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 and Rejections
Claims 1, 5, 12, and 17 have been amended.
Claims 1-17, 18, and 20 are pending.
Claims 12-17, 19 and 20 are withdrawn. As such, claims 1-11 are being examined.
In light of the claim amendments, the Section 112(b) rejection has been withdrawn.
In light of the claim amendments, the Section 101 rejection has been withdrawn.
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.
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 1-5, 7, and 11 are rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Appl. Publ. No. 2018/0228465 A1 to Southard et al. (hereinafter referred to as “SOUTHARD”), U.S. Patent Appl. Publ. No. 2023/0298757 A1 to Golan et al. (hereinafter referred to as “GOLAN”), U.S. Patent Appl. Publ. No. 2023/0363621 A1 to Mino et al. (hereinafter referred to as “MINO”), and the translation of CN 105107067 B (hereinafter “TOP MEDICAL”).
SOUTHARD discloses a portable ultrasound imaging system (Figure 1) for “assisting the placement of a catheter within the body of a patient.” (Abstract and [0006]). The system includes a probe, a console, and a display “for depicting the image, target location depth, and the proximity information of the medical device to the target location.” (Abstract). Prior to inserting a catheter within a patient, the system can help determine a catheter with a suitable length. (Id).
With respect to claim 1, SOUTHARD discloses a vasculature assessment device. The ultrasound imaging system “enable a clinician to determine, prior to insertion of the medical device, how much of the device will be disposed within the vessel, thus enabling the clinician to choose a catheter with suitable length.” (Abstract). SOUTHARD discloses an imaging probe configured to acquire raw image data of a blood vessel of a patient. “A probe 40, containing one or more transducer elements in a head 44 thereof for emitting and receiving ultrasonic signals, is operably attached to the console 20 via a cable or other suitable interface.” ([0025]). NOTE: Applicant’s disclosure does not provide a definition for “raw image data.” Examiner is interpreting “receiving ultrasonic signals” after emitting ultrasonic signals as “acquiring raw image data.” SOUTHARD also discloses a device module having a console coupled with the imaging probe, the console including a processor and a memory having logic stored thereon that, when executed by the processor, performs operations. The ultrasound imaging system 10 includes a console 20. ([0025] and Figure 1). “The console 20 includes therein a motherboard 64 for governing system functionality and includes a processor or other general or special purpose computer, memory, storage locations, and other components for system operation.” ([0027]). Furthermore, claim 1 recites “a processor; [and] a memory including instructions capable of causing the processor….” (see also [0061]).
SOUTHARD discloses that the operations include receiving the raw image data from the imaging probe. The probe 40 receives ultrasonic signals ([0025]) and communicates them to a motherboard 64 of the console 20 for subsequent processing. (Figure 1). SOUTHARD also discloses determining blood vessel data from the raw image data. “[A] processor or other suitable component of the motherboard 64 of system 10 (FIG. 2) can execute one or more algorithms to automatically detect the presence of a vessel in the ultrasound image 262 captured by the system 10 during operation. These algorithms take advantage of the fact that blood vessels represent a rapid gradient change compared to surrounding tissue when viewed ultrasonically, due to the relative density difference between the two. Further, vessels are typically round and possess a relatively ultrasonically homogenous interior structure, which further assist algorithms in detecting vessels in an ultrasonic image.” ([0055]) (see also [0050] in which depth to the vessel is calculated.”)
SOUTHARD also discloses depicting the suggested catheter parameters on a display of the device module. “The system 10 [after processing] can then inform the clinician as to the catheter(s) having overall lengths sufficient to provide at least the desired length of catheter within the vessel. This information is displayed to the left of the ultrasound image 102 in a suggested catheter length field 208, as shown in FIG. 12B.” (emphasis added) ([0050]). See also [0039] in which, after selecting a type of procedure, the system depicts “the procedure-specific gauge icon field 104.”
While SOUTHARD teaches an ultrasound probe having transducer elements disposed on a bottom surface of the ultrasound probe (see, e.g., [0025] and Figures 1 and 3), SOUTHARD does not explicitly teach that the imaging probe includes a plurality of ultrasound transducers disposed in a linear array on a bottom surface of the ultrasound probe and a near infrared imaging module, wherein the raw image data is acquired using the linear array and the near infrared imaging module.
