length, wherein a z-axis is from outside to inside of the subject (page 53 paragraph 7: in order to explain the position relationship between the longitudinal partition image and each connecting component, constructing a three-dimensional coordinate system in the CTPA image, comprising an X axis, a Y axis; Z axis, wherein the X axis is a positive view CTPA image from left to right coordinate axis; the Y axis is a coordinate axis of positive view CTPA image from top to bottom; the Z axis is a coordinate axis of positive view CTPA image from outside to inside. In another implementation, also can be the shooting CTPA image of the device of the world coordinate system as a reference, to describe the position relationship between the longitudinal image and each connecting component, the present disclosure is not specifically limited. Further, the position coordinate of the candidate communication component can be divided into 6 dimensions: the component-region; (0) represents the minimum coordinate of the candidate communication component on the X axis; the component-region; (1) represents the minimum coordinate of the candidate communication component on the Y-axis; the component-region; (2) represents the minimum coordinate of the candidate communication component on the Z-axis; the component-region; (3) represents the length of the candidate communication component on the X axis; the component-region_region( 4) represents the length of the candidate communication component on the Y-axis; the component-region_region( 5) represents the length of the candidate communication component on the Z-axis. Similarly, the position coordinate of the longitudinal image can be divided into 6 dimensions: MD-region (0) represents the minimum coordinate of middle-longitudinal-partition image on the X axis; MD-region (1) represents the minimum coordinate of the longitudinal-partition image on the Y-axis; MD-region (2) represents the minimum coordinate of the longitudinal-partition image on the Z-axis; MD-region( 3) represents the length of the longitudinal partition image on the X axis, MD-region( 4) represents the length of the longitudinal partition image on the Y-axis, MD-region( 5) represents the length of the longitudinal partition image on the Z-axis). 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 Ping’s knowledge of using z-axis length as a value for categories as taught and modify the system of Hsieh and Masumoto because such a system ensures the accuracy of lung parenchyma binary image (page 50 paragraph 6).
Although Ping teaches z-axis is from outside to inside of the subject, Ping does not explicitly teach the z-axis is from top to bottom of the subject. However, Ping teaches the y-axis is from top to bottom of the subject (page 53 paragraph 7). It would have been prima facie obvious for the z-axis to be from top to bottom of the subject and the y-axis to be from outside to inside. Whether the z-axis is from top to bottom or y-axis is from top to bottom of the subject is solely a matter of aesthetic design choice, and would not be sufficient to distinguish over the prior art. See MPEP 2144.04.
Regarding claim 9, Hsieh teaches the method of claim 1, wherein the categories of the values comprise at least six of: number of slices, slice thickness ([0034]: the nominal slice thickness in CT is defined as the full width at half maximum (FWHM) of the sensitivity profile, in the center of the scan field; [0041]: The metadata can also include information regarding the reconstruction parameters used to generate the 2D scan images included in the series (e.g., reconstruction kernel size, slice thickness, the modulation transfer function (MTF) value, slice sensitivity profile (SSP), point spread function (PSF) characteristics, etc.)), reconstruction kernel ([0041]: The metadata can also include information regarding the reconstruction parameters used to generate the 2D scan images included in the series (e.g., reconstruction kernel size, slice thickness, the modulation transfer function (MTF) value, slice sensitivity profile (SSP), point spread function (PSF) characteristics, etc.)), reconstruction kernel value, reconstruction diameter, contrast ([0040]: For dual energy or photon counting CT, an iodine map can be used to characterize the contrast phase of the scan. The 2D scan images may include native scan images that are generated relative to their native acquisition plane and/or reconstructed images that are regenerated from the 3D scan image volume data relative to a different acquisition plane. In some implementations, the scan data 102 may include two or more different medical image series. For example, the two or more different medical image series can correspond to different perspectives/orientations of the same anatomical region of interest (ROI) captured, different acquisition cycles, different contrast phases of a perfusion scan, and/or different ROIs), contrast value ([0072]: the secondary series characterization models can include one or more contrast phase detection models adapted to detect whether contrast injection was performed and if so, the particular contrast phase reflected in the series. The contrast phases can be classified as one of a defined set of possible contrast phases. For example, in some implementations, the defined set can include either non-contrast, arterial, portal/venous, and delayed. In other implementations, the defined set can include two or more of the following phases: pre-contrast phase (or unenhanced phase), intravenous (IV) phase (IVP), arterial phase (AP), early AP, late AP, extracellular phase (ECP), portal venous phase (PVP), delayed phase (DP), transitional phase (TP), hepatobiliary phase (HBP), and variants thereof. As noted above, these contrast phase detection models can include different models tailored to different series type classification. In this regard, the series characterization component 212 can select the appropriate contrast phase detection model (or models) for applying to the representative image (or images) based the series type classification. For example, the contrast phase detection models can be optimized by anatomy. Accordingly, if a first series characterization model detects a particular anatomical ROI and/or landmark present, the series characterization component 212 can invoke the contrast phase detection model for that anatomical ROI and/or landmark (e.g., cardiac, head, abdomen, etc.). The number of clinically relevant contrast phases present in a series often vary by anatomical