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 .
Specification
The abstract of the disclosure is objected to because it is over 150 words. A corrected abstract of the disclosure is required and must be presented on a separate sheet, apart from any other text. See MPEP § 608.01(b).
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.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The method of claim 1 is directed to a process, which is one of the statutory categories of invention, and passes Step 1: Statutory Category- MPEP § 2106.03. However, the following elements of Claim 1 recite steps that can be performed in the human mind or with pen and paper, therefore failing Step 2A Prong One. These steps constitute mental processes because they describe acts of observation, evaluation, and judgement that a human can perform mentally. For example, a human (e.g., radiologist) can observe an image and recognize lesions, visually identify a lesion and draw a box around it, outline lesion boundaries manually, and determine lesion boundaries and shade/outline affected regions.
detecting, using an object detection model, a region of interest in the medical image, the region of interest comprising a lesion in the organ;
demarcating, using the object detection model, a bounding box around the region of interest;
segmenting, using an image segmentation model that is independent from the object detection model, the localized image comprising the region of interest;
predicting, using the image segmentation model, a segmentation mask of the lesion in the localized image;
Claim 1 fails Step 2A Prong Two because the additional elements beyond the judicial exception, including an object detection model trained using a first dataset of training images comprising medical images of the organ, extraction of the localized image comprising the region of interest, an image segmentation model, and outputting the segmentation mask from the localized image to the medical image, do not integrate the judicial exception into a practical application. Extracting the localized image and outputting the segmentation mask are insignificant extra-solution activity (MPEP § 2106.05(g)), and the object detection and image segmentation model do not improve the functioning of a computer or any other technology or technical field (MPEP § 2106.05(a)) as they merely apply the abstract idea on a computer (MPEP § 2106.05(f)). The claim also does not impose meaningful limits on the computer components such that the method is tied to a particular machine; the object detection and image segmentation model may operate on any generic computing system (MPEP § 2106.05(b)).
Claim 1 also fails Step 2B, as these generic elements are well-understood, routine, and conventional (WURC), adding nothing significantly more than the abstract idea itself (MPEP § 2106.07(a)(III)). Extracting the localized image, training the object detection model using a first dataset of training images, and outputting the segmentation mask are WURC (see MPEP § 2106.05(d)). Additionally, the object detection model and image segmentation model are WURC; see Introduction section of Chilamkurthy et. al, “Development and Validation of Deep Learning Algorithms for Detection of Critical Findings in Head CT Scans.” Accordingly, claim 1 is rejected. Claims 2, 11, 12, and 13 contain this identical ineligible subject matter, with the only additional elements beyond the judicial exception being a processor and non-transitory computer readable medium. These additional elements do not integrate the judicial exception into a practical application (see claim 1 analysis above) and are WURC (see MPEP § 2106.05(d)). Therefore, they are rejected.
Claims 3-7 fail Step 2A Prong Two as the additional elements beyond the judicial exception, including the trained object detection model comprising a YOLOv5 model optimized using a genetic algorithm, a second dataset of training images comprising localized images of lesions in the organ used to train the image segmentation model, and the image segmentation model comprising a TransDeepLab and untrained Expectation-Maximization algorithm, do not integrate the judicial exception into a practical application (see claim 1 analysis above). Additionally, claims 8-10 fail Step 2A Prong Two as pre-processing the medical image is insignificant extra-solution activity (MPEP § 2106.05(g)), and specifying the organ as the brain and the lesion as an intracranial hemorrhage merely links the use of the judicial exception to a particular field of use (MPEP § 2106.05(h)). Regarding Step 2B, training the image segmentation model using a second dataset of localized images of lesions in the organ and pre-processing the medical image before detecting the region of interest (i.e., windowing the medical image and/or removing regions of skull tissue and calcification in the medical image) are WURC (see MPEP § 2106.05(d)). Additionally, the YOLOv5 model optimized using a genetic algorithm and the TransDeepLab and untrained Expectation-Maximization algorithm in the image segmentation model are WURC; see Introduction section of Khanam and Hussain (“WHAT IS YOLOV5: A DEEP LOOK INTO THE INTERNAL FEATURES OF THE POPULAR OBJECT DETECTOR”), section 3.1. Improved Genetic Algorithm of Zhou et. al (“Optimization of Hyperparameters in Object Detection Models Based on Fractal Loss Function”), Abstract and Introduction section of Azad et. al (“TransDeepLab: Convolution-Free Transformer-Based DeepLab v3+ for Medical Image Segmentation”), and Abstract of Balafar (“Spatial based Expectation Maximizing (EM)”). Accordingly, claims 3-10 fail Step 2B and are rejected. Claims 14-20 contain this identical ineligible subject matter and are also rejected.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-2, 5, 8-10, 11-13, 16, and 19-20 are rejected under 35 U.S.C. 102 (a)(1) as being anticipated by Chilamkurthy et. al (“Development and Validation of Deep Learning Algorithms for Detection of Critical Findings in Head CT Scans”).