In the same field of endeavor, TOP MEDICAL teaches “an infrared guided ultrasonic venography puncture system….” (p.1, lines 14-15). The technical solution in TOP MEDICAL is solved by using a puncture system that includes “an infrared guiding unit and an ultrasound positioning unit.” (p.1, line 58). The infrared guiding unit has an infrared light source, an infrared image processing unit, and an image projection unit or an infrared image display unit. (p.1, lines 58-60). The ultrasonic positioning unit includes an ultrasound probe, an ultrasound image processing unit, and an ultrasound image display unit. Infrared light (730 nm to 990 nm) is directed onto the human body and the reflected light is collected. (p.2, lines 5-7). An infrared image may be projected onto the body. (p.2, lines 8-9).
The display ultrasound image allows medical staff to observe the puncture depth of the puncture needle to determine whether it is successful. (p.2, lines 22-23). The ultrasound probe can be “set as a linear array probe….” (p.2, lines 23-24).
At least one of the “advantages” of TOP MEDICAL’s system is that the medical staff can use both infrared image/display to identify where to puncture the skin and the ultrasound image to track the needle as it extends a depth into the tissue. (p.3, lines 1-19). According to TOP MEDICAL, the system can improve accuracy, save time, and reduce the workload of the staff. (p.3, lines 18-19).
It would have been obvious to one skilled in the art to incorporate the combined “infrared guiding” and “ultrasound positioning” system as taught in TOP MEDICAL, which includes a linear array probe and a near infrared imaging module. One would have been motivated to use the TOP MEDICAL system because its dual-modality can “improve accuracy, save time, and reduce the workload of the staff.” (p.3, lines 18-19). There would have been a reasonable expectation of success as TOP MEDICAL demonstrates that it can be used to provide vascular access to a blood vessel.
NOTE: As discussed above with respect to GOLAN, Applicant’s disclosure does not provide a definition for “raw image data.” Examiner is interpreting TOP MEDICAL’s “ultrasonic signals received by the ultrasonic sensors” (p.4, lines 23-24) and collecting “light signal” (p.2, line 5) as “raw image data.”
However, neither SOUTHARD nor TOP MEDICAL teaches generating suggested catheter parameters through application of a trained machine learning (ML) model to the blood vessel data, wherein the suggested catheter parameters pertain to insertion of a catheter within the blood vessel.
GOLAN, on the other hand, teaches a system and method for computer-aided decision guidance. (Abstract). The system in GOLAN helps physicians make fast and accurate decisions by “warning surgeons of obstacles that may cause delays during a procedure (e.g., recommending a point of entry for a catheter, highlighting vascular geometries and properties which may be difficult or impossible to navigate with certain catheters, etc.); preventing surgeons from having to try multiple devices to successfully perform the surgery; reducing the waste associated with incorrect device choice; reducing the number of secondary procedures needed to correct for a non-optimal first procedure; and/or perform any other functions.” (emphasis added) ([0018]). The system can also automatically select “a catheter type or size based on a set of machine learning models).” ([0020]).
GOLAN’s system/method can include “selecting (e.g., recommending, initiating, etc.) a medical device for use in the procedure.” ([0109]). For example, recommending the medical device can include “selecting any or all of: a catheter diameter (e.g., based on vessel diameter, based on smallest vessel diameter, based on vessel diameter immediately before the occlusion, etc.), a catheter length (e.g., based on path length, based on length of one or more vessels, etc.), a catheter material (e.g., catheter flexibility based on vessel tortuosity), a catheter type (e.g., twist end catheter, suction catheter, etc.), whether or not aspiration is involved in the procedure (e.g., based on calcification of clot), a device type (e.g., catheter, revascularization device, coil, braid, aspiration system, etc.), a determination of whether or not to perform a procedure (e.g., based on a size of a clot, based on a calcification of a clot, etc.), and/or any other features.” ([0110]).
To make these determinations, the computing system in GOLAN can incorporate “artificial intelligence (AI), such as a with any or all of: a set of machine learning models and/or algorithms, a set of deep learning models and/or algorithms (e.g., neural networks, convolutional neural networks, etc.), a set of mappings, a decision tree, and/or with any other tools.” ([0042]). GOLAN’s system/method can also include “training and/or re-training (e.g., updating) any or all of a set of models (e.g., based on an outcome of a procedure performed based on an output from the method) and/or any other processes.” ([0122]).