region. Thus, by tailoring the different contrast phase detection models to specific anatomies, the respective models will be adapted to detect the appropriate clinically relevant contrast phases), field of view (scan field is functionally analogous to scanning field of view in medical imagining likes CT scans; [0034]: the nominal slice thickness in CT is defined as the full width at half maximum (FWHM) of the sensitivity profile, in the center of the scan field), times since medical image obtained ([0041]: In some implementations in which the scan was performed with perfusion, the metadata may also include time stamp information that identifies or indicates the relative timing of capture of each scan relative to perfusion process and/or the relative phase of the perfusion process at which time each scan image was captured/generated), x-axis resolution, y-axis resolution, z-axis resolution (thickness is inherently on z-axis and therefore resolution based on slice thickness is functionally analogous to z-axis resolution; [0034]: It should be appreciated that a thin slice has a smaller thickness than a thick slice. In accordance with most 3D medical imaging modalities (e.g., CT, MRI, PET, etc.), the native resolution of thin scan images (e.g., obtained with thin detectors) is higher than the native resolution of thicker scan images (e.g., obtained with thicker detectors). For example, the nominal slice thickness in CT is defined as the full width at half maximum (FWHM) of the sensitivity profile, in the center of the scan field; its value can be selected by the operator according to the clinical requirement and generally lies in the range between 1 millimeter (mm) and 10 mm In general, the larger the slice thickness, the greater the low contrast resolution in the image, while the smaller the slice thickness, the greater the spatial resolution), x-axis length, y-axis length, z-axis length, front margin of the x-axis length, back margin of the x-axis length, left margin of the y-axis length, right margin of the y-axis length, top margin of the z-axis length, and bottom margin of the z-axis length.
The combination of Hsieh and Masumoto does not explicitly teach wherein an x-axis is from front to back of the subject, a y-axis is from left to right of the subject, and a z-axis is from top to bottom of the subject.
Ping teaches an x-axis is from front to back of the subject, a y-axis is from left to right of the subject, and a z-axis is from top to bottom of the subject (page 53 paragraph 7: in order to explain the position relationship between the longitudinal partition image and each connecting component, constructing a three-dimensional coordinate system in the CTPA image, comprising an X axis, a Y axis; Z axis, wherein the X axis is a positive view CTPA image from left to right coordinate axis; the Y axis is a coordinate axis of positive view CTPA image from top to bottom; the Z axis is a coordinate axis of positive view CTPA image from outside to inside. In another implementation, also can be the shooting CTPA image of the device of the world coordinate system as a reference, to describe the position relationship between the longitudinal image and each connecting component, the present disclosure is not specifically limited. Further, the position coordinate of the candidate communication component can be divided into 6 dimensions: the component-region; (0) represents the minimum coordinate of the candidate communication component on the X axis; the component-region; (1) represents the minimum coordinate of the candidate communication component on the Y-axis; the component-region; (2) represents the minimum coordinate of the candidate communication component on the Z-axis; the component-region; (3) represents the length of the candidate communication component on the X axis; the component-region_region( 4) represents the length of the candidate communication component on the Y-axis; the component-region_region( 5) represents the length of the candidate communication component on the Z-axis. Similarly, the position coordinate of the longitudinal image can be divided into 6 dimensions: MD-region (0) represents the minimum coordinate of middle-longitudinal-partition image on the X axis; MD-region (1) represents the minimum coordinate of the longitudinal-partition image on the Y-axis; MD-region (2) represents the minimum coordinate of the longitudinal-partition image on the Z-axis; MD-region( 3) represents the length of the longitudinal partition image on the X axis, MD-region( 4) represents the length of the longitudinal partition image on the Y-axis, MD-region( 5) represents the length of the longitudinal partition image on the Z-axis). 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 Ping’s knowledge of using z-axis length as a value for categories as taught and modify the system of Hsieh and Masumoto because such a system ensures the accuracy of lung parenchyma binary image (page 50 paragraph 6).
Although Ping teaches the X axis is from left to right coordinate axis; the Y axis is from top to bottom; the Z axis is from outside to inside, Ping does not explicitly teach an x-axis is from front to back of the subject, a y-axis is from left to right of the subject, and a z-axis is from top to bottom of the subject. However, it would have been prima facie obvious for an x-axis is from front to back of the subject, a y-axis is from left to right of the subject, and a z-axis is from top to bottom of the subject. Whether the an x-axis is from front to back of the subject, a y-axis is from left to right of the subject, and a z-axis is from top to bottom of the subject is solely a matter of aesthetic design choice, and would not be sufficient to distinguish over the prior art. See MPEP 2144.04.
Claim(s) 18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Hsieh, in view of Masumoto, and further in view of Hershkovich et al. (US 2020/0372705, hereinafter Hershkovich).
Regarding claim 18, Hsieh teaches the method of claim 1, further comprising: creating a three-dimensional model of the subject ([0083], Fig.2: post-processing tasks; they can include automatic application of one or more image processing models/algorithms adapted to process medical images in the series. These image processing algorithms can include various medical image inferencing algorithms or models (e.g., AI models). For example, the image processing algorithms can include image synthesis algorithms used construct a three-dimensional image based on multiple two-dimensional images).
