Machine-learning enabled clinical decision support system for cognitive and movement disorders

Abstract

Machine-learning enabled clinical decision support system for cognitive and movement disorders are provided, enabling clinical next-step recommendations for brain health outcomes. A method comprises receiving a result of one or more multimodal assessments from a user interaction with a computing device; extracting one or more cognitive features from the result; applying a machine learning model to the one or more cognitive features to generate a composite score for the user; normalizing the composite score by age-category values; and predicting a cognitive impairment status based on the normalized composite score.

Description

MACHINE-LEARNING ENABLED CLINICAL DECISION SUPPORT SYSTEM FORCOGNITIVE AND MOVEMENT DISORDERSCROSS-REFERENCE TO RELATED APPLICATION(S)

[0001] This application claims the benefit of and priority to U.S. Provisional Application No. 63 / 569,885, filed March 26, 2024, which is hereby incorporated by reference in its entirety.BACKGROUND OF THE DISCLOSURE

[0002] Embodiments of the present disclosure relate to clinical next-step recommendations for brain health outcomes, and more specifically, to machine-learning clinical decision support systems for cognitive and movement disorders.BRIEF SUMMARY

[0003] In brain health, providers either know the clinical next-step recommendations in patient care or must look up next-step recommendations related to brain health outcomes. This may result in missing important clinical decisions impacting patient care, in cases where providers lack complete information. Further, many of the recommendations that could be made by providers are missed by the lack of sensitivity of the subjective observations of caregivers, often derived from paper and pencil tests. A more sensitive solution is required.

[0004] There is a need for a digital solution to automatically populate an all-inclusive results and person-specific next-step brain health recommendations for providers and patients.Embodiments of the present disclosure close the information gap by using information obtained from a digital health platform to inform potential directions in patient care. Using information from different subjective metrics (for example, patient-reported outcomes, questionnaires and / or surveys) and objective assessments (for example, screening cognitive, mental, neurologic and / or physical function, and available data from laboratory tests and electronic health records), embodiments of the present disclosure provide a patient result specific comprehensive clinical impression using the output of various modalities of the screened user’s cognitive performance. In addition to the overall impression, the platform generates next-step, patient-specific,actionable directions for the provider and patient specific to the patient’s brain health and cognitive assessment results.BRIEF DESCRIPTION OF THE DRAWINGS

[0005] The accompanying drawings, which are incorporated into and constitute a part of this specification, illustrate various exemplary embodiments and together with the description, serve to explain the principles of the disclosed embodiments.

[0006] FIG. 1 is schematic illustrating a clinical appropriateness score per indication risk pathway, in accordance with one or more embodiments of this disclosure.

[0007] FIG. 2 is schematic illustrating a clinical appropriateness score in accordance with one or more embodiments of this disclosure.

[0008] FIG. 3 is an exemplary clinical decision support pathway for clinical next steps, in accordance with one or more embodiments of this disclosure.

[0009] FIGs. 4A-4C illustrate an exemplary embodiment of the CDS solution for a digital trail making test, in accordance with one or more embodiments of this disclosure.

[0010] FIGs. 5A-5B is an exemplary anti-amyloid pathway implemented in the CDS, in accordance with one or more embodiments of this disclosure.

[0011] FIG. 6 is an exemplary computing node.DETAILED DESCRIPTION

[0012] Reference will now be made in detail to the exemplary embodiments of the present disclosure, examples of which are illustrated in the accompanying drawings. Wherever possible, the same reference numbers will be used throughout the drawings to refer to the same or like parts.

[0013] The systems, devices, and methods disclosed herein are described in detail by way of examples and with reference to the figures. The examples discussed herein are examples only and are provided to assist in the explanation of the apparatuses, devices, systems, and methods described herein. None of the features or components shown in the drawings or discussed below should be taken as mandatory for any specific implementation of any of these devices, systems, or methods unless specifically designated as mandatory.

[0014] Also, for any methods described, regardless of whether the method is described in conjunction with a flow diagram, it should be understood that unless otherwise specified or required by context, any explicit or implicit ordering of steps performed in the execution of a method does not imply that those steps must be performed in the order presented but instead may be performed in a different order or in parallel.

[0015] As used herein, the term “exemplary” is used in the sense of “example,” rather than “ideal.” Moreover, the terms “a” and “an” herein do not denote a limitation of quantity, but rather denote the presence of one or more of the referenced items.