GOLAN teaches determining a set of parameters related to the patient’s anatomy that will then produce/trigger an output to the user. ([0089], [0094], and [0103]). For example, “[t]he set of parameters preferably includes one or more geometric features associated with the set of images, such as…geometric features associated with the anatomical region(s) associated with the set of images.” ([0094]. GOLAN repeatedly describes parameters related to vessels, including vessel diameter, ([0096]), and recommending a device (e.g., catheter) based on the diameters. (see, e.g., [0120], [0125]). Both steps of (a) determining the set of parameters and (b) producing an output can be performed by machine-learning models. ([0093], [0106]).
GOLAN also teaches the ML models may be applied by a remote computing system. “Additionally or alternatively, any or all of the information can be determined with artificial intelligence (AI), such as a with any or all of: a set of machine learning models and/or algorithms, a set of deep learning models and/or algorithms (e.g., neural networks, convolutional neural networks, etc.), a set of mappings, a decision tree, and/or with any other tools.” ([0042]).
It would have been obvious to one skilled in the art to modify the system/method of SOUTHARD by generating suggested catheter parameters through application of a trained machine learning (ML) model to the blood vessel data, as taught in GOLAN, wherein the suggested catheter parameters pertain to insertion of a catheter within the blood vessel. One would have been motivated to apply the trained ML model because a trained ML model can assist a physician in making fast and accurate decisions, as taught in GOLAN, including decisions of what catheter to use to access a vessel. ([0020]). There would have been a reasonable expectation of success because, as taught in GOLAN, trained ML models can be applied to data for selecting a medical device (e.g., catheter) for a procedure.
However, none of SOUTHARD, TOP MEDICAL, or GOLAN teach that the trained ML model is trained by a computing device on blood vessel data sets acquired from a plurality of vasculature assessment devices and catheter parameter data sets, wherein the blood vessel data is obtained by the computing device from an electronic medical records system through a network communication over a network server and the catheter parameter data sets are correlated individually with the blood vessel data sets in a one-to-one relationship.
NOTE: Examiner is interpreting the above claim limitations as product-by-process claim limitations. More specifically, claim 1 recites a vascular assessment device having a processor that generates suggested catheter parameters by applying the blood vessel data to a ML model. However, claim 1 then recites how the ML model was trained (i.e., by a computing device). Claim 1 also recites how the blood vessel data is obtained by the computing device (i.e., from an electronic medical records system through a network communication over a network server), again rendering the claim limitation a product-by-process limitation.
“[E]ven though product-by-process claims are limited by and defined by the process, determination of patentability is based on the product itself. The patentability of a product does not depend on its method of production. If the product in the product-by-process claim is the same as or obvious from a product of the prior art, the claim is unpatentable even though the prior product was made by a different process.” In re Thorpe, 777 F.2d 695, 698, 227 USPQ 964, 966 (Fed. Cir. 1985) (MPEP 2113).
Examiner notes that “[t]he structure implied by the process steps should be considered when assessing patentability….” (MPEP 2113). However, in this case, the process steps do not cause any structural features or characteristics. For example, the process steps do not change the memory of the device module or the logic stored thereon.
MINO teaches a system that receives images of an anatomical target and, using a trained machine-learning (ML) model, generates a surgical plan that includes a recommended tool. (Abstract). “The ML model(s) 612 may be trained using supervised learning …. Supervised learning uses prior knowledge (e.g., examples that correlate inputs to outputs or outcomes) to learn the relationships between the inputs and the outputs. The goal of supervised learning is to learn a function that, given some training data, best approximates the relationship between the training inputs and outputs so that the ML model can implement the same relationships when given inputs to generate the corresponding outputs.” (emphasis added) ([0086]).
In one example of a trained ML model, the model is “trained to establish a correspondence between (i) endoscopic images and/or other external images of biopsy sites from past endoscopic procedures (optionally along with other information) and (ii) tissue acquisition tools used in those past procedures, and the tool characteristics including their types, sizes, and operational parameters. The trained first ML model(s) can be used by the biopsy tool selection unit 614 in an inference phase to automatically determine, from an input image (or a sequence of images or a live video) of an anatomical target (optionally along with other information), a tissue acquisition tool recommendation including a recommend tool of a particular type and size and operational parameters for manipulating to tool to collect tissue from the anatomical target.” ([0092]). Embodiments taught in MINO “can improve cannulation and endoscope navigation accuracy and efficiency and procedure success rate, especially for inexperienced physicians.” ([0102]).
