Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
DETAILED ACTION
In the amendments filed 22 January 2026:
Claim 6 is cancelled
Claims 1-3, 8, 10-15 are amended
Claim 1-5, 7-15 are pending
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 1-5, 7-15 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Claims 1,11-13 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1
The claims recite computer-implemented methods and systems, which are within a statutory category.
Step 2A1
The limitations of:
Claims 1, 11-13 (Claim 12 being representative):
determine an initial refractive power value for the intraocular lens to be inserted, [using] an initial training data volume consisting of tuples of previously measured ophthalmological biometry data of patients, associated postoperative refractive results values, generated by a previously inserted intraocular lens, and associated initial refractive power values of the previously inserted intraocular lens as ground truth data, wherein the previously measured ophthalmological biometry data and a postoperative target refraction value are used as input data;
assigning the postoperative refractive results value to the previously measured ophthalmological biometry data of the patient in order to form a new training data record;
determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume, wherein determining the importance indicator value comprises determining at least one element from the following group:
a specification regarding a nominal improvement in future refractive power value predictions;
and a specification regarding a refractive error that could have been avoided if the new training data record were known prior to the determination of the initial refractive power value, and retraining using the initial training data volume and the new training data record.
as drafted, is a process that, under the broadest reasonable interpretation, covers certain methods of organizing human activity (i.e., managing personal behavior including following rules or instructions) but for recitation of generic computer components. The claims encompass a series of rules or instructions for a person or persons to follow, with or without the aid of a computer, to determine the intraocular lens to be inserted in the manner described in the identified abstract idea, supra. The rules or instructions are the claimed steps of “measuring, determining, assigning, training” as indicated supra.
Other than reciting generic computer components (discussed infra), i.e., a system implemented by a processor (computer), the claimed invention amounts to managing personal behavior or interaction between people. If a claim limitation, under its broadest reasonable interpretation, covers managing personal behavior or interactions between people but for the recitation of generic computer components, then it falls within the “certain methods of organizing human activity” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
The claim further recites “retraining the machine learning system using the initial training data volume and the new training data record.” When given its broadest reasonable interpretation in light of the disclosure, the generation of training for a machine learning model to determine an initial refractive power value for an intraocular lens to be inserted represents the creation of mathematical interrelationships between data. As such, the generation of training data for the machine learning model represents a mathematical concept that is interpreted to be part of the identified abstract idea, supra. The types of identified abstract ideas are considered together as a single abstract idea for analysis purposes.
The Examiner notes that the retraining of a machine learning model is recited in the claim. The type of training utilized by the claimed invention is not described by the Applicant. As such the Examiner is required to analyze the retraining step given the broadest reasonable interpretation. The step(s) performed to retrain step(s) of the model/algorithm is/are considered to be part of the abstract idea because it/they fall(s) under data manipulations that humans perform (i.e., fitting a model to data) and thus are interpreted to be part of the abstraction--the rules or instructions that fall under Certain Methods of Organizing Human Activity. See, e.g., Recentive Analytics, Inc. v. Fox Corp., No. 2023-2437 at 12 (Fed. Cir. April 18, 2025) (finding that “[i]terative training using selected training material…are incident to the very nature of machine learning.”). As such, the training of the machine learning model represents a mathematical concept that is interpreted to be part of the identified abstract idea, supra. The types of identified abstract ideas are considered together as a single abstract idea for analysis purposes.
Step 2A2
This judicial exception is not integrated into a practical application. In particular, the claims recite the additional elements of a processor and memory that implements the identified abstract idea. The processor and memory are not described by the applicant and is recited at a high-level of generality (i.e., a generic server performing generic computer functions) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea.
The claim further recites the additional element of using a trained machine learning model to determine an initial refractive power value for an intraocular lens to be inserted. This represents mere instructions to implement the abstract idea on a generic computer. Implementing an abstract idea using a generic computer or components thereof does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. See, e.g., Recentive Analytics, Inc. v. Fox Corp., No. 2023-2437 at 10 (Fed. Cir. April 18, 2025) (finding that claims that do no more than apply established methods of machine learning to a new data environment are ineligible). Alternatively, or in addition, the implementation of the trained machine learning model to determine an initial refractive power value for an intraocular lens to be inserted merely confines the use of the abstract idea (i.e., the trained model) to a particular technological environment or field of use and thus fails to add an inventive concept to the claims.
Alternately, assuming that the measurement step(s) are not part of the abstract idea, these steps represent insignificant extra-solution data gathering activity. MPEP 2106.04(d)(I) indicates that extra-solution data gathering activity cannot provide a practical application. Accordingly, even in combination, these additional elements do not integrate the abstract idea into a practical application.
Step 2B
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements of using a processor and memory to perform the noted steps amounts to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept (“significantly more”).
As discussed above with respect to integration of the abstract idea into a practical application, the additional element of using the trained machine learning model to determine an initial refractive power value for an intraocular lens to be inserted was found to represent mere instructions to implement the abstract idea on a generic computer and/or confine the use of the abstract idea (i.e., the trained model) to a particular technological environment or field of use. This has been re-evaluated under the “significantly more” analysis and determined to be insufficient to provide significantly more. MPEP 2106.05(I) indicates that mere instructions to implement the abstract idea on a generic computer and/or confining the use of the abstract idea to a particular technological environment or field of use cannot provide significantly more. See also Recentive Analytics, Inc. v. Fox Corp., No. 2023-2437 at 17 (Fed. Cir. April 18, 2025) (finding that applying machine learning to an abstract idea does not transform a claim into something significantly more).
