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
This action is in reply to the present application filed on March 10, 2025.
Claims 1-20 are currently pending and have been examined.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 03/10/2025 was filed before the mailing date of the first office action on the merits. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Priority
Acknowledgment is made of Applicant’s claim for priority under 35 U.S.C. § 371 of International Application No. PCT/EP2023/074219 filed on September 5, 2023, which claims the benefit of U.S. Provisional Application No. 63/406,453, filed on September 14, 2022.
Drawings
The drawings are objected to as failing to comply with 37 CFR 1.84(p)(4) because
reference character “263” has been used to designate both a display component (para 63) and a prediction generator (FIG. 2).
Reference character “260” has been used to designate both the medical records component (paras 64) and the memory (FIG. 2 and paras 53, 58-59).
Corrected drawing sheets in compliance with 37 CFR 1.121(d) are required in reply to the Office action to avoid abandonment of the application. Any amended replacement drawing sheet should include all of the figures appearing on the immediate prior version of the sheet, even if only one figure is being amended. Each drawing sheet submitted after the filing date of an application must be labeled in the top margin as either “Replacement Sheet” or “New Sheet” pursuant to 37 CFR 1.121(d). If the changes are not accepted by the examiner, the applicant will be notified and informed of any required corrective action in the next Office action. The objection to the drawings will not be held in abeyance.
Claim Rejections - 35 USC § 112(b)
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 4, 13 and 15 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 4 recites the limitation “wherein the extracted plurality of different defined ICU ventilation prediction features for the patient comprises one or more of [list]”. A “plurality” of features clearly indicates there are at least two features being extracted, however, claim 4 recites that the extracted features may be "one or more of", indicating that there may be an instance when there is only one extracted feature. Therefore, the claim is indefinite. Claim 13 recites substantially similar limitations to those in claim 4 and is rejected on the same basis as stated above.
Claim 15 recites the limitation "the second processor" in line 1. There is insufficient antecedent basis for this limitation in the claim.
Claim 15 recites the limitation “the extracted plurality of different health features” in line 2. There is insufficient antecedent basis for this limitation in the claim.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1 analysis:
Claims 1, 11 and 20 are directed to a method, a system and a manufacture respectively and therefore all fall into one of the four statutory categories. (Step 1: Yes, the claims fall into one of the four statutory categories).
Step 2A analysis - Prong one:
The substantially similar independent method, system, and computer readable media claims, taking claim 20 as exemplary, recite the following limitations: obtain, from an electronic medical records database, a plurality of records for a patient in an ICU covering at least a first time period; extract, from the obtained plurality of records, a plurality of different defined ICU ventilation prediction features for the patient; analyze the extracted plurality of different defined ICU ventilation prediction features using a trained ICU ventilation prediction model; and predict, by the trained ICU ventilation prediction model, (i) a likelihood of a need for a ventilation of the patient in the ICU; and (ii) a duration of the need for the ventilation of the patient in the ICU.
The examiner interprets the above bolded limitations as additional elements as further discussed below. The remaining un-bolded limitations above, as drafted, is a process that, under the broadest reasonable interpretation, covers performance of the limitation in the mind but for recitation of generic computer components. That is, other than reciting the electronic medical records database and using a trained model, nothing in the claim precludes the step from practically being performed in the mind. For example, but for the identified additional elements, this claim encompasses a person obtaining and reading patient data, analyzing the data, and determining the likelihood of a need for ventilation and duration of ventilation for the patient in the manner described in the identified abstract idea, supra. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea. (Step 2A – Prong 1: Yes, the claims are abstract).
Step 2A analysis - Prong two:
This judicial exception is not integrated into a practical application. In particular, the claims recite an electronic medical records database (claims 1, 11 and 20), using a trained ICU ventilation prediction model (claims 1, 11 and 20), a processor (claims 11 and 20), an intensive care unit (ICU) ventilation prediction system (claim 11), non-transitory computer-readable storage medium (claim 20) and a computer program comprising instructions (claim 20) which are recited at a high-level of generality (i.e., as a generic processor performing generic computer functions) such that it amounts to no more than mere instructions to apply the exceptions using a generic computer component. For example, the trained model is used to generally apply the abstract idea without placing any limits on how the trained model functions. The limitations only recite the outcome of “analyzing the extracted plurality of different defined ICU ventilation prediction features” by “predicting a likelihood” and “a duration” of ventilation of the patient in the ICU and do not include any details about how the “analyzing” and “predicting” steps are accomplished. See MPEP 2016.05(f). As a further example, Applicant’s specification explains that the processor merely performs the method identified as the abstract idea by reading computer instructions/programs, executing the program components to analyze extracted data, cause results to be displayed, etc. and can essentially be any suitable processor (see Applicant’s specification paras 54-55, 59, 66-67, 69, 96). The identified additional elements equate to saying “apply it.” MPEP 2106.04(d)(I) indicates that merely saying “apply it” or equivalent to the abstract idea cannot provide a practical application.