Notably, the ML model can be trained by the training module of the system or trained by a separate unit (i.e., by a computing device other than the device that is using the ML model). ([0090]). This is consistent with GOLAN that teaches analysis may be performed remotely by a system using ML models. (see, e.g., [0042]).
Regarding the claimed “one-to-one relationship,” MINO teaches that “supervised learning uses prior knowledge (e.g., examples that correlate inputs to outputs or outcomes) to learn the relationships between the inputs and the outputs.” (emphasis added) ([0086]). MINO lists several “commonly used supervised-ML algorithms.” ([0087]). At least some of these algorithms are exclusively associated with one-to-one relationships (i.e., paired input-output data). These include Logistic Regression (LR), Naive-Bayes, Random Forest (RF), and Support Vector Machines (SVM). (Id).
It would have been obvious to one having ordinary skill in the art to use an ML model that was trained, by a computing device, on blood vessel data sets acquired from a plurality vasculature assessment devices (i.e., images showing a diameter of a vessel in a past procedure and the vasculature assessment device used during the procedure similar to MINO) in which the catheter parameter data sets are correlated individually with the blood vessel data sets in a one-to-one relationship (i.e., supervised learning as taught in MINO). One would be motivated to use a trained ML model in order to make the best possible selection early on because “an early choice of a proper device can improve the safety and efficacy of the procedure, reduce time to intervention, reduce cost and waste, and/or can confer any other benefits” as taught in GOLAN and MINO (see, e.g., [0102] of MINO). There would have been a reasonable expectation of success as ML trained models can receive input anatomical images and recommend a type of tool to use during a medical procedure as taught in GOLAN and MINO.
“Once the examiner provides a rationale tending to show that the claimed product appears to be the same or similar to that of the prior art, although produced by a different process, the burden shifts to applicant to come forward with evidence establishing an nonobvious difference between the claimed product and the prior art product.” (MPEP 2113).
With respect to claim 2, SOUTHARD teaches the blood vessel data including a diameter of the blood vessel and a depth of the blood vessel with respect to a skin surface. SOUTHARD teaches blood vessel diameter at, e.g., [0056], and teaches depth of the blood vessel at, e.g., [0028].
With respect to claim 3, SOUTHARD teaches the suggested catheter parameters include a catheter size and a catheter length. NOTE: Based on Applicant’s disclosure (e.g., [0034] and [0039]), Examiner is interpreting “catheter size” as including a catheter’s gauge or diameter. SOUTHARD teaches catheter size at, e.g., [0039], and catheter length at, e.g., [0050].
With respect to claim 4, SOUTHARD teaches that the imaging probe is an ultrasound probe. See, e.g., Figures 1 and 3 and [0025].
With respect to claim 5, while SOUTHARD teaches using an ultrasound imaging probe in which a head of the ultrasound probe includes a plurality of transducers (see, e.g.,[0025] describing transducer elements in a head).
With respect to claim 7, SOUTHARD teaches that the operations further include: determining an image of the blood vessel from the raw image data; and depicting the image on the display. “In another embodiment, color-coded icons, each representing a cross sectional size of a corresponding catheter or other suitable device, are depicted on a display of the ultrasound imaging system, together with and in relation to an ultrasound image of the vessel to be accessed.” ([0023], see also [0055]).
With respect to claim 11, SOUTHARD teaches that the imaging probe, the device module, and the display are combined into a single unit. See, e.g., ultrasound imaging system 10 in Figure 1.
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Appl. Publ. No. 2018/0228465 A1 to Southard et al. (hereinafter referred to as “SOUTHARD”), U.S. Patent Appl. Publ. No. 2023/0298757 A1 to Golan et al. (hereinafter referred to as “GOLAN”), U.S. Patent Appl. Publ. No. 2023/0363621 A1 to Mino et al. (hereinafter referred to as “MINO”), and the translation of CN 105107067 B (hereinafter “TOP MEDICAL”) as applied to claim 1 above, and further in view of U.S. Patent Appl. Publ. No. 2009/0118612 A1 to Grunwald et al. (hereinafter referred to as “GRUNWALD”).
With respect to claim 6, none of the cited art explicitly teach that the imaging probe is configured to obtain the raw image data via A-mode ultrasound imaging in combination with near infrared (NIR) imaging.