Also, as discussed above with respect to integration of the abstract idea into a practical application, the additional elements of (1) measuring a group of ophthalmological biometry data of a patient and (2) measuring a postoperative refractive results value were considered extra-solution activity. This has been re-evaluated under the “significantly more” analysis and determined to be well-understood, routine, conventional activity in the field. MPEP 2016.05(d)(II) indicates that measuring of data has been held by the courts to be well-understood, routine, conventional activity (citing Symantec, TLI Communications, OIP Techs., and buySAFE). The prior art of record indicates that measuring a group of ophthalmological biometry data of a patient is well-understood, routine, conventional activity (see US Publication No. 20050251254 at Para. 0109, “Clinically relevant biometry” and “Biometry and Formulas: Nailing the Outcome”). The prior art of record also indicates that measuring a postoperative refractive results value is well-understood, routine, conventional activity (see (see US Publication No. 20240090761 at Para. 0045, US Publication No. 20230284901 at Para. 0061 and US Publication No. 20200281461 at Para. 0003). Well-understood, routine, conventional activity cannot provide an inventive concept (“significantly more”). As such the claim is not patent eligible.
Claims 2-5, 7-10, 14-15 are similarly rejected because they either further define/narrow the abstract idea and/or do not further limit the claim to a practical application or provide as inventive concept such that the claims are subject matter eligible even when considered individually or as an ordered combination.
Claim(s) 2 merely describe(s) adding new training data temporarily, which further defines the abstract idea.
Claim(s) 3 merely describe(s) increasing importance indicator value, which further defines the abstract idea.
Claim(s) 4 merely describe(s) displaying data, which further defines the abstract idea.
Claim(s) 4 also includes the additional element of “graphical unit” which is analyzed the same as the “a processor” and does not provide a practical application or significantly more for the same reasons.
Claim(s) 5 merely describe(s) transmitting data, which further defines the abstract idea.
Claim(s) 5 also includes the additional element of “memory” which is analyzed the same as the “a processor” and does not provide a practical application or significantly more for the same reasons.
Claim(s) 7 merely describe(s) determining importance indicator value, which further defines the abstract idea.
Claim(s) 8 merely describe(s) increasing importance indicator value, which further defines the abstract idea.
Claim(s) 9 merely describe(s) determining importance indicator value, which further defines the abstract idea.
Claim(s) 10 merely describe(s) determining the number of postoperative refractive result measurements, which further defines the abstract idea.
Claim(s) 14-15 merely describe(s) increasing training data volume, which further defines the abstract idea.
Claim(s) 14-15 also includes the additional element of “computer or control units” which is analyzed the same as the “a processor” and does not provide a practical application or significantly more for the same reasons.
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.
The Examiner notes that the rejection will reference the translated documents (attached) corresponding to any foreign documents recited in the rejection.
Claims 1-5,7,9-10,12,14 is/are rejected under 35 U.S.C. 103(a) as being unpatentable over BURWINKEL et al (Foreign Publication WO-2021148517-A1) in view of Ladas et al (Foreign Publication JP-2022511115-A) in view of HAN et al (Foreign Publication CN-114239825-A) in view of ZHU et al (Foreign Publication CN-115101196-A).
Regarding Claim 1
BURWINKEL teaches a computer-implemented method for increasing a training data volume for a machine learning system for determining an initial refractive power value for an intraocular lens to be inserted, the method comprising:
measuring a group of ophthalmological biometry data of a patient [BURWINKEL at Page 8 Para 2 teaches according to an advantageous embodiment of the method, the ophthalmological data of an eye can have at least one from the group consisting of an axial length, an anterior chamber depth, a lens thickness, a posterior chamber depth, a corneal thickness, a corneal keratometry, a lens - Equatorial plane, white-to-white distance and a pupil size];
determining an initial refractive power value for the intraocular lens to be inserted, by a trained machine learning system [BURWINKEL at Page 3 Para 1 teaches according to one aspect of the present invention, a computer-implemented method for determining the refractive power for an intraocular lens to be inserted is presented. The method can include generating first training data for a machine learning system on the basis of a first physical model for a refractive power for an intraocular lens. The method can include training the machine learning system using the generated first training data to form a first learning model for determining refractive power and training the machine learning system trained with the first training data with clinical ophthalmological training data to form a second learning model for determining refractive power] … [ … ]
and retraining the machine learning system using the initial training data volume and the new training data record [BURWINKEL at Page 7 Para 7 teaches in addition, the speed advantage that arises when an already trained machine learning model is retrained using better or additional training data is used in each case. This can significantly shorten the overall training time, thereby significantly Computing power can be saved, and thus the existing computer capacities can be better used].