[0016] Methods for assessing cognition and movement diseases traditionally require at least one cognition and / or fine motoric activity parameter as input to reference from an existing dataset of activity measurements obtained from a subject using a mobile device. The dataset may be transmitted from the mobile device to an evaluating device, such as a computer, or may be processed in the mobile device in order to derive the said at least one parameter from the dataset.

[0017] Second, the determined parameter is compared to a reference by, e.g., using a computer-implemented comparison algorithm carried out by the data processor of the mobile device or by the evaluating device, e.g., the computer. The result of the comparison is assessed with respect to the reference used in the comparison and based on the said assessment the subject will be assessed with respect to the cognition and movement disease or disorder.

[0018] Third, the said assessment, e.g., the identification of the subject as being a subject suffering from the cognition and movement disease or disorder, or not, is indicated to the subject or other person, such as a medical practitioner.

[0019] Alternatively, a recommendation for a therapy, such as a drug treatment, or for a certain lifestyle, e.g., a certain nutritional diet, is provided automatically to the subject or other person. To this end, the established assessment is compared to recommendations allocated to different assessments in a database. Once the established assessment matches one of the stored and allocated assessments, a suitable recommendation can be identified due to the allocation of the recommendation to the stored assessment matching the established assessment. Typical recommendations involve therapeutic measures as described elsewhere herein.

[0020] These clinically informed next steps are delivered as part of the Clinical Decision Support (CDS) feature of digital cognitive assessments provided on a healthcare platform for healthcare delivery (HCD). An integrated and unified CDS can contain discreterecommendations triggered by objective and subjective measures and consolidated using a large language model (LLM) to avoid redundancies in similar recommendations generated by interrelated measures. These assessments include, but are not limited to the digital clock drawing test (DCTclock™), Immediate Recall, Delayed Recall (together being the Digital Clock and Recall, DCR™), Digital Trail Making Test-parts A and B (dTMT-A and B), Phonemic Fluency, Semantic Fluency, Backwards Digit Span Test, Philadelphia Repeatable Verbal Learning Task (PVLT), and digital Speech Hearing Screener (dSHS). Additional assessments include digital tools and questionnaires to assess personally defined treatment benefits, lifestyle, mood, functional dependence, frailty, fall risk, gait and balance, electronic medical records data, etc.

[0021] Embodiments of the present disclosure differ from those currently available in that other solutions provide a generic, non-customized set of recommendations that includes education in areas not specific to the patient’s identified areas of concern based on cognitive testing. Rather than solely identifying possible therapies for certain disorders, embodiments of the present disclosure provide a pathway to guide providers in further diagnosis of cognitive and motor impairments by identifying potential causes of transient impairments (e.g. sleep disorders, polypharmacy, medical history, etc.). Additionally, these embodiments guide diagnostic next steps; for example, suggesting subsequent assessments (e.g., neurological exam, MRI, blood biomarker testing, verbal / spatial assessment, etc.). After these listed steps, embodiments may provide clinicians individualized treatment options. Lastly, embodiments differs from currently known solutions by utilizing voice and speech metrics to inform CDS.

[0022] Embodiments of the present disclosure provide holistic CDS solutions for assessing a host of factors and modalities known to affect a patient’s cognition and functional independence. For example, known solutions provide generic therapy options for cognitive and motor disorders such as a drug treatment or a certain lifestyle, in contrast to embodiments of the present disclosure providing a pathway to guide providers in further diagnosis of cognitive and motor impairments and personalized next step and therapeutic recommendations.

[0023] Embodiments of the present disclosure provide patient-specific CDS targeted to clinicians, and do not provide individualized treatments and recommendations. Additionally, some embodiments integrate cognitive and motor data into electronic medical records. Some embodiments provide customized CDS, not generic education, for various subtypes of mild cognitive impairment (amnestic mild cognitive impairment (MCI) and non-amnestic MCI),dementia (Alzheimer’s Disease Related Dementia, Vascular Dementia, Lewy Body Dementia, etc.), tremor, and Parkinsonism. Other diagnostic indications for neurologic disorders such as Amyotrophic Lateral Sclerosis (ALS), Multiple Sclerosis, Traumatic Brain Injury (TBI), Chronic Traumatic Encephalopathy (CTE), etc., may be incorporated. Embodiments of the present disclosure integrate with electronic medical records to flag diagnostic indications (detailed below) and provide CDS recommendations based on multiple modalities and factors related to a patient's medical history, providing patient-specific, pertinent information related to clinical decision making. Some embodiments incorporate patient-prioritized outcomes or input from caregivers. In this way, these embodiments provide personalized outcomes and recommendations to act as a guide for health care providers.