In the same field of endeavor, GRUNWALD teaches a transcutaneous ultrasound vascular access guiding system. ([0012]). The system includes “a single element ultrasound device providing A-Mode imaging, Doppler and correlation-based blood velocity estimation; a processor to process and correlate ultrasound information.” (Id). The transcutaneous ultrasound system is able to emit and receive ultrasound signals (i.e., raw image data). ([0103]).
Figure 1 shows “an A-mode device that has a pencil or other shaped handheld device with the ultrasound device (i.e., disposable ultrasound transducer 502) at a distal tip… A needle 1/guide 11 or catheter 5/needle 1 combination may also be configured as part of the device [that] allows an operator to visualize the needle 1 as it punctures a blood vessel 6 of interest.” ([0122]). “The primary ultrasound modality is A-mode for visualizing the tissues on gray-scale with real time analysis; however the modality can also be switched manually or automatically to Doppler mode within the blood vessel lumen to confirm venous flow versus arterial flow based on velocity of blood flow and pulsatility pattern.” ([0124]).
GRUNWALD teaches using only “a single element imaging element comprising a body 13, shaped like a pen or a flashlight.” ([0127]). Despite its small size and having only a single element, GRUNWALD can acquire different types of ultrasound data. “The type of vascular access imaging may be free hand A-Mode obtained with the device. The imaging may be color A-mode imaging, whereby the colors indicate bidirectional blood flow velocities obtained using Doppler or cross-correlation calculations, or duplex A-Mode imaging mode, where the bidirectional Doppler spectral distribution (velocity distribution) is in a sample window.” ([0130]).
Accordingly, GRUNWALD teaches obtaining raw image data via A-mode ultrasound imaging technology.
It would have been obvious to one skilled in the art to modify the ultrasound imaging system to obtain the raw image data via A-mode ultrasound imaging technology in combination with near infrared (NIR) imaging technology. One would have been motivated to modify the system because GRUNWALD’s transcutaneous device has desirable qualities for a portable system. First, with only a single-element, the device can have a size similar to a pen or flashlight. Second, GRUNWALD’s device is not just capable of acquiring A-mode data but can be used “for color A-mode imaging, whereby the colors indicate bidirectional blood flow velocities obtained using Doppler or cross-correlation calculations, or duplex A-Mode imaging mode, where the bidirectional Doppler spectral distribution (velocity distribution) is in a sample window.” ([0130]). There would have been a reasonable expectation of success because TOP MEDICAL demonstrates that NIR imaging can be acquired with ultrasound imaging and GRUNWALD demonstrates that the ultrasound imaging can include A-mode ultrasound data.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Appl. Publ. No. 2018/0228465 A1 to Southard et al. (hereinafter referred to as “SOUTHARD”), U.S. Patent Appl. Publ. No. 2023/0298757 A1 to Golan et al. (hereinafter referred to as “GOLAN”), U.S. Patent Appl. Publ. No. 2023/0363621 A1 to Mino et al. (hereinafter referred to as “MINO”), and the translation of CN 105107067 B (hereinafter “TOP MEDICAL”) as applied to claim 1 above, and further in view of U.S. Patent Appl. Publ. No. 2021/0093383 A1 to Wang et al. (hereinafter referred to as “WANG”).
With respect to claim 8, none of the cited art explicitly teach that the suggested catheter parameters include a catheter model, the operations including choosing the catheter model from a list of catheter models stored in the memory.
WANG teaches a “system and method…[that]…can assist physicians in pre-operational planning (also referred to as ‘periprocedural planning’) of percutaneous procedures….” ([0017]). “Generally, the system and method described herein use advanced imaging and modeling strategies to accurately assess the location and size of various anatomical structures of interest and to determine or select an ideal or optimal type and size of medical device (e.g., catheter) to be used in the performance of a medical procedure.” (emphasis added) ([0017]). “In an illustrative embodiment, one or more user-inputtable or user-selectable fields (e.g., radio buttons, drop-down menus, text boxes for entering information, etc.) may be displayed to allow a user to provide certain information that the system 26 can use to narrow down the universe of models from which the selection is made.” (emphasis added) ([0086]).