BURWINKEL does not teach [ … ] … which was trained by an initial training data volume consisting of tuples of previously measured ophthalmological biometry data of patients, associated postoperative refractive results values, generated by a previously inserted intraocular lens, and associated initial refractive power values of the previously inserted intraocular lens as ground truth data for determining a corresponding machine learning model, wherein the previously measured ophthalmological biometry data and a postoperative target refraction value are used as input data for the trained machine learning system;
measuring a postoperative refractive results value;
assigning the postoperative refractive results value to the previously measured ophthalmological biometry data of the patient in order to form a new training data record;
determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume, wherein determining the importance indicator value comprises determining at least one element from the following group:
a specification regarding a nominal improvement in the future refractive power value predictions by the machine learning system;
and a specification regarding a refractive error that could have been avoided if the new training data record were known prior to the determination of the initial refractive power value;
Ladas teaches [ … ] … which was trained by an initial training data volume consisting of tuples of previously measured ophthalmological biometry data of patients, associated postoperative refractive results values, generated by a previously inserted intraocular lens, and associated initial refractive power values of the previously inserted intraocular lens as ground truth data for determining a corresponding machine learning model [Ladas at Page 8 Para 3 teaches another technique for determining an intraocular lens involves the use of machine learning. For example, the neural network may be trained using the result of the previous operation (interpret to combine with ophthalmological biometry data of BURWINKEL); Ladas at Page 12 Para 5 teaches in block 1030, method 1000 also includes determining the estimation error of the equation using a deep learning machine trained for verified postoperative results, including postoperative refraction corresponding to intraocular lens power], wherein the previously measured ophthalmological biometry data and a postoperative target refraction value are used as input data for the trained machine learning system [Ladas at Page 12 Para 5 (interpret to combine with ophthalmological biometry data of BURWINKEL)];
measuring a postoperative refractive results value [Ladas at Page 14 Para 4 teaches at block 1240, method 1200 can include obtaining post-operative refraction of the eye from an automated refraction that is communicably coupled with a biometer];
assigning the postoperative refractive results value to the previously measured ophthalmological biometry data of the patient in order to form a new training data record [Ladas at Page 3 Para 4 teaches the method can include correlating at least two eye measurement parameters, intraocular lens power, and postoperative refraction as a training set];
It would have been prima facie obvious skill in the art, at the time of effective filing, to combine biometry data of BURWINKEL with the refractive results of Ladas with the motivation to the IOL calculation [Ladas at Page 14 Para 10].
BURWINKEL/Ladas do not teach determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume, wherein determining the importance indicator value comprises determining at least one element from the following group:
a specification regarding a nominal improvement in the future refractive power value predictions by the machine learning system;
and a specification regarding a refractive error that could have been avoided if the new training data record were known prior to the determination of the initial refractive power value;
HAN teaches determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume, wherein determining the importance indicator value comprises determining at least one element from the following group [HAN at Page 2 Para 15 teaches determining Euclidean distances of the first training data and a plurality of second training data in different dimensions (first training data interpreted as initial training data, second training data interpreted as new training data)]:
It would have been prima facie obvious skill in the art, at the time of effective filing, to combine the references of BURWINKEL/Ladas with the indicator of HAN with the motivation to improve the data quality [HAN at Page 11 Para 5].
BURWINKEL/Ladas/HAN do not teach a specification regarding a nominal improvement in the future refractive power value predictions by the machine learning system; and a specification regarding a refractive error that could have been avoided if the new training data record were known prior to the determination of the initial refractive power value;
ZHU teaches a specification regarding a nominal improvement in the future refractive power value predictions by the machine learning system; and a specification regarding a refractive error that could have been avoided if the new training data record were known prior to the determination of the initial refractive power value [ZHU at Page 2 Para 3 teaches one of the above patent documents CN202011480616.5 is an intraocular lens refractive power calculation system based on machine learning. Through the design of a prediction model, an input module, a calculation module, an additional module and an output module, the artificial intelligence active learning data characteristics are fully utilized, the error calculation capability is automatically optimized, the refractive power of the artificial lens required to be implanted in cataract operation is accurately calculated, the eyeball biological parameters of all dimensions are matched, and the eyeball prediction accuracy of the extreme eyeball biological parameters is higher (improved accuracy interpreted as specification regarding a nominal improvement)];
It would have been prima facie obvious skill in the art, at the time of effective filing, to combine the references of BURWINKEL/Ladas/Han with the specification of ZHU with the motivation to improve the satisfaction degree of patients [ZHU at Page 9 Para 1].
Regarding Claim 2
BURWINKEL/Ladas/Han/ZHU teach the method of claim 1,
BURWINKEL/Ladas/Han/ZHU further teach wherein determining the specification regarding a nominal improvement in the future refractive power value predictions by the machine learning system comprises temporarily adding the new training data record to the initial training data volume [BURWINKEL at Page 7 Para 7 teaches In addition, the speed advantage that arises when an already trained machine learning model is retrained using better or additional training data is used in each case] and determining a prediction accuracy of the machine learning system [Ladas at Page 11 Para 4 teaches the training set 930 can be considered as labeled data, as the training set 930 can include postoperative refraction that can be used to determine the accuracy or error of the formula].
Regarding Claim 3
BURWINKEL/Ladas/Han/ZHU teach the method of claim 1,
BURWINKEL/Ladas/Han/ZHU further teach further comprising increasing the importance indicator value if there are no training data within a previously defined radius around the new training data record [ZHU at Page 2 Para 15 teaches and selecting at least one second training data from the plurality of second training data according to the Euclidean distance, and adding the second training data into the first training data].