[0024] In addition, with the approval of many of the 140+ plaque-clearing drugs at various phases of FDA approval, embodiments of the present disclosure provide the CDS functionalities to inform clinicians about the patient's potential eligibility for these drugs (e.g., lecanemab (Leqembi®)) or alternative treatment options for patients ineligible for these drugs. Nearly 90% of patients are ineligible for FDA approved anti-amyloid monoclonal antibodies (lecanemab and donanemab). Embodiments of the present disclosure may be applicable to the application of these drugs and additional downstream treatments of affected patients and patient eligibility for these drugs.

[0025] Embodiments of the CDS provide patient-specific, individualized care pathway next steps based on objective and subjective measures. The objective measures include the scores and subscores of a digital cognitive assessment and bluetooth stylus-derived graphomotor metrics (detailed below). Subjective measures include surveys and questionnaires such as medical history, activities of daily living, and lifestyle questionnaires). Additional cognitive (memory, attention, executive function, visuospatial acuity, etc.) and motor features are extracted from voice and speech, vision, electronic health records, patient / car egiver reports, questionnaires, and electronic Person-Specific Outcome Measures. Patients are presented with assessments on a computing device such as a tablet, mobile device, or other suitable processor, then the computing device receives results from the assessments and provides the results and clinical decision support to the user, healthcare provider, or test administrator. The assessments are administered and scored, and the results with clinical decision support are provided all within the same system. Using these combined data, integrated and unified clinically appropriate next-steprecommendations are generated to guide clinicians in diagnostic next steps and personalized care. In the present context, “integrated” refers to no issue being raised as to whether the triggering source for the recommendation is an object or subjective measure, as the recommendations are presented as part of the same output. The CDS is also unified, meaning that the recommendations are consolidated to remove redundant or overlapping recommendations using a large language model. Integrated and unified CDS recommendations are presented as personalized clinical next steps.

[0026] Additionally, some embodiments allow for configurability at the level of healthcare organization (from a small practice to large healthcare systems) or individual providers. The preference determines which recommendations are to be shown to the user / provider / administrator by default and which recommendations are not shown by default (but can be selected if a provider chooses to do so). The CDS system for that organization or provider will remember the choices each site has deemed acceptable or appropriate for the population of interest at that specific location. A recommendation will populate a template specific to the organization’s or the provider’s preferences.

[0027] In brain health, providers (psychologists, neurologists, primary care providers, and other clinicians) must know off hand the clinical next-step recommendations in patient care or providers can look up next-step recommendations related to brain health outcomes.Embodiments of the present disclosure provide holistic CDS for a host of modalities known to potentially affect patients’ cognition or functional independence.

[0028] Another advantage of CDS system is the capability of our system to consider and adapt to clinicians’ preferences for certain diagnostic procedures for their patient population.Embodiments of the CDS system can be configured - whether at the provider or the organization level (from a small clinic to a major healthcare system) to incorporate clinicians’ preferences for including (or not including) certain diagnostic procedures, e.g., lab tests, types of imaging, genetic testing, etc. at certain stages during the workup for their patients.

[0029] An important advantage of some embodiments of the CDS solutions is its clinical validation the recommendations by senior, board-certified cognitive / behavioral neurologists affiliated with reputable academic institutions in the U.S. An IRB-approved clinical validation investigation was performed to evaluate the clinical appropriateness of the recommendations generated by embodiments of the present disclosure based on the patients’ performance in digitalcognitive assessment and their answers to the Linus Health Life and Health Questionnaire (LHQ) that screens for lifestyle and psychosocial risk factors related to brain health. The digital cognitive assessment used in some embodiments of the present disclosure is the Core Cognitive Evaluation (CCE™) which comprises the Digital Clock Drawing Test (DCTclock™), Immediate Recall, Delayed Recall (together being the DCR™), and the LHQ. The experts reviewed a set of 21 de-identified CCE reports of real-life patients, which in total contained at least 3 instances of CDS recommendations for each diagnostic indication. The CDS recommendation for each indication was rated for clinical appropriateness and relevance by expert cognitive / behavioral neurologists. Similarly, the experts rated the clinical appropriateness of the Linus Clinical Pathways; 9 parts in total, including whole pathways and partial nodes of the CDS decision tree for Red, Yellow, and Green DCR scores and a preliminary version of the Linus Clinical Pathway for disease-modifying treatment. The results of the specialists’ ratings indicated acceptable ratings for all CDS pathway recommendations for diagnostic indications related to cognitive impairment or risk of developing dementia. All CDS recommendations related to cognitive impairment received a median rating of 7 (out of 9) or above, whereas CDS recommendations related to concerns for tremor, Parkinsonism, and cholinergic impairment received median ratings of 5-5.5 (as shown graph 102 and table 104 in FIG. 1). Similarly, all CDS pathways except those for Green DCR scores (which included recommendations for tremor, Parkinsonism, and cholinergic impairment) and anti-amyloid (e.g., lecanemab) pathway received a median rating of 7 or above (as shown in FIG.2).