It would have been obvious to one skilled in the art to modify the ultrasound imaging system to include, among the suggested catheter parameters, a catheter model that has been chosen from a list of catheter models stored in memory. One would have been motivated to include a chosen catheter model because, as taught in WANG, providing the user with a chosen catheter model can reduce the time necessary for deciding which catheter to use, thereby reducing the time to complete the medical procedure. There would have been a reasonable expectation of success because, as taught in WANG, a system can be configured to provide such information to the user.
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Appl. Publ. No. 2018/0228465 A1 to Southard et al. (hereinafter referred to as “SOUTHARD”), U.S. Patent Appl. Publ. No. 2023/0298757 A1 to Golan et al. (hereinafter referred to as “GOLAN”), U.S. Patent Appl. Publ. No. 2023/0363621 A1 to Mino et al. (hereinafter referred to as “MINO”), and the translation of CN 105107067 B (hereinafter “TOP MEDICAL”) as applied to claim 1 above, and further in view of International Publ. No. WO 2018/138343 A1 (hereinafter referred to as “FRESENIUS”).
With respect to claim 9, the cited art does not explicitly teach that the operations further include communicating the suggested catheter parameters to an electronic medical record (EMR) system for inclusion in an EMR for the patient. However, GOLAN teaches that the system may “provide one or more outputs to…a database associated with the healthcare facility (e.g., EMR, EHR, PACS, etc.).” (emphasis added) ([0046]).
In the same field of endeavor, FRESENIUS teaches an automated cannulation system. (Abstract). Prior to the system performing an automated cannulation, “the selection method is designed to access a data matrix stored in the data storage device 90 in which the necessary program parameters are linked as a function of patient data and treatment data.” ([0112] of translation). “The data matrix may contain information that for the treatment ‘hemodialysis’ the program parameters ‘vascular structure data’,…’type of cannula to be used during cannulation’ must be observed.” ([0113]). FRESENIUS teaches storing the patient data as well: “The vascular structure data generated by the vascular structure measurement device can be saved as patient data and are available as historical data for subsequent treatments, which can improve cannulation in a patient-specific manner.” ([0118]).
It would have been obvious to one skilled in the art to modify the teachings of GOLAN so that the outputs that are communicated to an electronic medical records (EMR) system include patient-specific data, as taught in FRESENIUS. One would have been motivated to store such data in order to better prepare for subsequent procedures requiring catheters. There would have been a reasonable expectation of success because, as taught in GOLAN and FRESENIUS, patient-specific data can be communicated to an EMR system.
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Appl. Publ. No. 2018/0228465 A1 to Southard et al. (hereinafter referred to as “SOUTHARD”), U.S. Patent Appl. Publ. No. 2023/0298757 A1 to Golan et al. (hereinafter referred to as “GOLAN”), U.S. Patent Appl. Publ. No. 2023/0363621 A1 to Mino et al. (hereinafter referred to as “MINO”), and the translation of CN 105107067 B (hereinafter “TOP MEDICAL”) as applied to claim 1 above, and further in view of U.S. Patent Appl. Publ. No. 2020/0027210 A1 to Haemel et al. (hereinafter referred to as “HAEMEL”).
With respect to claim 10, the cited art does not explicitly teach receiving the trained ML model from an external computing system and storing the trained ML model in the memory. However, SOUTHARD does teach that its system may be used in network computing environments and that data can be transferred through a network or other communications connection. ([0062], [0064]).
HAEMEL teaches “a virtualized computing platform for advanced computing, such as image inferencing and image processing in medical applications” including ultrasound. ([0020]). In HAEMEL, a “training system 104 may be used to perform training, deployment, and implementation of machine learning models (e.g., neural networks, object detection algorithms, computer vision algorithms, etc.) for use in the deployment system 106. The deployment system 106 may be configured to offload processing and compute resources among a distributed computing environment to reduce the infrastructure requirements at the facility 102.” ([0021]). The facilities include “medical facilities, hospitals, healthcare institutes, clinics, research or diagnostic labs, and/or the like.” (Id). “The machine learning models may be trained at the facility 102 using data 108 (such as imaging data) generated at the facility 102 (and stored on one or more picture archiving and communication system (PACS) servers at the facility 102).” (emphasis added) ([0022]).