Accordingly, this/these additional element(s), when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because it/they does/do not impose any meaningful limits on practicing the abstract idea. Therefore, claims 1, 11 and 20 are directed to an abstract idea without practical application. (Step 2A – Prong 2: No, the additional claimed elements are not integrated into a practical application).
Step 2B analysis:
The claim does 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 an electronic medical records database, using a trained ICU ventilation prediction model, a processor, an intensive care unit (ICU) ventilation prediction system, non-transitory computer-readable storage medium and a computer program comprising instructions to perform the noted steps amounts to no more than mere instructions to apply the exception using a generic computer component. Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually. There is no indication that the combination of elements improves the functioning of a computer or improves any other technology. The collective functions appear to be implemented using conventional computer systemization. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept (“significantly more”).
For the role of a computer in a computer implemented invention to be deemed meaningful in the context of this analysis, it must involve more than performance of well-understood, routine, and conventional activities previously known to the industry. Further, the mere recitation of a generic computer cannot transform a patent ineligible abstract idea into a patent-eligible invention. See MPEP 2106.05(d).
Applicant’s specification discloses the following:
Applicant describes embodiments of the disclosure at a very high level to include the use of a wide variety of processors, memories, storage mediums, networks, input/output devices, etc. (see paras 53-70, 96-103). The invention may use “any suitable processor or collection of processors” (see para 96).
Generic computer components recited as performing generic computer functions that are well-understood, routine and conventional activities amount to no more than implementing the abstract idea with a computerized system.
In summary, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because mere instructions to apply an exception using a generic computer component cannot provide an inventive concept (“significantly more”). The claims do not provide an inventive concept significantly more than the abstract idea. Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. (Step 2B: No, the claims do not provide significantly more).
Dependent Claims 2-10 and 12-19 further define the abstract idea that is presented in independent Claims 1, 11 and 20 and are further grouped as a mental process and are abstract for the same reasons and basis as presented above. Further, claims 5-7 and 14-19 recite additional elements beyond the abstract idea. Claims 5 and 14 recite an ICU ventilation prediction system, a patient data management systems (PDMS) and a patient monitoring system. Claims 6 and 16 recite an electronic medical records database. Claim 7 recites a gradient-boosted decision tree model. Claims 15 and 18 recite a second processor. Claim 16 further recites training the ICU ventilation prediction model and storing the trained ICU ventilation prediction model. Claims 17 and 19 recite a user interface. This/these additional element(s) is/are recited at a high level of generality such that it amounts to no more than mere instructions to apply the exception using a generic computer component. For example, as noted above, the Applicant’s specification indicates the use of known processors and interfaces.
Accordingly, this/these additional element(s), when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claims do not recite additional elements that integrate the judicial exception into a practical application when considered both individually and as an ordered combination. Therefore, the dependent claims are also directed to an abstract idea.
Thus, claims 1-20 are rejected under 35 U.S.C. 101 as being directed to abstract ideas without significantly more.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1-6, 8, 11-15, 17 and 19-20 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Suresh et al. (Suresh, H., Hunt, N., Johnson, A., Celi, L. A., Szolovits, P., & Ghassemi, M. (2017, May 23). Clinical intervention prediction and understanding using Deep Networks. arXiv.org. https://doi.org/10.48550/arXiv.1705.08498).
