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 .
Applicants filed response on 9/11/2025 has been received and entered.
Claims 1-37 have been canceled.
Claims 38-47 are pending, and claims 38-44 are withdrawn from further consideration.
Currently claims 45-47 are under examination.
The rejection on claims 45-47 under 35 USC 101 judicial exception is maintained and of record.
Applicants’ remarks on 35 USC 101 judicial exception are summarized below-
Step 1: The Claims Fall Within a Statutory Category (not in dispute)
The claims are directed to a "process" comprising a concrete sequence of diagnostic steps that assess type 2 diabetes risk based on multiple biomarkers-a statutory process under § 101. Accordingly, the claims satisfy the threshold inquiry for subject matter eligibility.
Step 2A, Prong One: The Claims Are Not Directed to a Judicial Exception (in dispute)
Applicants’ state “the technological improvement provided by the claimed invention addresses and overcomes significant inefficiencies and delays in accessing diabetes detection, prevention, and risk assessment for type 2 diabetes. Prior methods for assessing type 2 diabetes risk, such as the oral glucose tolerance test (OGTT) and fasting plasma glucose (FPG), detect abnormal glucose levels only after the disease has occurred or complications may have already begun.
Furthermore, glucose alone does not capture systemic disease effects, as diabetes is a systemic disease, assessing risk based solely on glucose is inherently limited, leaving substantial variability in future risk unaddressed. In current clinical practice, a potential diabetes diagnosis relies exclusively on glucose measurements, categorizing patients as either diabetic (high glucose) or non-diabetic (normal glucose). Yet among the "non-diabetic" population, individuals may have widely varying future risks of developing diabetes-risks that current techniques and biomarker measurements are unable to assess efficiently or accurately.
The claimed invention describes and provides a significant advancement in diabetes diagnostics by enabling multivariable, individualized risk assessment and facilitating treatment planning through risk-based scores for the assessment of preventive interventions. It provides longitudinal risk predictions over a defined period, such as less than 10 years, if the patient's condition is left untreated. This tangible, stepwise process improves diabetes diagnosis and prevention and cannot be performed mentally or with pen-and-paper analysis alone, distinguishing the claims from abstract ideas as recognized by Federal Circuit precedent. This process addresses the inefficiencies and delays inherent in current diabetes intervention approaches, particularly for individuals at risk or with a history indicating elevated risk, where existing methods are limited in providing timely, actionable guidance for preventive care. This aligns with claims upheld as non- abstract in Federal Circuit precedent as detailed below.”
Applicant’s arguments have been considered but are not persuasive.
The gist of the above argument stresses the novelty of the current invention which distinguishes the conventional methodology and/or technology of relying on glucose tolerance test (OGTT) and fasting plasma glucose (FPG) in detecting abnormal glucose level correlating to type 2 diabetes. Whereas the current invention employs a significant advancement in diagnosis by enabling multivariable, individualized risk assessment (which requires measuring follistatin, HbA1c, proinsulin and C-peptide via K-means clustering and recursive feature elimination, although applicants did not specify and mention in this portion of argument)(emphasis added). Applicants alleges that such improvement provides solution to the inefficiencies and delays inherent in the conventional approach.
Concerning the novelty, the criteria for determining judicial exception is distinct from those of prior art evaluations. True, the current invention uses the specific four biomarkers in combining two mathematical algorithms analysis, namely k-clustering and recursive feature elimination, in predicting risk of type 2 diabetes. The method may be novel and unobvious to one ordinary skilled in the art under different patent law statutory, i.e. 35 USC 102 and 35 USC 103. However, the test for judicial exception is not based on novelty and unobviousness. Rather the test is on 35 USC 101 law of nature, i.e. correlation and mathematical concepts abstract idea/mental process.
The claims are directed to a specific, concrete process, not a mental process.
Applicants argue that “The claims do not preempt mental diagnosis but instead require a biomarker-based multi- biomarker computational model system that automates and enhances diabetes risk prediction with features impossible in the mind, such as laboratory measurement of multiple blood analytes, algorithmic clustering and feature elimination to build a predictive model, and generation of a
quantified risk score and clinical report that enable early, actionable intervention. Like the self- referential database in Enfish, LLC v. Microsoft Corp., 822 F.3d 1327 (Fed. Cir. 2016), which was eligible as "a specific improvement to the way computers operate" the present claims likewise recite a specific improvement in medical diagnostics by transforming raw biomarker measurements into a clinically useful prediction that conventional diagnostic methods could not achieve.