Regarding Claim 4
BURWINKEL/Ladas/Han/ZHU teach the method of claim 1,
BURWINKEL/Ladas/Han/ZHU further teach further comprising displaying, on a graphical unit, the importance indicator value and/or the specification about a nominal improvement and/or the specification about the refractive error [Ladas at Page 12 Para 9 teaches further, the device 1100 can include a display 1108 capable of displaying the user interface 1102 (interpret to combine with improvements of ZHU)].
Regarding Claim 5
BURWINKEL/Ladas/Han/ZHU teach the method of claim 1,
BURWINKEL/Ladas/Han/ZHU further teach further comprising transmitting the measured postoperative refractive results value to a training data memory [Ladas at Page 3 Para 4 teaches the method can include correlating at least two eye measurement parameters, intraocular lens power, and postoperative refraction as a training set].
Regarding Claim 7
BURWINKEL/Ladas/Han/ZHU teach the method of claim 1,
BURWINKEL/Ladas/Han/ZHU further teach wherein the importance indicator value is determined after a number of postoperative refractive results values for one type of inserted intraocular lenses, the number being determined in advance [ZHU at Page 7 Para 2 teaches as a preferred technical scheme, the database building module is as follows: and establishing a high-myopia cataract deep learning database which comprises 1852 real and complete high-myopia cataract clinical data and biometric parameters, and randomly dividing the database into a training set, a verification set and a test set; ZHU at Page 7 Para 3 teaches as a preferred technical solution, the module for screening learning features includes: a) basic learning features including patient age, eye axis, corneal steepness axis curvature and its axial direction, corneal flatness axis curvature and its axial direction, anterior chamber depth, lens thickness, white-to-white distance, intraocular lens power, intraocular lens a constant; b) a learning characteristic of the base transform, including a ratio of lens thickness to anterior chamber depth; c) learning characteristics of complex transformation, including predicted diopter and effective intraocular lens position calculated by theoretical formula].
Regarding Claim 9
BURWINKEL/Ladas/Han/ZHU teach the method of claim 1,
BURWINKEL/Ladas/Han/ZHU further teach wherein the importance indicator value is determined after collecting an amount of training data in the form of postoperative refractive results values for given ophthalmological biometry data and inserted intraocular lenses [Ladas at Page 13 Para 5 teaches in one aspect, the training set 1130 can include postoperative results received from the automatic refractor 1180], the amount being determined in advance [ZHU at Page 7 Para 2 teaches As a preferred technical scheme, the database building module is as follows: and establishing a high-myopia cataract deep learning database which comprises 1852 real and complete high-myopia cataract clinical data and biometric parameters, and randomly dividing the database into a training set, a verification set and a test set].
Regarding Claim 10
BURWINKEL/Ladas/Han/ZHU teach the method of claim 1,
BURWINKEL/Ladas/Han/ZHU further teach further comprising:
determining a number of measurements of postoperative refractive results value by one entity [Ladas at Page 10 Para 6 teaches FIG. 7 shows a graph 700 representing an exemplary eye dataset (exemplary eye dataset interpret to include a number of measurements of postoperative refractive results). The dataset may be collected from one or more surgeons; Ladas at Page 12 Para 9 teaches for example, in one aspect, validated results from multiple devices 1100 provide postoperative measurements to a database located in the network in addition to or instead of using postoperative measurements locally. can do.];
and transferring the number of measurements of the postoperative refractive value by the one entity in comparison with the number of comparable measurements by other entities [Ladas at Page 14 Para 4 teaches at block 1240, method 1200 can include obtaining post-operative refraction of the eye from an automated refraction that is communicably coupled with a biometer. In one aspect, for example, the automatic refraction device 1180 can obtain postoperative refraction of the eye. The automatic refraction device 1180 may be coupled to the biometer 1170 so that the same device is used to obtain at least two eye measurement parameters and postoperative refraction. Further, the measured values may be stored in a common patient data storage device 1160 (interpreted to be the same number of measurements for any surgeon)].
Regarding Claim 12
BURWINKEL teaches a system for increasing a training data volume for a machine learning system for determining a refractive power value for an intraocular lens to be inserted, the system comprising:
a processor and a memory which is operatively connected to the processor and which stores program code elements which, when executed, cause the processor to [BURWINKEL at Page 13 Para 2 (see Claim 11 for explanation)]:
measure a group of ophthalmological biometry data of a patient [BURWINKEL at Page 8 Para 2 (see Claim 1 for explanation)];
determine an initial refractive power value for the intraocular lens to be inserted, by a trained machine learning system [BURKINWEL at Page 3 Para 1 (see Claim 11 for explanation)]… [ … ]
and retraining the machine learning system using the initial training data volume and the new training data record [BURWINKEL at Page 7 Para 7 (see Claim 1 for explanation)].