Regarding Claim 1, Suresh discloses the following limitations:
A method for predicting a likelihood of ventilation of a patient in an intensive care unit (ICU), the method comprising: obtaining, from an electronic medical records database, a plurality of records for the patient in the ICU covering at least a first time period; (Suresh discloses integrating data from all available ICU sources (vitals, labs, notes, demographics) (the patient in the ICU) and focus on learning rich representations of this data to predict onset and weaning of multiple invasive interventions. Data from the Multiparameter Intelligent Monitoring in Intensive Care database is used, which contains over 58,000 hospital admissions from approximately 38,600 adults. For each patient, they extract 5 static variables, 29 time-varying variable, and all available de-identified clinical notes for each patient as timeseries across their entire stay (obtaining, from an electronic medical records database, a plurality of records for the patient in the ICU covering at least a first time period). Static variables were replicated across all timesteps for each patient. Vital and lab measurements are given timestamps that are rounded to the nearest hour (covering at least a first time period).– abstract; section 3.1 and 3.2)
extracting, from the obtained plurality of records, a plurality of different defined ICU ventilation prediction features for the patient; (Suresh discloses extracting many variables for data extraction and preprocessing. – section 3.2)
analyzing the extracted plurality of different defined ICU ventilation prediction features using a trained ICU ventilation prediction model; and (Suresh discloses using a model that is trained on general ICU populations (using a trained ICU ventilation prediction model). The model utilizes the collected variable/data (analyzing the extracted plurality of different defined ICU ventilation prediction features) and predicts the status of intervention in a prediction window after a gap time. – sections 2, 3.3; Figures 1, 2)
predicting, by the trained ICU ventilation prediction model, (i) a likelihood of a need for a ventilation of the patient in the ICU; and (ii) a duration of the need for the ventilation of the patient in the ICU. (Suresh discloses using the model (by the trained ICU ventilation prediction model) to predict ventilation, non-invasive ventilation or vasopressors using the model to classify the prediction window as one of four outcomes: 1) Onset, 2) Wean, 3) Staying on intervention, 4) Staying off intervention. A prediction window is an onset if there is a transition from a label of 0 to 1 for the patient during that window; weaning is the opposite: a transition from 1 to 0. A window is classified as ”stay on” if the label for the entire window is 1 or ”stay off” if 0 (predicting a likelihood of a need for a ventilation of the patient in the ICU and a duration of the need for the ventilation of the patient in the ICU). – sections 3.3, 3.4; Figures 1, 2; Table 1)
Regarding Claim 2, Suresh discloses all the limitations above and further discloses the following limitations:
The method of claim 1, further comprising: predicting, by the trained ICU ventilation prediction model, one or more of: (i) no ventilation; (ii) invasive ventilation only; (iii) non-invasive ventilation only; and (iv) both invasive and non-invasive ventilation. (Suresh discloses prediction of five intervention tasks: invasive ventilation (predicting both invasive), non-invasive ventilation (and non-invasive ventilation), vasopressors, colloid boluses, and crystalloid boluses. The model also classifies the prediction window as one of four possible outcomes, the fourth being staying off intervention (no ventilation). – abstract; section 3.4)
Regarding Claim 3, Suresh discloses all the limitations above and further discloses the following limitations:
The method of claim 1, further comprising: predicting, by the trained ICU ventilation prediction model, one or more of: (i) ventilation versus no ventilation; (ii) invasive ventilation versus no invasive ventilation; and (iii) non-invasive ventilation versus no non-invasive ventilation. (Suresh discloses predicting ventilation (predicting ventilation versus no ventilation; invasive ventilation versus no invasive ventilation), non-invasive ventilation (predicting non-invasive ventilation versus no non-invasive ventilation), or vasopressors, the model classifies the prediction window as one of four possible outcomes: 1) Onset, 2) Wean, 3) Staying on intervention, 4) Staying off intervention. A prediction window is an onset if there is a transition from a label of 0 to 1 for the patient during that window; weaning is the opposite: a transition from 1 to 0. A window is classified as ”stay on” if the label for the entire window is 1 or ”stay off” if 0. – section 3.4)
Regarding Claim 4, Suresh discloses all the limitations above and further discloses the following limitations:
wherein the extracted plurality of different defined ICU ventilation prediction features for the patient comprises one or more of BMI, age, gender, pre-ICU admission lead time, ventilation status, admitted with elective surgery status, mean blood pressure, systolic blood pressure, diastolic blood pressure, heart rate, respiratory rate, oxygen saturation, blood glucose, white blood cell count, blood sodium, blood potassium, blood creatinine, blood hemoglobin, blood albumin, blood lactate, arterial blood gas, admission diagnosis, and Total Glasgow Coma Scale. (Suresh discloses extracting static demographics for each patient (the extracted plurality of different defined ICU ventilation prediction features for the patient) such as age (age), gender (gender), as well as vitals and labs including diastolic blood pressure (diastolic blood pressure), glucose (blood glucose), systolic blood pressure (systolic blood pressure), mean blood pressure (mean blood pressure), heart rate (heart rate), white blood cell count (white blood cell count), sodium (blood sodium), respiratory rate (respiratory rate), oxygen saturation (oxygen saturation), blood potassium (blood potassium), blood creatinine (blood creatinine), hemoglobin (blood hemoglobin), and Glasgow coma scale total (Total Glasgow Coma Scale). – Sections 3.2, 3.3; Figure 1; Table 3)