[0030] The results of the clinical validation study demonstrate the clinical appropriateness of the recommendations generated by embodiments of the CDS for diagnostic indications related to cognitive impairment via ratings by independent experts in cognitive neurology. A study was conducted soliciting ratings (and written feedback for improvement) from a number of independent clinical experts unaffiliated - and no history of collaboration - with Linus Health who provided ratings on each aspect of our CDS solution. Overall ratings are described in FIG.l and FIG. 2. Results indicated favorable evaluations of each aspect of embodiments of the CDS solution related to the evaluation of cognitive performance or estimating the risk of developing dementia in the future. The results also indicate the clinical appropriateness and satisfaction with the vast majority of the CDS recommendations and pathways implemented in CCE v2.0 as ratedby senior cognitive neurologists and the utility of the CDS functionality in the CCE for supporting clinical decision making among patients with cognitive impairment.

[0031] Importantly, the solution is configurable and personalized, and flags diagnostic indications for various subtypes of both mild cognitive impairment (e.g., amnestic MCI and non- amnestic (executive / vascular) MCI) and dementia (ADRD, VaD, LBD, etc.) specific to each individual. As a potential tool to identify patients who are likely to be suitable candidates for approved treatments, the solution provides both recommendations for therapies for suitable candidates while flagging unsuitable candidates for specialist referrals for further assessments, thus, substantially accelerating healthcare delivery while avoiding the imposition of undue harm upon patients who are at increased risk of side effects from such treatments. Further, some embodiments learn from clinicians’ interaction with the system. For example, some embodiments refine the recommendations according to the actions taken by the clinician and the results of that action. For example, the CDS recommends an MRI, the clinician orders it, and the results of the MRI aid in further diagnosis or not, the CDS remembers these steps and adapts the recommendations accordingly. Adaptation is done through the integration with electronic medical records to provide CDS recommendations based on multiple modalities and factors related to a patient's medical history including current and past medications and conditions, providing patient-specific, pertinent information related to clinical decision making. Lastly, another advantage of some embodiments is the personalized interventions and potential therapies based on personally defined treatment priorities for each patient.

[0032] Further, embodiments of the CDS system are iterative: the patient-specific next-step recommendations and pathways can be updated as more data is incrementally introduced to the patient profile. For example, the first pass of the CDS system on CCE results may note that there may be a hearing issue, and recommend a hearing screener. If the provider completes a hearing screener that is integrated with the CDS system, the recommendations can be automatically updated based on those results to suggest the next steps, such as proceeding to audiology (if results were positive), or proceeding to addressing other reversible causes (if the results were negative). This allows the recommendations to remain up to date as the patient completes additional assessments on the HCD platform.