Regarding Claim 1, Chilamkurthy teaches a computerized method for segmenting a medical image of an organ, the method comprising:
Introduction, pg. 2: “In this manuscript, we describe the development, validation and clinical testing of fully automated deep learning algorithms that are trained to detect abnormalities requiring urgent attention from non-contrast head CT scans.”
detecting, using an object detection model, a region of interest in the medical image, the region of interest comprising a lesion in the organ;
Introduction, pg. 2: “The trained algorithms detect five kinds of intracranial hemorrhages (ICH) namely intraparenchymal (IPH), intraventricular (IVH), subdural (SDH), extradural (EDH) and subarachnoid (SAH), and calvarial/cranial vault fractures. Algorithms also detect mass effect and midline shift, both used as indicators of severity of the brain injury.”
Explanation: The hemorrhage/lesion locations constitute the claimed region of interest. The deep learning algorithms detect lesions in the organ (brain).
demarcating, using the object detection model, a bounding box around the region of interest;
2.3.3 Calvarial Fractures: “Each slice in these scans was annotated by marking a tight bounding box around fractures.”
Explanation: Explicit disclosure of bounding boxes around detected abnormalities/regions of interest.
extracting a localized image comprising the region of interest from the medical image, the localized image defined by the bounding box;
2.3.3 Calvarial Fractures: “We engineered features representative of local fracture lesions and their volumes from the generated heatmaps of the whole scan.”
Explanation: The reference processes localized lesion regions generated from the bounded/localized area. The extraction/localization is shown from using tight bounding boxes and localized lesion features.
segmenting, using an image segmentation model that is independent from the object detection model, the localized image comprising the region of interest;
2.3.1 Intracranial Hemorrhage: “We used ResNet18 [35], a popular convolutional neural network architecture with a slight modification to predict SoftMax based confidences [41] for the presence of each type of hemorrhage in a slice…We further trained a model to localize the following type of hemorrhages: IPH, SDH, EDH. Localization requires dense prediction [32] of presence or absence of bleed for every pixel in the scan…we used a UNet [33] based architecture for segmentation of each type of hemorrhage.”
Explanation: The reference teaches separate models: a ResNet18-based detection/classification model and an independent UNet segmentation model.
predicting, using the image segmentation model, a segmentation mask of the lesion in the localized image; and
PNG
media_image1.png
328
580
media_image1.png
Greyscale
2.3.1 Intracranial Hemorrhage: “Localization requires dense prediction [32] of presence or absence of bleed for every pixel in the scan.”
4. Discussion: “Since we also trained segmentation networks for hemorrhage detection algorithms, we can also output a mask representing the precise location and extent of the hemorrhage (except for the subarachnoid hemorrhage), in addition to detecting its presence.”