Regarding Claim 5, Suresh discloses all the limitations above and further discloses the following limitations:
The method of claim 1, wherein the predicting is performed by an ICU ventilation prediction system that is a component of a patient data management systems (PDMS) or a patient monitoring system. (Suresh discloses predictions done in a forward-facing manner to enable “real-time” performance (is a component of a patient data management systems (PDMS)), and predictions are made with a six-hour gap time to support clinically actionable planning by using deep architectures such as LSTM and CNN models. – abstract)
Regarding Claim 6, Suresh discloses all the limitations above and further discloses the following limitations:
The method of claim 1, wherein obtaining, from an electronic medical records database, a plurality of records for each of a plurality of patients in an ICU comprises obtaining a plurality of records for a plurality of historical ICU patients. (Suresh discloses utilizing data from the MIMIC-III database (an electronic medical records database). We consider patients 15 and older who had ICU stays (comprises obtaining a plurality of records for a plurality of historical ICU patients) from 12 to 240 hours and consider each patient’s first ICU stay only. – section 3.1)
Regarding Claim 8, Suresh discloses all the limitations above and further discloses the following limitations:
The method of claim 1, wherein the trained ICU ventilation prediction model is configured to analyze the extracted plurality of different defined ICU ventilation prediction features and predict a likelihood of the need for the ventilation of the patient in the ICU when at least one of the plurality of different ICU ventilation prediction features are missing from the obtained plurality of records. (Suresh discloses comparing raw physiological data to physiological words, where the vitals data are categorized by first converting each value into a z-score based on the population mean and standard deviation for that variable, and then rounding this score to the nearest integer and capping it to be between -4 and 4 (analyze the extracted plurality of different defined ICU ventilation prediction features). Each z-score value then becomes its own column, which explicitly allows for a representation of missingness (e.g., all columns for a particular variable zeroes) that does not require imputation (Figure 7 in Appendix B) (Wu et al., 2016) (when at least one of the plurality of different ICU ventilation prediction features are missing from the obtained plurality of records). For example, Figure 7 further shows that a row containing all zeros for a given variable indicates that the value for that variable was missing at the timestep. – section 3.3; Figure 7)
Regarding Claim 11, Suresh discloses the following limitations:
An intensive care unit (ICU) ventilation prediction system configured to predict a likelihood of ventilation of a patient in an ICU, the system comprising: an electronic medical records database comprising a plurality of records for a plurality of patients; (Suresh discloses using machine learning models (An intensive care unit (ICU) ventilation prediction system) for the prediction of five intervention tasks: invasive ventilation, non-invasive ventilation (configured to predict a likelihood of ventilation of a patient in an ICU), vasopressors, colloid boluses, and crystalloid boluses. Data from the MIMIC-III database is used. MIMIC is publicly available and contains over 58,000 hospital admissions from approximately 38,600 adults (an electronic medical records database comprising a plurality of records for a plurality of patients). – abstract; section 3.1)
and a processor configured to: obtain, from the electronic medical records database, a plurality of records for the patient in the ICU covering at least a first time period; (Suresh discloses integrating data from all available ICU sources (vitals, labs, notes, demographics) (the patient in the ICU) and focus on learning rich representations of this data to predict onset and weaning of multiple invasive interventions. For each patient, they extract 5 static variables, 29 time-varying variable, and all available de-identified clinical notes for each patient as timeseries across their entire stay (obtain, from the electronic medical records database, a plurality of records for the patient in the ICU covering at least a first time period). Static variables were replicated across all timesteps for each patient. Vital and lab measurements are given timestamps that are rounded to the nearest hour (covering at least a first time period).– abstract; section 3.2)
extract, from the obtained plurality of records, a plurality of different defined ICU ventilation prediction features for the patient; (Suresh discloses extracting many variables for data extraction and preprocessing. – section 3.2)
analyze the extracted plurality of different defined ICU ventilation prediction features using a trained ICU ventilation prediction model; and (Suresh discloses using a model that is trained on general ICU populations (using a trained ICU ventilation prediction model). The model utilizes the collected variable/data (analyze the extracted plurality of different defined ICU ventilation prediction features) and predicts the status of intervention in a prediction window after a gap time. – sections 2, 3.3; Figures 1, 2)