[0033] The assessments on the health platform include an aggregate assessment of cognitive health that combines metrics from objective assessments, such as the Digital Clock and Recallcomprising the DCTclock™, Immediate and Delayed Recall speech assessments and subjective measures, including patient’s answers to the items on the Life and Health Questionnaire (LHQ), electronic Person-Specific Outcome Measures (ePSOM) to capture personally defined treatment priorities, and other questionnaires. The platform incorporates bluetooth stylus-derived praxis metrics to the CDS algorithm to capture brain dysfunction and guide clinical next steps and patient-specific recommendations. During the recall component of the DCR, subjects hear three words, which they are asked to immediately repeat (Immediate Recall). Subjects then proceed to complete the DCTclock™ assessment, after which they are asked to repeat the three words they heard (Delayed Recall). The DCTclock™ consists of several algorithm layers that together produce features that are then used as inputs for a series of stacked logistic regressions producing Composite Scales and a final DCTclock™ Score. The four composite scales correspond to Drawing Efficiency, Simple and Complex Motor, Speed of Information Processing, and Visuospatial Reasoning. Composite scales are calculated once for each of Command and Copy Clock drawings for a total of 8 composite scales.• Drawing Efficiency: The efficiency the individual demonstrated during the process of drawing each clock. This considers metrics such as time spent relative to properties of the drawing including number of pen strokes, stroke length, and size of the drawing.• Simple and Complex Motor: The graphomotor components involved in the process of drawing each clock. This considers metrics such as pen stroke speeds and oscillatory motion and can be helpful in parsing out motor and non-motor cognitive functions.• Speed of Information Processing: The non-motor cognitive functions used during the drawing process. This considers metrics such as absolute and relative duration of latencies, number of pauses, and relative time spent thinking vs. actively drawing with pen on the paper.• Visuospatial Reasoning: The visuospatial abilities demonstrated during the drawing process. This considers metrics pertaining to the geometric properties of the drawing including the circularity of the clock circle, placement of clock components, and drawing placement on the page.

[0034] Composite scores are generated by passing specific key features into three separate logistic regression classifiers trained to differentiate between cognitively unimpaired andcognitive impaired subjects. Thus, composite scales are weighted combinations of several underlying key features. The coefficients are hard-coded within the DCTclock algorithm codebase. All features, composite scales and the final DCTclock score are then standardized by age-category means and standard deviations for appropriate values. Mean and standard deviation values are hardcoded into the DCTclock algorithm code.

[0035] The final DCTclock score is used to categorize the participant into a DCTclock Score Classification. If the drawing is deemed unanalyzable by previous data validation steps it is given a DCTclock Score Classification integer of 3. Otherwise, a value is assigned according to the following criteria:• Score > 75: string = ‘No Cognitive Impairment Detected’, integer = 2• Score > 60 AND < 75: string = ‘Borderline for Cognitive Impairment’, integer = 1• Score > 0 AND < 60: string = ‘Likely Cognitively Impaired’, integer = 0

[0036] The combination of the DCTclock Classification metric based on the DCTclock, which is defined as a string in the form of either ‘Inside’, ‘Indeterminate’, or ‘Outside’, and the number of correctly remembered words from the Delayed Recall is used to produce a summary of the patient’s overall cognitive health. Incoming Amazon Simple Queue Service messages are first decoded by the DCR analytics engine. These messages contain all the upstream metrics required to execute the DCR algorithm. Next, DCTclock Classification is converted from a string [‘Inside’, ‘Indeterminate’, ‘Outside’] to an integer [2, 1, 0], respectively. This DCTclock Classification integer is then summed with the number of words correctly remembered during the delayed recall portion of the IDR assessment, which can range from 0 to 3. This produces a DCR Score ranging from 0 to 5, inclusive. The DCR Score is converted to a DCR Classification according to the following rules:• if DCR Score > 4, DCR Classification = ‘Not Indicative of Cognitive Impairment’• if DCR Score < 4 AND > 2, DCR Classification = ‘Borderline for Cognitive Impairment’• if DCR Score < 2, DCR Classification = ‘Indicative of Cognitive Impairment’

[0037] Based on the Immediate Recall, DCTclock, and Delayed Recall scores, the system executes a set of rules to determine what recommendation content is shown in the report. These rules may check anything from the following:The overall DCR scoreDCR subscores (i.e. Immediate and / or Delayed Recall Scores)• DCTclock composite scales and subscales (number of standard deviations below / above the age-matched norm) for either Command or Copy clocks• Memory Impairment Probability (MIP), described below

[0038] The following examples (in pseudocode) illustrate the level of complexity expected from the rules.Example rule: if (delayed recall > 2 & MIP <50% & DCTclock score < 60) { / / show "Concern for Executive Vascular Cognitive Impairment" concern in pathwayExample rule: if [(Delayed Recall = 0 or 1 AND Immediate Recall > 2) OR MIP > 75%>]& DCTclock score < 60 { / / show "Concern for Mixed-Domain Impairment" in pathway

[0039] The addition of voice and speech metrics to the DCR - DCR plus voice & speech (DCRP) - improves its cognitive-classification capabilities. One of the capabilities enabled by the DCR is the Linus Health Memory Impairment Probability (MIP). The MIP is a percentage score (0-100%) based on a logistic-regression classifier predicting delayed recall performance on the Rey Auditory Verbal Learning Test (RAVLT), which is a widely used neuropsychological test to measure verbal memory and its impairments in clinical and research settings. Predicting a patient’s delayed recall performance (e.g., on the RAVLT) is crucial because it reflects a given individual’s verbal episodic memory - the cognitive function that is the most relevant to AD pathophysiology.