BURWINKEL does not teach [ … ] … which was trained by an initial training data volume consisting of tuples of previously measured ophthalmological biometry data of patients, associated postoperative refractive results values, generated by a previously inserted intraocular lens, and associated initial refractive power values of the previously inserted intraocular lens as ground truth data for determining a corresponding machine learning model, wherein the previously measured ophthalmological biometry data and a postoperative target refraction value are used as input data for the trained machine learning system;
measuring a postoperative refractive results value;
assigning the postoperative refractive results value to the previously measured ophthalmological biometry data of the patient in order to form a new training data record;
determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume, wherein determining the importance indicator value comprises determining at least one element from the following group:
a specification regarding a nominal improvement in future refractive power value predictions by the machine learning system;
and a specification regarding a refractive error that could have been avoided if the new training data record were known prior to the determination of the initial refractive power value;
Ladas teaches [ … ] … which was trained by an initial training data volume consisting of tuples of previously measured ophthalmological biometry data of patients, associated postoperative refractive results values, generated by a previously inserted intraocular lens, and associated initial refractive power values of the previously inserted intraocular lens as ground truth data for determining a corresponding machine learning model [Ladas at Page 8 Para 3, Page 12 Para 5 (see Claim 1 for explanation)], wherein the previously measured ophthalmological biometry data and a postoperative target refraction value are used as input data for the trained machine learning system [Ladas at Page 12 Para 5 (see Claim 1 for explanation)];
measuring a postoperative refractive results value [Ladas at Page 14 Para 4 (see Claim 1 for explanation)];
assigning the postoperative refractive results value to the previously measured ophthalmological biometry data of the patient in order to form a new training data record [Ladas at Page 3 Para 4 (see Claim 1 for explanation)];
It would have been prima facie obvious skill in the art, at the time of effective filing, to combine biometry data of BURWINKEL with the refractive results of Ladas with the motivation to the IOL calculation [Ladas at Page 14 Para 10].
BURWINKEL/Ladas do not teach determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume, wherein determining the importance indicator value comprises determining at least one element from the following group:
a specification regarding a nominal improvement in future refractive power value predictions by the machine learning system; and a specification regarding a refractive error that could have been avoided if the new training data record were known prior to the determination of the initial refractive power value;
HAN teaches determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume, wherein determining the importance indicator value comprises determining at least one element from the following group [HAN at Page 2 Para 15 (see Claim 1 for explanation)]:
It would have been prima facie obvious skill in the art, at the time of effective filing, to combine the references of BURWINKEL/Ladas with the indicator of HAN with the motivation to improve the data quality [HAN at Page 11 Para 5].
BURWINKEL/Ladas/HAN do not teach a specification regarding a nominal improvement in future refractive power value predictions by the machine learning system; and a specification regarding a refractive error that could have been avoided if the new training data record were known prior to the determination of the initial refractive power value;
ZHU teaches a specification regarding a nominal improvement in future refractive power value predictions by the machine learning system; and a specification regarding a refractive error that could have been avoided if the new training data record were known prior to the determination of the initial refractive power value [ZHU at Page 2 Para 3 (see Claim 1 for explanation)];
It would have been prima facie obvious skill in the art, at the time of effective filing, to combine the references of BURWINKEL/Ladas/Han with the specification of ZHU with the motivation to improve the satisfaction degree of patients [ZHU at Page 9 Para 1].
Claim 8 rejected under 35 U.S.C. 103(a) as being unpatentable over BURWINKEL, Ladas, HAN, ZHU as applied to claim 1 above, and further in view of SEAN et al (Foreign Publication CN-113705720-A).
Regarding Claim 8
BURWINKEL/Ladas/Han/ZHU teach the method of claim 1,
BURWINKEL/Ladas/Han/ZHU do not teach further comprising increasing the importance indicator value for a selected type of an intraocular lens by a predefined factor,
wherein the predefined factor is increased depending on an amount of training data for the selected type of intraocular lens in comparison with an overall amount of training data,
or wherein the predefined factor is increased if the amount of training data for the selected type of intraocular lens is less than a threshold value determined in advance.
SEAN teaches further comprising increasing the importance indicator value for a selected type of an intraocular lens by a predefined factor, wherein the predefined factor is increased depending on an amount of training data for the selected type of intraocular lens in comparison with an overall amount of training data, or wherein the predefined factor is increased if the amount of training data for the selected type of intraocular lens is less than a threshold value determined in advance [SEAN at Page 3 Para 2 teaches generally, to avoid the problem that the probability of the rare group appearing in the training data is too low due to the small number of samples and is therefore raised to be ignored by the classifier, the user of the machine learning algorithm will raise the weight of the rare group, for example, all groups will be given the same weight to train the Bayes classifier (number of samples being too low interpreted as amount of training data below a threshold; interpret to combine with intraocular lens selection and importance indicator value of BURWINKEL and HAN)].
It would have been prima facie obvious skill in the art, at the time of effective filing, to combine the references of BURWINKEL, Ladas, Han, ZHU with the adjustment of SEAN with the motivation to improve the correctness of the machine learning classifier.
Claims 11, 13, 15 is/are rejected under 35 U.S.C. 103(a) as being unpatentable over BURWINKEL et al (Foreign Publication WO-2021148517-A1) in view of Ladas et al (Foreign Publication JP-2022511115-A) in view of HAN et al (Foreign Publication CN-114239825-A).