The combination of Hsieh and Masumoto does not explicitly teach creating a three-dimensional model of the subject based on the anchor medical image and the medical images; generating a plurality of transducer layouts for application of tumor treating fields to the subject based on the three-dimensional model of the subject; selecting at least two of the transducer layouts as recommended transducer layouts; presenting the recommended transducer layouts; receiving a user selection of at least one recommended transducer layout; and providing a report for the at least one selected recommended transducer layout.
Hershkovich teaches creating a three-dimensional model of the subject based on the anchor medical image and the medical images ([0091]: the primary ("anchor") image will be used to generate the computational 3D model and/or transducer array layout map); generating a plurality of transducer layouts for application of tumor treating fields to the subject based on the three-dimensional model of the subject ([0075]: Sets of transducer array layout maps of the plurality of sets of transducer array layout maps may be determined that include two or more transducer array layout maps that include non-overlapping positions for transducer array placement); selecting at least two of the transducer layouts as recommended transducer layouts ([0075]: The plurality of sets of transducer array layout maps may be displayed, for example, to a user, and/or be selectable, for example, via a user interface); presenting the recommended transducer layouts ([0075]: The plurality of sets of transducer array layout maps may be displayed, for example, to a user, and/or be selectable, for example, via a user interface); receiving a user selection of at least one recommended transducer layout ([0075]: The user interface may be used to select a transducer array layout map; [0089]-[0098], Fig.11A-D: screens of an example interface); and providing a report for the at least one selected recommended transducer layout ([0076]: A report that describes the selected transducer array layout may be generated). 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 Hershkovich’s knowledge as taught and modify the system of Hsieh and Masumoto because such a system optimizes array placement on the patient's scalp to increase the intensity in the diseased region of the brain ([0055]).
Relevant Prior Art
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Singhal et al. (US 2024/0193851) describes determining, by an image aesthetics predictor module in the 360-degree view generation system, the representative image for each view category of the plurality of view categories comprises calculating, by the image aesthetics predictor module, a global quality score for each image from the view category using a deep-learning-based image aesthetics predictor model, wherein the representative image has a maximum quality score in the view category.
Iwasaki (US 2012/0250961) describes three-dimensional medical image data are volume data composed of three-dimensionally-arranged voxels, and which are obtained by imaging by a CT apparatus, an MRI apparatus, and the like. Further, the medical report database 8 registers an image ID for identifying an image-reading target image or a representative image, an image-read person's ID for identifying a doctor who read the image, position information about a region of interest, findings, and information such as a degree of certainty of the findings.
Allowable Subject Matter
Claims 6 and 11-15 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
The following is a statement of reasons for the indication of allowable subject matter:
Regarding claim 6, none of the cited prior art references of record, teach either individually or in combination, “wherein determining the values comprises: accessing data stored for each medical image to determine a z-axis length for each medical image, wherein a z-axis is from top to bottom of the subject; and analyzing each medical image to determine at least one of a top margin of the z-axis length above a designated area of the subject or a bottom margin of the z-axis length below the designated area of the subject”.
Regarding claim 11, none of the cited prior art references of record, teach either individually or in combination, “wherein the score for each medical image is based on a weighted sum of at least a contrast value, a normalized field of view value, and a normalized z- axis length value for each medical image, wherein a z-axis is from top to bottom of the subject”.
Regarding claim 12, none of the cited prior art references of record, teach either individually or in combination, “wherein the score for each medical image is based on a weighted sum of at least a contrast value, a normalized z-axis resolution value, a normalized field of view value, a normalized slice thickness value, a reconstruction kernel value, and a normalized z-axis length value for each medical image, wherein a z-axis is from top to bottom of the subject”.
Regarding claim 13, none of the cited prior art references of record, teach either individually or in combination, “wherein the score for each medical image is based on a weighted sum of values corresponding to the values for the categories for each medical image, wherein the categories for each medical image comprise contrast value and field of view, wherein a weight for a contrast value is larger than a weight for a normalized field of view value”.
Regarding claim 14, none of the cited prior art references of record, teach either individually or in combination, “wherein the score for each medical image is based on a weighted sum of values corresponding to the values for the categories for each medical image, wherein the categories for each medical image comprise field of view and z-axis length, wherein a weight for a normalized field of view value is larger than a weight for a normalized z-axis length value, wherein a z-axis is from top to bottom of the subject”.
Regarding claim 15, none of the cited prior art references of record, teach either individually or in combination, “wherein the score for each medical image is based on a weighted sum of values corresponding to the values for the categories for each medical image, wherein the categories for each medical image comprise field of view, resolution, and slice thickness, wherein a normalized field of view value, a resolution value, and a normalized slice thickness value have a same weight”.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JWALANT B AMIN whose telephone number is (571)272-2455. The examiner can normally be reached Monday-Friday 10am - 630pm CST.
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, Said Broome can be reached at 571-272-2931. 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.
/JWALANT AMIN/Primary Examiner, Art Unit 2612
Prosecution Insights