The Federal Circuit emphasized eligibility for claims that provide a specific, concrete improvement in technology, illustrated in that case by faster searches and improved database efficiency-here, the claimed method is eligible because it addresses a real-world problem in medical diagnostics. It applies laboratory measurements of specific biomarkers and computational techniques to generate a quantified risk score and clinical report, enabling early intervention in type 2 diabetes. These steps extend beyond mental processes or abstract analysis, offering a practical and actionable improvement in healthcare.
Furthermore, the subject matter of the pending claims relate to rules-based processing analogous to McRO, Inc. v. Bandai Namco Games America Inc., 837 F.3d 1299 (Fed. Cir. 2016), which was held patent-eligible for using "limited rules... specifically designed to achieve an improved technological result." Here, the Applicant applies k-means clustering and recursive feature elimination to biomarker measurements to generate a quantified diabetes risk score and clinical report, providing a practical improvement in medical diagnostics by enabling earlier and more accurate identification of diabetes risk in various individuals than conventional approaches. These steps extend beyond mental or abstract processes, yielding a tangible and actionable result.
The claimed system offers a targeted solution to the challenge of early and accurate prediction of type 2 diabetes risk in individuals. This includes a quantified diabetes risk score and clinical report that can guide preventive intervention, which are not provided by existing conventional diagnostic methods or single-biomarker tests. Thus, the claimed system offers a practical and significant advancement in diabetes diagnosis, assessment, prediction, and treatment, enabling earlier and more accurate identification of risk and actionable clinical intervention. Like the cardiac device in CardioNet, LLC v. InfoBionic, Inc. 955 F.3d 1358 (Fed. Cir. 2020), eligible for improving arrhythmia detection via specific determinators analyzing beat variability, the claimed system advances diabetes assessment, prognosis, and subsequent preventive treatment by analyzing biomarker data (e.g., follistatin, HbAlc, proinsulin, C-peptide, or a combination thereof)
to generate clinically actionable outputs (e.g., a quantified diabetes risk score and early prediction report). In CardioNet, the Federal Circuit reversed an initial finding of ineligibility, noting the claims solved technical problems (e.g., inaccurate monitoring) with a device-centric solution. Here, the claimed system likewise addresses inefficiencies in diabetes diagnosis and treatment by requiring specific features-laboratory measurement of defined biomarkers and computational techniques (e.g., k-means clustering and recursive feature elimination)-to generate a quantified risk score and clinical report, thereby providing an actionable diagnostic tool.
Accordingly, Applicant respectfully asserts that the claims are not drawn to an abstract mental process but rather, and in sharp contrast, embody a specific biomarker-based computational method for early diabetes risk prediction that improves medical care by transforming laboratory measurements into quantified risk scores and clinical reports, enabling earlier, actionable intervention. Therefore, the claims are directed to patent-eligible subject matter under Step 2A, Prong One.”
Applicant’s arguments have been considered but are not persuasive.
First, applicants considered that the features of k-means clustering and recursive feature elimination are not in category of mental process and/or abstract idea because applicants’ usage provides an improved technological result for early and accurate prediction of type 2 diabetes risk in individual (cited Enfish, LLC v. Microsoft Corp., 822 F.3d 1327 (Fed. Cir. 2016), McRO, Inc. v. Bandai Namco Games America Inc., 837 F.3d 1299 (Fed. Cir. 2016), and CardioNet, LLC v. InfoBionic, Inc. 955 F.3d 1358 (Fed. Cir. 2020)). Examiner disagrees.
It is noted that combining K-means clustering and recursive feature elimination for more accurate and efficient predictor(s) in diseases are known and have been practiced in the field. For instance, Yousef et al. review using combined K-means clustering and recursive feature elimination approaches in identifying biomarkers in various diseases (J. Intelligent Learning Systems and Applications 2014 Vol. 6, page 153-161; See Abstract and Figure 4). Also Veredas discloses using K-means clustering and recursive feature elimination in identifying optimal classifiers “…to be effective in keeping the efficacy of the classifiers up and significantly reducing the number of necessary predictors” (Neurocomputing 2015 164:112-122; See Abstract). Shen teaches combing K-means clustering and recursive feature elimination in identifying more accurate biomarkers, i.e. HOXA9, KRTAP8-1, CCND1 and TULP2 for lung adenocarcinoma (Annals of Oncology 2018 29: Suppl. 8, page Viii33, Abstract No. 106P) and similarly Pradhan uses K-means clustering and recursive feature elimination to remove sample bias and increase efficiency in identifying more accurate tumor biomarkers (WO 2019220459; claims 7). It is noted that none of the above references discloses or suggests applying the K-means clustering and recursive feature elimination for detailed analysis of type 2 diabetes risk, particularly using the four recited biomarkers. Still, the mathematical models of combing K-means clustering and recursive feature elimination in identifying biomarkers for disease is not an inventive concept in the field.