In one scenario, “the facility 102 needs a machine learning model for use in performing one or more processing tasks for one or more applications in the deployment system 106, but the facility 102 may not currently have such a machine learning model.” (emphasis added) ([0025]). In this example, the facility may select a machine learning model from a model registry in which the machine learning model was previously trained at another facility. (Id). The model registry is in the cloud. ([0023]). “In some embodiments, the machine learning model may then be retrained, or updated, at any number of other facilities, and the retrained or updated model may be made available in the model registry 124.” ([0025]). One benefit of training locally is that patient privacy can be better protected. “[W]hen being trained on imaging data from a specific location, the training may take place at that location, or at least in a manner that protects the confidentiality of the imaging data or restricts the imaging data from being transferred off-premises (e.g., to comply with HIPAA regulations, privacy regulations, etc.).” (Id).
It would have been obvious to one skilled in the art to configure the ultrasound imaging system to receive the trained ML model from an external computing system and store the trained ML model in the memory. One would have been motivated to receive a pre-trained ML model and store the ML model locally because pre-trained ML models are configured to perform certain tasks and storing the ML model locally protects patient privacy, as taught by HAEMEL. ([0025]). There would have been a reasonable expectation of success as HAEMEL demonstrates that ML models can be received from remote locations and stored locally.
RESPONSE TO APPLICANT’S ARGUMENTS
Applicant's arguments filed February 17, 2026 have been fully considered but they are not persuasive. Applicant argues at the bottom of page 11 and top of page 12 that the distinction between the claimed device and the cited art is that the “’trained ML model is trained by a computing device” (i.e., different from the vasculature assessment device as any broadest reasonable interpretation must be consistent with the specification) or whether “the blood vessel data is obtained by the computing device from an electronic medical records (EMR) system through a network communication over a network server.’”
As explained above, Applicant’s amendments are product-by-process claim limitations that do not provide structural features to the vascular assessment device that are different than what is taught by the prior art. More specifically, the prior art teaches ML models being trained on blood vessel data sets. Whether the ML models are trained by the device module of the vascular assessment device or a different computing device does not provide a structural difference. The burden shifts to Applicant to provide evidence that there is a nonobvious difference between the claimed invention and the prior art. (MPEP 2113).
Notwithstanding the above, the prior art also teaches that ML models can be trained by a host system or trained by a different system other than the host system. (see, e.g., MINO at [0090] and GOLAN at [0042]).
Moreover, Examiner notes that it is known for mobile ultrasound devices to use remote artificial intelligence networks, some of which use the data obtained by the mobile ultrasound devices. (see SONKO and PAGOULATOS discussed briefly below).
Prior Art Made of Record
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Sonko, Momodou L., T. Campbell Arnold, and Ivan A. Kuznetsov. “Machine learning in point of care ultrasound.” POCUS journal 7.Kidney (2022): 78 (hereinafter “SONKO”). SONKO teaches using point-of-care ultrasound devices to obtain training data for ML model. “DL [deep learning] is uniquely suited for analysis of POCUS because it is able to generate high-level abstractions from a wide array of raw imaging data of varying quality.” (p.79, right column). Moreover, SONKO describes “an open-access web service (POCOVIDScreen) that deploys the predictive model, allowing clinicians to both perform predictions on ultrasound lung images and upload their captured images to add to the database.” (p.82, right column).
US 20170262982 A1 (hereinafter “PAGOULATOS”) teaches artificial intelligence training networks that are used by handheld or mobile ultrasound systems. (Abstract and [0023]). “The systems and methods provided herein may be particularly useful for ultrasound imaging performed by novice ultrasound technicians and/or for ultrasound imaging utilizing a handheld or mobile ultrasound imaging device which may be deployed in a non-traditional clinical setting. Utilizing artificial intelligence approaches, the systems and methods provided herein are capable of determining whether acquired ultrasound images accurately depict or represent a desired view of a patient's organ or other tissue, feature or region of interest in a patient.” ([0023]). Notably, the AI networks are not trained by the mobile ultrasound systems. “An artificial intelligence ultrasound image recognition training network (“AI training network”) may be a cloud-based or distributed computing network, and may be accessible to a large number of ultrasound imaging devices.” ([0025]).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JASON P GROSS whose telephone number is (571)272-1386. The examiner can normally be reached Monday-Friday 9:00-5:00CT.
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, Anne M. Kozak can be reached at (571) 270-5284. 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.
/JASON P GROSS/Examiner, Art Unit 3797
/SERKAN AKAR/Primary Examiner, Art Unit 3797
Prosecution Insights