[0040] The MIP leverages algorithmically selected voice / speech metrics extracted from the Immediate / Delayed Recall segments and features of the DCTclock to enable the DCR to estimate - with excellent precision (± 3%) - the likelihood of episodic verbal memory impairment in a given patient. The following examples illustrate the levels of the likelihood of verbal memory impairment estimated by the MIP:0 -49%: Low likelihood of verbal memory impairment50-74%: Moderate likelihood of verbal memory impairment75-100%: High likelihood of verbal memory impairment

[0041] These probability levels help determine the CDS recommendations for next diagnostic or interventional steps suggested to the PCPs based on the likelihood of the presence of amnestic (likely AD-related) cognitive impairment.

[0042] Subjective measures that inform the CDS recommendations may include questionnaires such as the LHQ, ePSOM, mood, functional dependence, frailty, fall risk, gait and balance assessments, and electronic medical records data, to name a few. The LHQ and ePSOM are briefly described herein. Information obtained from mood, functional, and other assessments are used to provide personalized actionable recommendations that an individual can take toward attaining better brain health both in the present and in the future.

[0043] These probability levels help determine the CDS recommendations for next diagnostic or interventional steps suggested to the PCPs based on the likelihood of the presence of amnestic (likely AD-related) cognitive impairment.

[0044] The LHQ component of the CCE is a quick screening of lifestyle and psychosocial factors relevant to overall brain health. Examples of LHQ questions include “I do tasks that challenge my brain (like reading, writing, drawing, or playing a musical instrument) every day or almost every day;” “I am generally satisfied with the course my life has taken;” and “Over the past year, I have been doing some form of moderate (cycling, fast walking) or vigorous exercise (running, spinning, playing basketball) at least three times a week.” The answers to the LHQ items are necessary for two important recommendation outputs of the Linus Health Healthcare Delivery Platform. First, the LHQ supports the CDS functionality of CCE in two ways, in the detection of concerns and in the form of recommendations that guide clinicians on best practices in cognitive decline and dementia care. Second, the patient’s answers to the LHQ items are the basis for generating individualized, patient-facing educational material that aims to empower individuals to take active steps toward better brain health. This patient education tool is informed by scientific evidence, highlights areas that need attention, and provides personalized actionable recommendations that an individual can take toward attaining better brain health both in the present and in the future.

[0045] ePSOM enables each patient to define what matters the most to them personally as they age. This questionnaire asks the patient to input free text / open responses to questions suchas “What matters to you about your daily tasks?”, “How confident are you in your ability to take part in [unique patient answer] ?”ePSOM is used in two important components of the Linus Healthcare Delivery Platform. First, the results of the machine-learning CCE algorithms and other recommended digital assessments are used to pre-populate the content of ePSOM, where questions relate specifically to an individual’s performance on cognitive testing. Second, ePSOM responses are the basis for provider- and patient-facing educational material that is customized to the individual’s unique needs and personally defined priorities.

[0046] Results of the objective cognitive assessment and subjective measures input into a large language model (LLM) trained to optimize output interventions for each cognitive domain taking into account lifestyle factors, patient’s priority areas, and other health related information obtained through assessments and questionnaires. This output is used to generate actionable CDS recommendations to guide clinicians in further assessments to rule out potential causes of cognitive impairment. These next steps contribute to the final set of recommendations given to the provider specific to the patient's objective and subjective assessment results. Accordingly, CDS includes actionable recommendations for providers, including additional diagnostic tools that can be referenced or even completed on the platform, follow-up clinical referral suggestions, and personalized care plans tailored by providers for their patients. The CDS also incorporates voice and speech metrics that will be used to identify potential reasons for cognitive impairment such as mood (e.g., depression, anxiety, and PTSD).