predict, by the trained ICU ventilation prediction model, (i) a likelihood of a need for a ventilation of the patient in the ICU; and (ii) a duration of the need for the ventilation of the patient in the ICU. (Suresh discloses using the model (by the trained ICU ventilation prediction model) to predict ventilation, non-invasive ventilation or vasopressors using the model to classify the prediction window as one of four outcomes: 1) Onset, 2) Wean, 3) Staying on intervention, 4) Staying off intervention. A prediction window is an onset if there is a transition from a label of 0 to 1 for the patient during that window; weaning is the opposite: a transition from 1 to 0. A window is classified as ”stay on” if the label for the entire window is 1 or ”stay off” if 0 (predict (i) a likelihood of a need for a ventilation of the patient in the ICU and (ii) a duration of the need for the ventilation of the patient in the ICU). – sections 3.3, 3.4; Figures 1, 2; Table 1)
Regarding Claim 12, Suresh discloses all the limitations above and further discloses the following limitations:
The system of claim 11, wherein the processor is further configured to predict one or more of: (i) no ventilation; (ii) invasive ventilation only; (iii) non- invasive ventilation only; and (iv) both invasive and non-invasive ventilation; (Suresh discloses prediction of five intervention tasks: invasive ventilation (predicting both invasive), non-invasive ventilation (and non-invasive ventilation), vasopressors, colloid boluses, and crystalloid boluses. The model also classifies the prediction window as one of four possible outcomes, the fourth being staying off intervention (no ventilation). – abstract; section 3.4)
or one or more of: (i) ventilation versus no ventilation; (ii) invasive ventilation versus no invasive ventilation; and (iii) non-invasive ventilation versus no non-invasive ventilation. (Suresh discloses predicting ventilation (predict ventilation versus no ventilation; invasive ventilation versus no invasive ventilation), non-invasive ventilation (predicting non-invasive ventilation versus no non-invasive ventilation), or vasopressors, the model classifies the prediction window as one of four possible outcomes: 1) Onset, 2) Wean, 3) Staying on intervention, 4) Staying off intervention. A prediction window is an onset if there is a transition from a label of 0 to 1 for the patient during that window; weaning is the opposite: a transition from 1 to 0. A window is classified as ”stay on” if the label for the entire window is 1 or ”stay off” if 0. – section 3.4)
Regarding Claim 13, Suresh discloses all the limitations above and further discloses the following limitations:
The system of claim 11, wherein the extracted plurality of different defined ICU ventilation prediction features for the patient comprises one or more of BMI, age, gender, pre-ICU admission lead time, ventilation status, admitted with elective surgery status, mean blood pressure, systolic blood pressure, diastolic blood pressure, heart rate, respiratory rate, oxygen saturation, blood glucose, white blood cell count, blood sodium, blood potassium, blood creatinine, blood hemoglobin, blood albumin, blood lactate, arterial blood gas, admission diagnosis, or Total Glasgow Coma Scale. (Suresh discloses extracting static demographics for each patient (the extracted plurality of different defined ICU ventilation prediction features for the patient) such as age (age), gender (gender), as well as vitals and labs including diastolic blood pressure (diastolic blood pressure), glucose (blood glucose), systolic blood pressure (systolic blood pressure), mean blood pressure (mean blood pressure), heart rate (heart rate), white blood cell count (white blood cell count), sodium (blood sodium), respiratory rate (respiratory rate), oxygen saturation (oxygen saturation), blood potassium (blood potassium), blood creatinine (blood creatinine), hemoglobin (blood hemoglobin), and Glasgow coma scale total (Total Glasgow Coma Scale). – Sections 3.2, 3.3; Figure 1; Table 3)
Regarding Claim 14, Suresh discloses all the limitations above and further discloses the following limitations:
The system of claim 11, wherein the ICU ventilation prediction system is a component of a patient data management systems (PDMS) or a patient monitoring system. (Suresh discloses predictions done in a forward-facing manner to enable “real-time” performance (is a component of a patient data management systems (PDMS)), and predictions are made with a six-hour gap time to support clinically actionable planning by using deep architectures such as LSTM and CNN models. – abstract)
Regarding Claim 15, Suresh discloses all the limitations above and further discloses the following limitations:
The system of claim 17, wherein the second processor is further configured to bin at least one of the extracted plurality of different health features into bins prior to training the ICU ventilation prediction model, and wherein the bins comprise a bin comprising missing data. (Suresh discloses comparing raw physiological data to physiological words, where the vitals data are categorized by first converting each value into a z-score based on the population mean and standard deviation for that variable, and then rounding this score to the nearest integer and capping it to be between -4 and 4 (bin at least one of the extracted plurality of different health features into bins). Each z-score value then becomes its own column, which explicitly allows for a representation of missingness (e.g., all columns for a particular variable zeroes) that does not require imputation (Figure 7 in Appendix B) (Wu et al., 2016) (the bins comprise a bin comprising missing data). For example, Figure 7 further shows that a row containing all zeros for a given variable indicates that the value for that variable was missing at the timestep. – section 3.3; Figure 7)
Regarding Claim 17, Suresh discloses all the limitations above and further discloses the following limitations:
The method of claim 1, further comprising: presenting, via a user interface, at least one of (i) the predicted likelihood of the need for the ventilation of the patient in the ICU; and (ii) the predicted duration of the need for the ventilation of the patient in the ICU. (Suresh discloses that real-time prediction of clinical interventions remains a challenge within intensive care units (ICUs). This task is complicated by data sources that are noisy, sparse, heterogeneous and outcomes that are imbalanced. In this paper, we integrate data from all available ICU sources (vitals, labs, notes, demographics) and focus on learning rich representations of this data to predict onset and weaning of multiple invasive interventions. Our predictions are done in a forward-facing manner to enable “real-time” performance (presenting the predicted likelihood and the predicted duration), and predictions are made with a six hour gap time to support clinically actionable planning. – abstract)
Regarding Claim 19, Suresh discloses all the limitations above and further discloses the following limitations:
The system of claim 11, further comprising: a user interface configured to provide at least one of (i) the predicted likelihood of the need for the ventilation of the patient in the ICU; and (ii) the predicted duration of the need for the ventilation of the patient in the ICU. (Suresh discloses that real-time prediction of clinical interventions remains a challenge within intensive care units (ICUs). This task is complicated by data sources that are noisy, sparse, heterogeneous and outcomes that are imbalanced. In this paper, we integrate data from all available ICU sources (vitals, labs, notes, demographics) and focus on learning rich representations of this data to predict onset and weaning of multiple invasive interventions. Our predictions are done in a forward-facing manner to enable “real-time” performance (presenting the predicted likelihood and the predicted duration), and predictions are made with a six hour gap time to support clinically actionable planning. – abstract)
Regarding Claim 20, Suresh discloses the following limitations:
A non-transitory computer-readable storage medium having stored a computer program comprising instructions, which, when executed by at least one processor, cause the at least one processor to: obtain, from an electronic medical records database, a plurality of records for a patient in an ICU covering at least a first time period; (Suresh discloses integrating data from all available ICU sources (vitals, labs, notes, demographics) (the patient in the ICU) and focus on learning rich representations of this data to predict onset and weaning of multiple invasive interventions. For each patient, they extract 5 static variables, 29 time-varying variable, and all available de-identified clinical notes for each patient as timeseries across their entire stay (obtain, from an electronic medical records database, a plurality of records for the patient in the ICU covering at least a first time period). Static variables were replicated across all timesteps for each patient. Vital and lab measurements are given timestamps that are rounded to the nearest hour (covering at least a first time period).– abstract; section 3.2)
extract, from the obtained plurality of records, a plurality of different defined ICU ventilation prediction features for the patient; (Suresh discloses extracting many variables for data extraction and preprocessing. – section 3.2)
analyze the extracted plurality of different defined ICU ventilation prediction features using a trained ICU ventilation prediction model; and (Suresh discloses using a model that is trained on general ICU populations (using a trained ICU ventilation prediction model). The model utilizes the collected variable/data (analyze the extracted plurality of different defined ICU ventilation prediction features) and predicts the status of intervention in a prediction window after a gap time. – sections 2, 3.3; Figures 1, 2)
predict, by the trained ICU ventilation prediction model, (i) a likelihood of a need for a ventilation of the patient in the ICU; and (ii) a duration of the need for the ventilation of the patient in the ICU. (Suresh discloses using the model (by the trained ICU ventilation prediction model) to predict ventilation, non-invasive ventilation or vasopressors using the model to classify the prediction window as one of four outcomes: 1) Onset, 2) Wean, 3) Staying on intervention, 4) Staying off intervention. A prediction window is an onset if there is a transition from a label of 0 to 1 for the patient during that window; weaning is the opposite: a transition from 1 to 0. A window is classified as ”stay on” if the label for the entire window is 1 or ”stay off” if 0 (predict a likelihood of a need for a ventilation of the patient in the ICU and a duration of the need for the ventilation of the patient in the ICU). – sections 3.3, 3.4; Figures 1, 2; Table 1)
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.
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 7, 9-10, 16 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Suresh in view of Wang et al. (Wang H, et al. Early prediction of noninvasive ventilation failure after extubation: development and validation of a machine-learning model. BMC Pulm Med. 2022 Aug 8;22(1):304. doi: 10.1186/s12890-022-02096-7. PMID: 35941641; PMCID: PMC9358918.).
Regarding Claim 7, Suresh discloses all the limitations above, however, does not disclose the following limitations met by Wang:
The method of claim 1, wherein the ICU ventilation prediction model is a gradient-boosted decision tree model. (Wang teaches that ten different models were derived on the training set and compared with the CatBoost model, including a Gradient Boosting Decision Tree (GBDT) model (a gradient-boosted decision tree model). – page 4 statistical analysis)
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified the real-time predictions of clinical interventions as disclosed by Suresh to incorporate a gradient boosting decision tree model as taught by Wang in order so that patients might benefit from early triage and more intensive monitoring (see Wang abstract conclusions).