Explanation: Explicit disclosure of segmentation masks for hemorrhages/lesions. Figure 4a visually demonstrates the segmentation mask output.
outputting the segmentation mask from the localized image to the medical image to facilitate medical diagnosis of the lesion (Figure 4 (shown above)),
5. Conclusion: “The strong performance of deep learning algorithms suggest they could be a helpful adjunct for identification of acute Head CT finding in a trauma setting, providing a lower performance bound for quality and consistency of radiologic interpretation.”
Explanation: The segmentation/localization outputs are overlaid/displayed for diagnostic assistance, facilitating medical diagnosis.
wherein the object detection model is trained using a first dataset of training images comprising medical images of the organ, each medical image in the first dataset comprising a ground-truth bounding box for one or more lesions in the organ.
Abstract: “We retrospectively collected a dataset containing 313,318 head CT scans along with their clinical reports from various centers.”
2.1 Datasets: “Of these scans, we earmarked scans of 23,163 randomly selected patients (Qure25k dataset) for validation and used the scans of rest of the patients (development dataset) to train/develop the algorithms.”
2.3.3 Calvarial Fractures: “Each slice in these scans was annotated by marking a tight bounding box around fractures.”
Explanation: Explicit disclosure of training datasets of medical images (head CT scans) and ground-truth bounding box annotations for lesions/abnormalities.
Regarding Claim 2, Chilamkurthy teaches the method according to claim 1, comprising:
detecting, using the object detection model, a plurality of regions of interest in the medical image, each region of interest comprising a lesion in the organ;
Introduction, pg. 2: “The trained algorithms detect five kinds of intracranial hemorrhages (ICH) namely intraparenchymal (IPH), intraventricular (IVH), subdural (SDH), extradural (EDH) and subarachnoid (SAH), and calvarial/cranial vault fractures.”
2.3.1 Intracranial Hemorrhage: “Each slice in these scans was manually labeled with the hemorrhages that are visible in that slice.”
Explanation: The reference detects multiple hemorrhage lesions/abnormalities in the same medical image, corresponding to a plurality of regions of interest.
demarcating, using the object detection model, the bounding box around the plurality of regions of interest;
2.3.3 Calvarial Fractures: “Each slice in these scans was annotated by marking a tight bounding box around fractures.”
Explanation: Explicit disclosure of bounding boxes surrounding lesion regions. Multiple fracture regions/lesions may be bounded in scans.
extracting the localized image comprising the plurality of regions of interest from the medical image;
2.3.3 Calvarial Fractures: “We engineered features representative of local fracture lesions and their volumes from the generated heatmaps of the whole scan.”
Explanation: The system derives localized lesion regions from the overall medical image for further analysis.
segmenting, using the image segmentation model, the localized image comprising the plurality of regions of interest;
2.3.1 Intracranial Hemorrhage: “We further trained a model to localize the following type of hemorrhages: IPH, SDH, EDH. Localization requires dense prediction [32] of presence or absence of bleed for every pixel in the scan…we used a UNet [33] based architecture for segmentation of each type of hemorrhage.”
Explanation: The UNet segmentation model segments multiple hemorrhage lesion regions within localized image data.
generating, using the image segmentation model, segmentation masks of the respective lesions in the localized image (Figure 4 (shown above)); and
4. Discussion: “Since we also trained segmentation networks for hemorrhage detection algorithms, we can also output a mask representing the precise location and extent of the hemorrhage (except for the subarachnoid hemorrhage), in addition to detecting its presence.”
Explanation: Explicit disclosure of generating segmentation masks corresponding to detected hemorrhage lesions.
outputting the segmentation masks from the localized image to the medical image to facilitate medical diagnosis of the lesions (Figure 4 (shown above)).
5. Conclusion: “The strong performance of deep learning algorithms suggest they could be a helpful adjunct for identification of acute Head CT finding in a trauma setting, providing a lower performance bound for quality and consistency of radiologic interpretation.”