Nevertheless K-means clustering and recursive feature elimination are in fact mathematical models known in the field statistical analysis (See above; emphasis added). These two are in fact mathematical concept, albeit no explicit formula or equations are recited in claim 45. While the claim directs to a “new” abstract idea (i.e. improving diagnosis of type 2 diabetes based on scores of the selected biomarkers) here, a mathematical calculation, is still an abstract (Synopsys, Inc. vs. Mentor Graphics Corp., 839 F.3d 1138, 1151 (Fed. Cir. 2016)). Claim 45 is directed to an abstract concept of collecting, analyzing, and displaying data as scores/reports while using K-means clustering and recursive feature elimination (mathematical analyzing).
Step 2A, Prong Two: The Claims Integrate Any Judicial Exception into a Practical Application.
Applicants argue that “[e]ven if viewed as reciting a judicial exception (which they do not), the claims integrate any such exception into a practical application and add significantly more. Applicant respectfully asserts that even if the claims were considered to recite a mental process (which they do not), the claims integrate any such process into a practical application under the 2019 PEG factors: (1) improving technology by applying specific computational techniques (k-means clustering and recursive feature elimination) to laboratory biomarker measurements; (2) effecting treatment by producing clinically actionable outputs such as a quantified diabetes risk score and early prediction report that enable preventive medical intervention; and (3) being tied to particular machines and laboratory instruments required for obtaining biomarker measurements and implementing the computational analysis.
For example, the method of claim 45 applies laboratory biomarker measurements (follistatin, HbAlc, proinsulin, and C-peptide) and computational techniques (k-means clustering and recursive feature elimination) to generate outputs such as a quantified diabetes risk score and early prediction report, enabling preventive intervention before disease onset-a "particular treatment" that improves medical outcomes by allowing earlier, more accurate identification of at- risk individuals than conventional single-biomarker testing. Applicant respectfully asserts USPTO Guidance (October 2019 Update, Example 42), that claims integrating specific data analyses to produce actionable, real-world diagnostic outputs are deemed eligible subject matter. Here, the method integrates biomarker and algorithmic data into clinically meaningful outputs (e.g., quantified scores and reports) that meaningfully apply the exception beyond mere mathematical concepts or general data analysis (see MPEP 2106.05(e)). Thus, the claimed features include meaningful limitations that integrate the judicial exception into a practical application for example, providing early, actionable risk stratification and clinical reporting unavailable in prior diagnostic resources.
Thus, the claimed outputs integrate the alleged abstract idea into the practical application of biomarker-based diabetes risk prediction and management. Indeed, the recited outputs represent a significant improvement for the patient and clinician, who receive a quantified diabetes risk score and early prediction report that indicate low, intermediate, or high risk of developing type 2 diabetes. These outputs enable timely identification of at-risk individuals and support preventive intervention, which are outcomes not achievable through conventional single-biomarker tests or generalized diagnostic methods.
The method of claim 45 requires the features of obtaining specific biomarker data (follistatin, HbAlc, proinsulin, and C-peptide) from a patient and applying defined computational techniques (k-means clustering and recursive feature elimination) to that data to provide specific outputs that result in a greatly improved system for early prediction of type 2 diabetes. Reasonably, the outputs provided by the method of claim 45 represent a significant advancement in diabetes diagnostics and care, as the patient and clinician are directly provided with a quantified diabetes risk score and an early prediction report indicating whether the individual has a low, intermediate, or high risk of developing type 2 diabetes. These outputs enable preventive medical intervention before disease onset, in sharp contrast to conventional single-biomarker tests (e.g., HbAlc alone) that do not offer sufficient predictive power. As described in the application, the presently claimed method eliminates guesswork in early diabetes assessment and provides for a significantly improved system of risk prediction and preventive treatment.”
Applicant’s arguments have been considered but are not persuasive.
The test for whether the claim has “practical application” adding more significant weight than law of nature is by (1) identifying whether there are any additional elements recited in the claim beyond the judicial exception, and (2) evaluating those additional elements individually and in combination to determine whether the claim as a whole integrates the exception into a practical application. See MPEP 2106.04(d).
Viewing claim 45 at issue, the active steps are (1) measuring blood levels of four biomarkers; (2) comparing and selecting biomarkers using K-means clustering and recursive feature elimination; (3) generating diabetic scores based on mathematical calculations from (2) and (4) providing diagnosis report.