Regarding Claim 9, Suresh and Wang disclose all the limitations above and further disclose the following limitations:
The method of claim 16, further comprising binning at least one of the extracted plurality of different health features into bins prior to training the ICU ventilation prediction model, and wherein the bins comprise a bin comprising missing data. (Suresh discloses comparing raw physiological data to physiological words, where the vitals data are categorized by first converting each value into a z-score based on the population mean and standard deviation for that variable, and then rounding this score to the nearest integer and capping it to be between -4 and 4 (bin at least one of the extracted plurality of different health features into bins). Each z-score value then becomes its own column, which explicitly allows for a representation of missingness (e.g., all columns for a particular variable zeroes) that does not require imputation (Figure 7 in Appendix B) (Wu et al., 2016) (the bins comprise a bin comprising missing data). For example, Figure 7 further shows that a row containing all zeros for a given variable indicates that the value for that variable was missing at the timestep. – section 3.3; Figure 7)
Regarding Claim 10, Suresh and Wang disclose all the limitations above and further disclose the following limitations:
The method of claim 9, wherein the binned extracted plurality of different health features comprise one or more of blood albumin, blood lactate, arterial blood gas pH, arterial blood gas PaCO2, arterial blood gas PaO2, PF ratio, and Glasgow Coma Scale (GCS) score. (Suresh discloses that the extracted data may include lactate (blood lactate) and Glascow coma scale total (Glasgow Coma Scale (GCS) score). – Appendix A-Table 3)
Regarding Claim 16, Suresh and Wang disclose all the limitations above and further disclose the following limitations:
The method of claim 1, wherein the ICU ventilation prediction model is trained by:(i) obtaining, from an electronic medical records database, a plurality of historical records for each of a plurality of historical patients in an ICU, wherein at least one of the plurality of historical patients in the ICU did not receive any ventilation, and wherein at least one of the plurality of historical patients in the ICU received non-invasive ventilation for a known non-invasive ventilation period, and wherein at least one of the plurality of historical patients in the ICU received invasive ventilation for a known invasive ventilation period, (Suresh discloses using data from the MIMIC-III database (obtaining, from an electronic medical records database, a plurality of historical records for each of a plurality of historical patients in an ICU). MIMIC is publicly available and contains over 58,000 hospital admissions from approximately 38,600 adults. We consider patients 15 and older who had ICU stays from 12 to 240 hours (a known period) and consider each patient’s first ICU stay only. This yields 34,148 unique ICU stays. When predicting ventilation (received invasive ventilation), non-invasive ventilation (received non-invasive ventilation), or vasopressors, the model classifies the prediction window as one of four possible outcomes: 1) Onset, 2) Wean, 3) Staying on intervention, 4) Staying off intervention (did not receive any ventilation). A prediction window is an onset if there is a transition from a label of 0 to 1 for the patient during that window; weaning is the opposite: a transition from 1 to 0. A window is classified as ”stay on” if the label for the entire window is 1 or ”stay off” if 0. – sections 3.1, 3.4; Table 1)
(ii) extracting, from the obtained plurality of historical records, a plurality of different health features for each of the plurality of historical patients; (Suresh discloses extracting, for each patient, 5 static variables, 29 time-varying variables and all available, de-identified clinical notes. – section 3.2; Figure 1)
(iii) curating the extracted plurality of different health features to identify a plurality of different historical ICU ventilation prediction features; (Suresh discloses extracting and preprocessing the data from the MIMIC-III database (curating the extracted plurality of different health features) in order to compare and categorize the data (to identify a plurality of different historical ICU ventilation prediction features). – sections 3.2, 3.3; Figure 1)
(iv) training the ICU ventilation prediction model using the plurality of different historical ICU ventilation prediction features; and (Suresh discloses training the models on general ICU populations (training the ICU ventilation prediction model). Table 4 shows the data set statistics for the training dataset, testing dataset and the total. We use a train/validation/test split of 70/10/20 and stratify the splits based on outcome. – section 4.3; Table 4)
(v) storing the trained ICU ventilation prediction model. (Suresh discloses that the results are an interesting start to extracting interpretability from neural networks on patient data, and future work to expand this will enable these models to be adopted (storing the trained ICU ventilation prediction model) in real clinical settings. – section 6)
Suresh does not disclose the following limitations met by Wang:
and wherein at least one of the plurality of historical patients in the ICU received both invasive and non-invasive ventilation each for a known ventilation period; (Wang teaches determining early predictors of noninvasive ventilation (NIV) failure and constructing an accurate machine-learning model to identify patients at risks of NIV failure after extubation in intensive care units (ICUs) (at least one of the plurality of historical patients in the ICU). Patients who underwent NIV after extubation in the eICU Collaborative Research Database (eICU-CRD) were included. NIV failure was defined as need for invasive ventilatory support (reintubation or tracheotomy) or death after NIV initiation (received both invasive and non-invasive ventilation). – page 1 abstract; page 4 Statistical analysis)
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified the real-time predictions of clinical interventions as disclosed by Suresh to incorporate a machine learning model trained on data containing patients who experienced noninvasive ventilation failure as taught by Wang in order so that patients might benefit from early triage and more intensive monitoring (see Wang abstract conclusions).