Regarding Claim 5, Chilamkurthy teaches the method according to claim 1, wherein the image segmentation model is independently trained using a second dataset of training images comprising localized images of lesions in the organ, each localized image in the second dataset comprising one or more ground-truth segmentation masks for one or more lesions in the organ.
2.3 Developing the algorithm: “In this study, we trained separate deep learning models for each of the subtasks viz. intracranial bleeds, midline shift/mass effect and calvarial fractures which we describe below.”
2.3.1 Intracranial Hemorrhage: “To train models for dense predictions, pixels corresponding to each bleed were annotated for a subset of the above slice-annotated images to provide the ground truth for the model. This set contained 1706 images of which number of images with IPH, SDH, EDH and neither of these are 506, 243, 277 and 750 respectively. We used a UNet [33] based architecture for segmentation of each type of hemorrhage.”
Explanation: This explicitly teaches independently trained models for different subtasks, including segmentation. The segmentation model is trained using a separate annotated image dataset containing lesion-localized hemorrhage images. Pixel-level based annotations correspond to ground-truth segmentation masks for lesions.
Regarding Claim 8, Chilamkurthy teaches the method according to claim 1, further comprising pre-processing the medical image before detecting the region of interest.
2.3.4 Preprocessing: “For a given CT scan, we selected the non-contrast axial series which uses soft reconstruction kernel and resampled it so that slice thickness is around 5mm. We then resized all the slices of this series to a size of 224 × 224 pixels before passing to our deep learning models.”
Regarding Claim 9, Chilamkurthy teaches the method according to claim 8, wherein said pre-processing of the medical image comprises windowing the medical image and/or removing regions of skull tissue and calcification in the medical image.
2.3.4 Preprocessing: “Instead of passing the whole dynamic range of CT densities as a single channel, we windowed the densities by using three separate windows and stacking them as channels. Windows used were brain window (l = 40, w = 80), bone window (l = 500, w = 3000) and subdural window (l = 175, w = 50) … Subdural window helps differentiate between the skull and an extra axial bleed that might have been indistinguishable in a normal brain window [46].”
Explanation: Explicit disclosure of image windowing preprocessing, where the preprocessing differentiates skull tissue from bleed regions.
Regarding Claim 10, Chilamkurthy teaches the method according to claim 1, wherein the organ is a brain, and the lesion is an intracranial hemorrhage.
Abstract: “Non-contrast head CT scan is the current standard for initial imaging of patients with head trauma or stroke symptoms.”
Introduction, pg. 2: “The trained algorithms detect five kinds of intracranial hemorrhages (ICH) namely intraparenchymal (IPH), intraventricular (IVH), subdural (SDH), extradural (EDH) and subarachnoid (SAH), and calvarial/cranial vault fractures.”
Regarding Claim 11, Chilamkurthy teaches the all of the limitations of claim 1 above. Chilamkurthy further teaches a non-transitory computer-readable medium having stored thereon instructions that, when executed, cause a processor to perform the computerized method according to claim 1 through the disclosed deep learning models.
Regarding Claim 12, Chilamkurthy teaches the all of the limitations of claim 1 above. Chilamkurthy further teaches a processor that performs substantially the same steps as claim 1 through the disclosed deep learning models.
Regarding Claim 13, Chilamkurthy teaches the system according to claim 12, and additional limitations are met as in the consideration of claim 2 above.
Regarding Claim 16, Chilamkurthy teaches the system according to claim 12, and additional limitations are met as in the consideration of claim 5 above.
Regarding Claim 19, Chilamkurthy teaches the system according to claim 12, and additional limitations are met as in the consideration of claim 8 above.
Regarding Claim 20, Chilamkurthy teaches the system according to claim 19, and additional limitations are met as in the consideration of claim 9 above.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 3 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Chilamkurthy et. al in view of Kothala et. al (“Localization of mixed intracranial hemorrhages by using a ghost convolution-based YOLO network”).