As has been discussed above, step (2) and (3) using K-means clustering and recursive feature elimination falls into category of mathematical concept, and step (4) providing diagnosis report. Applicants consider the report of step (4) as actionable stratification and clinical management a greatly improved system for early prediction of type 2 diabetes. Applicants stress that such outputs enable preventive medical intervention before disease onset compared to the conventional biomarker test. Examiner disagrees.
Step (4) of claim 45 directs “providing an early prediction report of the human’s diabetes risk score to enable identification of high-risk individuals for potential intervention before disease onset.” Such step is merely advisable (see 2012 Mayo vs. Prometheus Supreme Court case where the claim “indicates a need to decrease” but not limiting doctors actually decrease (or increase) the dosage of 6-thioguanine)(emphasis added). Also under Vanda Pharmaceuticals vs. West-Ward Pharmaceutical, the Court held a treatment to the patient with an amount of a particular medication is an integrated practical application more than law of nature (887 F.3d at 1134-36, 126 USPQ2d at 1279-81). Taken together, at most the current invention provides merely “advice”, not actionable “treatment” to the identified subject. Thus no patentable weight of practical application is found in the claim.
Step 2B: Whether The Additional Elements Contribute an "Inventive Concept"
Applicants argue that “[f]irst, while measuring blood levels of certain biomarkers may have been known individually, the claims require a specific set of four biomarkers (follistatin, HbAlc, proinsulin, and C-peptide) that were not conventionally measured together for predictive purposes. At the time of the invention, conventional diagnostic methods primarily relied on HbAlc alone, which lacks sufficient predictive value for identifying pre-disease risk.
Second, the claims do not end with merely evaluating or reporting those measurements. Rather, they recite application of defined computational techniques-k-means clustering and recursive feature elimination-to select features and construct a biomarker model, which generates a quantified diabetes risk score. This ordered integration of laboratory data with algorithmic feature selection is not shown or suggested in the art relied upon by the Examiner, nor is it a conventional diagnostic practice.
Third, unlike in Mayo, the claims do not merely recite observing a natural correlation and instructing doctors to "apply it." Instead, the claims transform raw laboratory measurements into actionable clinical outputs: a quantified diabetes risk score and an early prediction report that stratifies individuals into varying risk score. These outputs enable preventive intervention that was not possible with routine or conventional methods. Courts have repeatedly found claims eligible where they integrate data analysis into a concrete improvement in diagnostics or treatment (see, e.g., CardioNet, LLC v. InfoBionic, Inc., 955 F.3d 1358 (Fed. Cir. 2020)).
Accordingly, the claims recite significantly more than a judicial exception, because the specific biomarker panel, coupled with clustering and recursive feature elimination to generate a clinically actionable output, constitutes an inventive concept that was not well-understood, routine, or conventional in the field.
Thus, Applicant respectfully asserts that claims 45-47 are not directed to a judicial exception. Rather, claims 45-47 recite multiple features that, when considered as an ordered combination, integrate any alleged abstract idea into a specific practical application, and impose meaningful limitations that provide a biomarker-based system for early prediction of type 2 diabetes with clear improvements over conventional diagnostic methods. These improvements include the measurement of a specific panel of biomarkers, the application of clustering and recursive feature elimination to construct a predictive model, and the generation of a quantified risk score and early prediction report that enable preventive intervention.
Applicant’s arguments have been considered but are not persuasive.
As to the first point, examiner already pointed out that the novelty itself is not the evaluation for judicial exception. Rather law of nature, natural phenomena, natural product, abstract idea and/or mental process are the tests for judicial exception.
As to the second point, the computation by K-means clustering and recursive feature elimination are in fact mathematical calculation concept. In addition, the combined use of these two models are known in the field as a more efficient approach in identifying suitable biomarkers in diseases, albeit not for type 2 diabetes specifically (see above cited prior art).
For the third point, the output report based on the risk score indeed is merely for doctor’s advice and lacks sufficient weight for “significant more” than law of nature (See Vanda Pharmaceuticals vs. West-Ward Pharmaceutical).
Therefore the current invention as a whole falls into an abstract idea under law of nature.
Conclusion
No claim is allowed.
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 nonprovisional extension fee (37 CFR 1.17(a)) 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 CHANGHWA J CHEU whose telephone number is (571)272-0814. The examiner can normally be reached 8 am to 8 pm.
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, Gregory Emch can be reached at 5712728149. 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.
CHANGHWA J. CHEU
Primary Examiner
Art Unit 1678
/CHANGHWA J CHEU/Primary Examiner, Art Unit 1678