Regarding Claim 11
BURWINKEL teaches a computer-implemented method for increasing a training data volume for a machine learning system for determining a postoperative refractive results value of an intraocular lens to be inserted, the method comprising:
measuring a group of ophthalmological biometry data of a patient [BURWINKEL at Page 8 Para 2 (see Claim 1 for explanation)];
determining a postoperative refractive results value of the intraocular lens to be inserted, by a trained machine learning system [BURKINWEL at Page 3 Para 1 teaches according to one aspect of the present invention, a computer-implemented method for determining the refractive power for an intraocular lens to be inserted is presented. The method can include generating first training data for a machine learning system on the basis of a first physical model for a refractive power for an intraocular lens] … [ … ]
and retraining the machine learning system using the initial training data volume and the new training data record [BURWINKEL at Page 7 Para 7 (see Claim 1 for explanation)].
BURWINKEL does not teach [ … ] …. which was trained by an initial training data volume consisting of tuples of previously measured ophthalmological biometry data of patients, associated initial refractive power values of a previously inserted intraocular lens, and associated postoperative refractive results values, generated by the previously inserted intraocular lens, as ground truth data for determining a corresponding machine learning model, wherein the previously measured ophthalmological biometry data and an initial refractive power value of the intraocular lens to be inserted are used as input data for the trained machine learning system;
measuring the postoperative refractive results value;
assigning the postoperative refractive results value to the previously measured ophthalmological biometry data of the patient in order to form a new training data record;
determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume;
Ladas teaches [ … ] …. which was trained by an initial training data volume consisting of tuples of previously measured ophthalmological biometry data of patients, associated initial refractive power values of a previously inserted intraocular lens, and associated postoperative refractive results values, generated by the previously inserted intraocular lens, as ground truth data for determining a corresponding machine learning model [Ladas at Page 8 Para 3, Page 12 Para 5 (see Claim 1 for explanation)], wherein the previously measured ophthalmological biometry data and an initial refractive power value of the intraocular lens to be inserted are used as input data for the trained machine learning system [Ladas at Page 12 Para 5 (see Claim 1 for explanation)];
measuring the postoperative refractive results value [Ladas at Page 14 Para 4 (see Claim 1 for explanation)];
assigning the postoperative refractive results value to the previously measured ophthalmological biometry data of the patient in order to form a new training data record [Ladas at Page 3 Para 4 (see Claim 1 for explanation)];
It would have been prima facie obvious skill in the art, at the time of effective filing, to combine biometry data of BURWINKEL with the refractive results of Ladas with the motivation to the IOL calculation [Ladas at Page 14 Para 10].
BURWINKEL/Ladas do not teach determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume;
HAN teaches determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume [HAN at Page 2 Para 15 (see Claim 1 for explanation)];
It would have been prima facie obvious skill in the art, at the time of effective filing, to combine the references of BURWINKEL/Ladas with the indicator of HAN with the motivation to improve the data quality [HAN at Page 11 Para 5].
Regarding Claim 13
BURWINKEL teaches a system for increasing a training data volume for a machine learning system for determining a postoperative refractive results value of an intraocular lens to be inserted, the system comprising:
a processor and a memory which is operatively connected to the processor and which stores program code elements which, when executed, cause the processor to [BURWINKEL at Page 13 Para 2 (see Claim 11 for explanation)]:
measure a group of ophthalmological biometry data of a patient [BURWINKEL at Page 8 Para 2 (see Claim 1 for explanation)];
determine a postoperative refractive results value of the intraocular lens to be inserted, by a trained machine learning system [BURKINWEL at Page 3 Para 1 (see Claim 11 for explanation)] … [ … ]
and retraining the machine learning system using the initial training data volume and the new training data record [BURWINKEL at Page 7 Para 7 (see Claim 1 for explanation)].
BURWINKEL does not teach [ … ] … which was trained by an initial training data volume consisting of tuples of previously measured ophthalmological biometry data of patients, associated initial refractive power values of a previously inserted intraocular lens, and associated postoperative refractive results values, generated by the previously inserted intraocular lens, as ground truth data for determining a corresponding machine learning model, wherein the previously measured ophthalmological biometry data and an initial refractive power value of the intraocular lens to be inserted are used as input data for the trained machine learning system;
measuring the postoperative refractive results value;
assigning the postoperative refractive results value to the previously measured ophthalmological biometry data of the patient in order to form a new training data record;
determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume;
Ladas teaches [ … ] … which was trained by an initial training data volume consisting of tuples of previously measured ophthalmological biometry data of patients, associated initial refractive power values of a previously inserted intraocular lens, and associated postoperative refractive results values, generated by the previously inserted intraocular lens, as ground truth data for determining a corresponding machine learning model [Ladas at Page 8 Para 3, Page 12 Para 5 (see Claim 1 for explanation)], wherein the previously measured ophthalmological biometry data and an initial refractive power value of the intraocular lens to be inserted are used as input data for the trained machine learning system [Ladas at Page 12 Para 5 (see Claim 1 for explanation)];
measuring the postoperative refractive results value [Ladas at Page 14 Para 4 (see Claim 1 for explanation)];
assigning the postoperative refractive results value to the previously measured ophthalmological biometry data of the patient in order to form a new training data record [Ladas at Page 3 Para 4 (see Claim 1 for explanation)];
It would have been prima facie obvious skill in the art, at the time of effective filing, to combine biometry data of BURWINKEL with the refractive results of Ladas with the motivation to the IOL calculation [Ladas at Page 14 Para 10].