Regarding Claim 3, Chilamkurthy teaches the method according to claim 1, but fails to teach that the trained object detection model comprises a YOLOv5 model.
However, Kothala teaches a YOLOv5-based object detection model for detecting and localizing intracranial hemorrhages in CT images, stating that “we proposed a novel YOLOv5x-GCB model that can be able to detect multiple hemorrhages with limited resources by employing a ghost convolution process” (Abstract), and “YOLOv5x was used in the localization of mixed ICH” (Introduction, pg. 2).
Thus, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to substitute the generic object detection/localization approach of Chilamkurthy with the known YOLOv5 object detection model of Kothala in order to improve hemorrhage localization speed and detection accuracy in medical CT images. Kothala teaches that YOLOv5 improves detection efficiency and accuracy, stating that “YOLO has become a subject of interest in the computer vision field because of its fast-computing capability” (3. Proposed methodology, pg. 3) and “YOLOv5x has the highest accuracy for detecting mixed ICH” (3.1.2. Backbone, pg. 7). One of ordinary skill in the art would have recognized that substituting YOLOv5 into the detection framework of Chilamkurthy would have predictably yielded improved object detection speed and accurate hemorrhage localization because Kothala teaches YOLOv5’s fast-computing capability and high detection performance for mixed intracranial hemorrhage localization. This represents the simple substation of one known element for another, which would have yielded predictable results.
Regarding Claim 14, Chilamkurthy teaches the system according to claim 12, and additional limitations are met as in the consideration of claim 3 above.
Claims 4 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Chilamkurthy et. al in view of Kothala et. al, further in view of Zhou et. al (“Optimization of Hyperparameters in Object Detection Models Based on Fractal Loss Function”).
Regarding Claim 4, Chilamkurthy in view of Kothala teaches the method according to claim 3, but fails to teach that the hyperparameters of the YOLOv5 model are optimized using a genetic algorithm.
However, Zhou teaches using a genetic algorithm to optimize hyperparameters in object detection neural networks, stating that an “improved genetic algorithm is proposed to optimize the hyperparameters of the network by taking the loss function as the research object…the improved genetic algorithm is used to optimize the hyperparameters of the neural network” (Abstract).
Thus, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the system of Chilamkurthy and Kothala to optimize the hyperparameters of the YOLOv5 object detection model of Kothala using the genetic algorithm optimization technique taught by Zhou in order to improve hemorrhage detection accuracy and model performance. Zhou teaches that hyperparameters directly affect object detection model performance, stating that “hyperparameters involved in neural networks (NNs) have a significant impact on the accuracy of model predictions” (Abstract) and “the setting of hyperparameters has a direct impact on the performance of the model, not only in terms of detection accuracy and training speed” (Introduction, pg. 2). Zhou additionally teaches that the optimized model improves object detection performance, stating that “our proposed method achieves the highest prediction accuracy in object detection” (Abstract). One of ordinary skill in the art would have recognized that the YOLOv5 detector of Kothala was a known neural-network based object detector ready for improvement, and that applying Zhou’s known genetic-algorithm optimization technique to the YOLOv5 hyperparameters would have predictably improved detection accuracy and training performance. This represents applying a known technique to a known device ready for improvement to yield predictable results.
Regarding Claim 15, Chilamkurthy in view of Kothala teaches the system according to claim 14, and additional limitations are met as in the consideration of claim 4 above.
Claims 6 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Chilamkurthy et. al in view of Azad et. al (“TransDeepLab: Convolution-Free Transformer-Based DeepLab v3+ for Medical Image Segmentation”).
Regarding Claim 6, Chilamkurthy teaches the method according to claim 5, but fails to teach that the image segmentation model comprises a TransDeepLab model.
However, Azad teaches this, stating that “this paper proposes TransDeepLab, a novel DeepLab-like pure Transformer for medical image segmentation” (Abstract).