Regarding Claim 18, Suresh and Wang disclose all the limitations above and further disclose the following limitations:
The system of claim 11, wherein a second processor is configured to train the ICU ventilation prediction model, the second processor is configured to: obtain, from an electronic medical records database, a plurality of historical records for each of a plurality of historical patients in an ICU, wherein at least one of the plurality of historical patients in the ICU did not receive any ventilation, and wherein at least one of the plurality of historical patients in the ICU received non-invasive ventilation for a known non-invasive ventilation period, and wherein at least one of the plurality of historical patients in the ICU received invasive ventilation for a known invasive ventilation period, (Suresh discloses using data from the MIMIC-III database (obtain, from an electronic medical records database, a plurality of historical records for each of a plurality of historical patients in an ICU). MIMIC is publicly available and contains over 58,000 hospital admissions from approximately 38,600 adults. We consider patients 15 and older who had ICU stays from 12 to 240 hours (a known period) and consider each patient’s first ICU stay only. This yields 34,148 unique ICU stays. When predicting ventilation (received invasive ventilation), non-invasive ventilation (received non-invasive ventilation), or vasopressors, the model classifies the prediction window as one of four possible outcomes: 1) Onset, 2) Wean, 3) Staying on intervention, 4) Staying off intervention (did not receive any ventilation). A prediction window is an onset if there is a transition from a label of 0 to 1 for the patient during that window; weaning is the opposite: a transition from 1 to 0. A window is classified as ”stay on” if the label for the entire window is 1 or ”stay off” if 0. – sections 3.1, 3.4; Table 1)
extract, from the obtained plurality of historical records, a plurality of different health features for each of the plurality of historical patients; (Suresh discloses extracting, for each patient, 5 static variables, 29 time-varying variables and all available, de-identified clinical notes. – section 3.2; Figure 1)
curate the extracted plurality of different health features to identify a plurality of different historical ICU ventilation prediction features; (Suresh discloses extracting and preprocessing the data from the MIMIC-III database (curate the extracted plurality of different health features) in order to compare and categorize the data (to identify a plurality of different historical ICU ventilation prediction features). – sections 3.2, 3.3; Figure 1)
train the ICU ventilation prediction model using the plurality of different historical ICU ventilation prediction features; (Suresh discloses training the models on general ICU populations (train the ICU ventilation prediction model). Table 4 shows the data set statistics for the training dataset, testing dataset and the total. We use a train/validation/test split of 70/10/20 and stratify the splits based on outcome. – section 4.3; Table 4)
and store the trained ICU ventilation prediction model. (Suresh discloses that the results are an interesting start to extracting interpretability from neural networks on patient data, and future work to expand this will enable these models to be adopted (store the trained ICU ventilation prediction model) in real clinical settings. – section 6)
Suresh does not disclose the following limitations met by Wang:
and wherein at least one of the plurality of historical patients in the ICU received both invasive and non-invasive ventilation each for a known ventilation period; (Wang teaches determining early predictors of noninvasive ventilation (NIV) failure and constructing an accurate machine-learning model to identify patients at risks of NIV failure after extubation in intensive care units (ICUs) (at least one of the plurality of historical patients in the ICU). Patients who underwent NIV after extubation in the eICU Collaborative Research Database (eICU-CRD) were included. NIV failure was defined as need for invasive ventilatory support (reintubation or tracheotomy) or death after NIV initiation (received both invasive and non-invasive ventilation). – page 1 abstract; page 4 Statistical analysis)
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified the real-time predictions of clinical interventions as disclosed by Suresh to incorporate a machine learning model trained on data containing patients who experienced noninvasive ventilation failure as taught by Wang in order so that patients might benefit from early triage and more intensive monitoring (see Wang abstract conclusions).
Relevant Prior Art of Record Not Currently Being Applied
The prior art made of record and not relied upon is considered pertinent to applicant's
disclosure.
Bendavid et al. (Bendavid I, et al. A novel machine learning model to predict respiratory failure and invasive mechanical ventilation in critically ill patients suffering from COVID-19. Sci Rep. 2022 Jun 22;12(1):10573. doi: 10.1038/s41598-022-14758-x. PMID: 35732690; PMCID: PMC9216294.) discloses a two-step model; the first step was to train a machine learning predictive model on a large dataset of non COVID-19 critically ill hypoxemic patients from the United States (MIMIC-III). The second step was to apply transfer learning and adapt the model to a smaller COVID-19 cohort. An XGBoost algorithm was trained on data from the MIMIC-III database to predict if a patient would require IMV within the next 6, 12, 18 or 24 h. Patients’ datasets were used to construct the model as time series of dynamic measurements and laboratory results obtained during the previous 6 h with additional static variables, applying a sliding time-window once every hour. See Bendavid abstract.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KIMBERLY VANDER WOUDE whose telephone number is (703)756-4684. The examiner can normally be reached M-F 9 AM-5 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, PETER H CHOI can be reached at (469) 295-9171. 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.
/K.E.V./Examiner, Art Unit 3681
/PETER H CHOI/Supervisory Patent Examiner, Art Unit 3681