BURWINKEL/Ladas do not teach determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume;
HAN teaches determining an importance indicator value for the new training data record, with the importance indicator value being indicative of a distance value of the new training data record in comparison with the initial training data volume [HAN at Page 2 Para 15 (see Claim 1 for explanation)];
It would have been prima facie obvious skill in the art, at the time of effective filing, to combine the references of BURWINKEL/Ladas with the indicator of HAN with the motivation to improve the data quality [HAN at Page 11 Para 5].
Regarding Claim 14
BURWINKEL/Ladas/Han/ZHU further teach a non-transitory computer program product for increasing training data volume for a machine learning system for determining a refractive power value for an intraocular lens to be inserted, wherein the computer program product comprises program instructions stored thereon, the program instructions being executable by one or more computers or control units and prompting the one or more computers or control units to carry out the method as per claim 1 [BURWINKEL at Page 7 Para 3 teaches furthermore, embodiments may relate to a computer program product that is accessible from a computer-usable or computer-readable medium that has program code for use by, or in connection with, a computer or other instruction processing system. In the context of this description, a computer-usable or computer-readable medium can be any device suitable for storing, communicating, forwarding or transporting the program code].
Regarding Claim 15
BURWINKEL/Ladas/Han further teach a non-transitory computer program product for increasing training data volume for a machine learning system for determining a postoperative refractive results value of an intraocular lens to be inserted, wherein the computer program product comprises program instructions stored thereon, the program instructions being executable by one or more computers or control units and prompting the one or more computers or control units to carry out the method as per claim 11 [BURWINKEL at Page 7 Para 3 (see Claim 14 for explanation).
Response to Arguments
Rejection under 35 U.S.C. § 101
Regarding the non-statutory rejection of Claims 14-15, the Applicant has amended the claims to overcome the basis of the rejection.
Regarding the rejection of Claims 1-5, 7-15, the Examiner has considered the Applicant’s arguments; however the arguments are not persuasive. Any arguments inadvertently not addressed are unpersuasive for at least the following reasons. Applicant argues:
The amended limitation of “retraining the machine learning system using the initial training data volume and the new training data record”, is a practical application of the alleged abstract idea.
Cites Ex parte Desjardins
Cites Enfish v. Microsoft
The Present Application in at least [0004]-[0008] indicates that the performance of machine learning systems can be negatively impacted when the volume and/or quality of available training data is low. Thus, the Present Application discloses and claims methods and systems for increasing a training data volume for a machine learning system for determining an initial refractive power value for an intraocular lens to be inserted. Therefore, the claims are directed to an improvement to how the claimed machine learning system operates and, thus, as found in Exparte Desjardin, the claims recite a practical application of the alleged abstract idea and are therefore patent eligible.
Regarding (a), the Examiner respectfully disagrees. As recited in the Office Action, the retraining of the machine learning model does not provide a practical application. As will be discussed infra, the background of the Specification does not describe a technical problem nor is there any improvement to the computer within the meaning of Desjardens/Enfish.
Regarding (b, d), the Examiner respectfully disagrees. The Examiner respectfully submits that there is no improvement to the claimed machine learning as there is in Desjardins. As found by the Panel, the claimed “training strategy allows the model to preserve performance on earlier tasks even as it learns new ones, directly addressing the technical problem of 'catastrophic forgetting' in continual learning systems" represents “technical improvements over conventional systems by addressing challenges in continual learning and model efficiency by reducing storage requirements and preserving task performance across sequential training.” This analysis represents implementation of the practical application-“improvement” analysis of MPEP 2106.04(d)(I) to the facts before the Panel.
Applicant’s claims do not provide such an improvement. There is no indication in the cited portion of the Specification nor the remainder of the Specification that the claimed invention provides an improvement to the machine learning model. Improving the accuracy of a machine learning model by supplying it with specific data is not an improvement to how the model operates within the meaning of Desjardins (see quotations from Recentive, infra). Applicant has cited no improvements to the actual machine learning model itself. The machine learning model does not learn new tasks while protecting knowledge of previous tasks, use less storage capacity, or reduce system complexity. Potentially improving the accuracy of the machine learning model by retraining it with new data is not an improvement to the model itself.
The Examiner’s position is wholly supported by the CAFC decision in Recentive Analytics, Inc. v. Fox Corp., No. 2023-2437, (Fed. Cir., April 18, 2025). Recentive held that non-specifically claimed training (which encompasses retraining) of a machine learning algorithm is insufficient to provide a practical application or significantly more because it does not result in “improving the mathematical algorithm or making machine learning better.” Recentive a 12. The decision further held that “[i]terative training using selected training material…are incident to the very nature of machine learning” and thus does not provide for an improvement. Recentive at 12.
Further, to the extent that the Applicant argues that the Background of the Specification describes a technical problem that the claimed invention is solving, the Examiner disagrees. What is described in Paras. 0004-0008 is a data insufficiency/collection problem rather than a technical problem caused by either the computer or the machine learning itself. Not having enough data to provide for an accurate model is a problem with the data itself. A machine learning model trained on specific data, yet deficient, data will provide a result based on that data. A machine learning model trained again (retrained) on a greater volume of applicable data will provide a result based on that data. That the greater volume of data provides a “better prediction” is a result of the content/volume of the data. Hence what is described is a data insufficiency/collection problem.