Thus, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to substitute the UNet segmentation architecture of Chilamkurthy with the known TransDeepLab segmentation architecture of Azad in order to improve medical image segmentation performance, including improved capture of long-range contextual dependencies and multi-scale feature representations. Azad teaches that conventional CNN-based segmentation architectures suffer from limitations because “CNNs fail to capture long-range dependencies and spatial correlations due to the intrinsic property of confined receptive field size of convolution operations” (Abstract) and teaches that TransDeepLab addresses this issue through Transformer-based segmentation, stating that “Transformer, profiting from global information modeling that stems from the self-attention mechanism, has recently attained remarkable performance” (Abstract). One of ordinary skill in the art would therefore have been motivated to use the known TransDeepLab architecture in the hemorrhage segmentation framework of Chilamkurthy to obtain the predictable benefit of improved medical image segmentation accuracy and contextual feature extraction. This represents the simple substation of one known element for another, which would have yielded predictable results.
Regarding Claim 17, Chilamkurthy teaches the system according to claim 16, and additional limitations are met as in the consideration of claim 6 above.
Claims 7 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Chilamkurthy et. al in view of Karkkainen et. al (“Unsupervised Acute Intracranial Hemorrhage Segmentation with Mixture Models”).
Regarding Claim 7, Chilamkurthy teaches the method according to claim 1, but fails to teach that the image segmentation model comprises an untrained Expectation-Maximization algorithm.
However, Karkkainen teaches an unsupervised (i.e., not pre-trained using labeled training data) hemorrhage segmentation model utilizing an Expectation-Maximization algorithm, stating that “we propose a fully-unsupervised algorithm for acute intracranial hemorrhage segmentation based on the idea of Mixture Models… this representation allows us to find the optimal distribution parameters using the Expectation-Maximization algorithm” (Introduction, pg. 2) and “to determine the optimal parameter values, we use the Expectation Maximization (EM) algorithm” (C. Mixture Models, pg. 4).
Thus, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the segmentation approach of Chilamkurthy with the unsupervised Expectation-Maximization algorithm of Karkkainen in order to provide an alternative segmentation model that does not require extensive manually segmented training data. Karkkainen expresses the benefit of unsupervised segmentation techniques, stating that “publicly-available training data remains scarce due to privacy concerns…this problem can be alleviated by unsupervised algorithms” (Abstract), and “supervised algorithms can be inaccessible to most unless expert annotators are available…unsupervised algorithms help alleviate this problem and make the segmentation algorithms more accessible” (Introduction, pg. 2). One of ordinary skill in the art would have recognized that substituting the UNet segmentation model of Chilamkurthy with the unsupervised Expectation-Maximization algorithm of Karkkainen would have predictably yielded the known benefit of reducing dependence on manually segmented training datasets while still performing hemorrhage segmentation. This represents the simple substation of one known element for another and the use of a known technique to improve similar devices in the same way, which would have yielded predictable results.
Regarding Claim 18, Chilamkurthy teaches the system according to claim 12, and additional limitations are met as in the consideration of claim 7 above.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
The following references would also render a 102 rejection for claims 1-2, 5, 8-10, 11-13, 16, and 19-20:
Dong et. al (“An unsupervised domain adaptation brain CT segmentation method across image modalities and diseases”)
Kuo et. al (“Expert-Level Detection of Acute Intracranial Hemorrhage on Head Computed Tomography using Deep Learning”)
Ye et. al (“Precise diagnosis of intracranial hemorrhage and subtypes using a three-dimensional joint convolutional and recurrent neural network”)
Any inquiry concerning this communication or earlier communications from the examiner should be directed to WILLIAM ADU-JAMFI whose telephone number is (571)272-9298. The examiner can normally be reached M-T 8:00-6:00.
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, Andrew Bee can be reached at (571) 270-5183. 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.
/WILLIAM ADU-JAMFI/Examiner, Art Unit 2677
/ANDREW W BEE/Supervisory Patent Examiner, Art Unit 2677
Prosecution Insights