Finally, circling back to the ‘machine learning improvement’ issue, the fact that the described problem is a data insufficiency/collection problem is also why there is no improvement to the machine learning model as in Desjardins. The machine learning algorithm will provide a solution based on whatever data it is provided with. Nothing in the claim changes (i.e., improves) how the solution for a particular set of training data is arrived at. The claim is just changing the inputted data. The actual internal functioning of the machine learning model, such as the internal parameters of Desjardins, are not affected at all. This indicates that an improvement is not present.
Regarding (c), the Examiner respectfully disagrees. The “improvement” analysis of Enfish v. Microsoft parallels that of Desjardens, discussed sura. Because the claimed invention does not provide an improvement within the meaning of Desjardens, it does not provide an improvement as in Enfish either. There is no improvement to the software and thus there is no improvement to the computer.
Rejection under 35 U.S.C. § 112
Regarding the indefiniteness rejection of Claims 10, the Applicant has amended the claims to overcome the basis of the rejection.
Rejection under 35 U.S.C. § 102/103
Regarding the rejection of Claims 1-5, 7-15, the Examiner has considered the Applicant’s arguments; however the arguments are not persuasive. Applicant argues:
The core feature of claims 1 and 11-13 - per-case "importance indicator value" tied to distance and counterfactual benefit - is not taught or suggested by the cited art. None of Burwinkel, Ladas, Han, and Zhu discloses or suggests this per-record "value-of-data" computation, which both quantifies out-of-distribution novelty (distance to the training distribution), and computes a concrete, case-specific improvement metric (nominal improvement) or counterfactual avoidable error attributable specifically to adding that single record.
Moreover, the art solves different problems; it does not teach, suggest or motivate the claimed solution. In fact, the cited art teaches away from the claimed inventions.
The asserted combination is hindsight and requires selective importation without a teaching to combine. The Office Action stitches together. (i) Ladas's collection of pre/post measurements and training set, (ii) Zhu's Euclidean distance (for selection/cleaning), and (iii) Han's general claims of "better accuracy," to allege the claimed "importance indicator value" with "distance" plus "nominal improvement/avoidable error specification." But there is no teaching/suggestion/motivation in Burwinkel, Ladas, Han, and Zhu to: compute a per-record "importance indicator value" as a function of its distance to the training data volume; run a counterfactual "what would have happened if this record had been available earlier" calculation; or communicate that per-record quantified benefit back to the contributor to drive an ongoing, closed-loop increase in training data volume.
Claim-specific gaps remain, even with OAl's combinations. "Importance indicator value ... indicative of a distance value ... in comparison with the initial training data volume" is not taught by Burwinkel-Han, and Zhu's use of Euclidean distance is for a different purpose (cleaning/selection), domain (real estate), and without the claimed per- record benefit specification. Determining at least one ... specification regarding a nominal improvement ... and/or a specification regarding a refractive error that could have been avoided if the new training data record were known prior to the determination ..." is absent from Burwinkel, Ladas, Han, and Zhu.
Neither Ladas's error-prediction network nor Han's general accuracy claims instruct per- record counterfactual improvement computations triggered upon ingesting a new record.
Technical effect and unexpected synergy. The Present Application's per-record importance indicator (distance-based novelty plus per- record nominal improvement/avoidable error spec) produces a concrete technical effect: a closed feedback loop that measurably increases training set coverage in underrepresented biometric regions and expedites continual retraining with the most impactful data. This closes a "data-network" gap not addressed in Burwinkel, Ladas, Han, and Zhu, and would not have been obvious from them.
Regarding (a), the Examiner respectfully disagrees. The importance indicator value being tied to counterfactual benefit not recited in the claims, nor is the term “counterfactual benefit” defined or recited in the claims. Furthermore, the distance value is not defined in the claims, and is therefore given its broadest reasonable interpretation.
Regarding (b), the Examiner respectfully disagrees. The art does not need to be solving the same problem to cover the claims. Algorithms, and especially machine learning models, have applications in numerous fields, and are not tied exclusively to the application.
Regarding (c), the Examiner respectfully disagrees. Hindsight is not a requirement to cover a limitation or claim. As long as a motivation to combine is provided, then the requirement
Regarding (d), the Examiner respectfully disagrees. The distance value is not defined in the claims. Therefore, given its broadest reasonable interpretation, the prior of Han at Page 2 Para 15 covers the limitation. The limitation of “and a specification regarding a refractive error that could have been avoided if the new training data record were known prior to the determination of the initial refractive power value” is not required because only one limitation of the two needs to be covered.
Regarding (e), the Examiner respectfully disagrees. The term “counterfactual improvement” is not recited in the claims.
Regarding (f), the Examiner respectfully disagrees. None of arguments provided in the statement is recited in the claims.
Conclusion
The prior art made of record and not relied upon in the present basis of rejection are noted in the attached PTO 892 and include:
Bor et al (US Publication No. 10888380) discloses systems and methods for intraocular lens selection.
KIM et al (US Publication No. 20210059756) discloses a method for determining a lens to be inserted into an eyeball during lens implant surgery
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JONATHAN C EDOUARD whose telephone number is (571)270-0107. The examiner can normally be reached M-F 730 - 430.
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, Robert Morgan can be reached on (571) 272 - 6773. 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.
/JONATHAN C EDOUARD/Examiner, Art Unit 3683
/JASON S TIEDEMAN/Primary Examiner, Art Unit 3683
Prosecution Insights