[0047] An example of the CDS for the CCE is depicted in FIG. 3. FIG. 3 depicts a flowchart 300 where DCR score ranges 302, 305, and 306 are input to the CCE. If the DCR flags an individual as borderline for cognitive impairment in the score range 304 or as indicative of cognitive impairment in the score range 306, CDS recommends first assessing the Immediate Recall component for a possible hearing concern. If no hearing concern is present, the clinical next steps are to screen for potentially reversible causes of cognitive impairment such as mood and medications (examples in Table 1). Further assessment of the reversible causes provides insights as to the cause of the cognitive impairment, the likely disease trajectory, and opens doors for interventions. If no reversible causes are present, or all such causes have been resolved in previous visits, CDS recommends considering cognitive impairment findings detected by the DCR such as diagnostic concerns for verbal memory or executive / vascular cognitive impairment triggered by specific rules based on established neuropsychological evidence and norm-basedthresholds. Each diagnostic concern identified elicits its own set of CDS recommendations, the Recommendations range from testing for verbal / spatial delayed recall, obtaining an MRI with specific notes on features to consider based on the flagged diagnostic indication, checking for blood-based biomarkers of AD, evaluating cardiovascular factors, and specialist referrals to name a few. If the DCR score is with the sore range 302, the output recommendation may include encouraging lifestyle management and retesting at a later date.Table 1: Potential reversible causes of cognitive impairment.

[0048] Using results from the cognitive assessments, analysis of hearing and vision data, eyetracking and facial-sentiment data, LHQ, ePSOM, and metrics obtained from electronic medical records, embodiments of the CDS will provide recommendations for next steps in personalized care such as lifestyle interventions and novel therapeutics. Based on the results of these assessments and taking into account the patient’s treatment priorities and any other objective or subjective information related to what is personally meaningful to them, the CDS provides a brain health action plan for the next steps in lifestyle and health management.

[0049] Another example of an embodiment of the CDS solution for a digital trail making test is shown in FIGs. 4A-4C. In FIG. 4A, exemplary text is provided where an individual is found to be impaired on the dTMT-Part B completion time. The corresponding number of errors (shown in FIG. 4B) indicate the number of errors a user would receive for respective recommendations and provides recommendations for dTMT-part B completion times. FIG. 4C is an exemplary text of the dTMT-part B consolidated recommendations.

[0050] Additionally, the pathway identifies suitable candidates for a specialist referral with blood biomarker testing for evaluation for anti-amyloid treatment, e.g., lecanemab (Leqembi®) and any similar anti-amyloid mAbs that are approved for patients with MCI or dementia due to AD. Patients who are classified as likely cognitively impaired and amyloid-beta positive - but are not classified as likely functionally dependent indicative of moderate-to-severe dementia are assessed with respect to inclusion and exclusion criteria.

[0051] Patients for whom disease modifying treatment is deemed to be safe and effective are selected for specialist referral. FIGs. 5A-5B illustrates the implementation of such a pathway 500 in the health assessment platform. Patients who receive a DCR score in score range 502, or convert from a score in the score range 504 to a score in the score range 502 are evaluated forrelative risk, as described above. Prioritized candidates are referred for blood biomarker testing and referral to a specialist to be considered for anti-amyloid treatment.

[0052] For patients with DCR scores in the score range 502, the predicted probability of the PET A0+ status is reviewed. For low / intermediate status, there is a recommendation to reevaluate after 3 months. For high status with a DCT score in the score range 502 or high status with a DCT score in the score range 504, functional impairment is predicted or an iADL QNR is completed. If no impairment is realized, this may be indicative of mild cognitive impairment or mild dementia. If all inclusion criteria is met with this recommendation, and none of the exclusion criteria is met, a non-contrast MRI may be performed, including a T1 FLAIR and T2 SWI or GRE, and review a DWI (preferably 3T) within the past three months. If MRI safety criteria is met, blood biomarker testing may be run as well as APOE4 genetic testing. Based on these results, the patient may be referred to a neurology department for a PET scan or CFS. If the patient is found to be AD positive, they are deemed eligible and high priority for an antibody prescription by the Neurology department. At any point throughout the process, the patient may be referred to Neurology for further assessment.

[0053] For patients with DCR scores in score range 504 and the low or intermediate predicted probability PET A0+ status, a reevaluation in 6 months is generally recommended. If the DCR score is in the score range 506, the reevaluation may take place within a year.

[0054] It should be noted that the Disease-Modifying (DMT) pathway below is integrated into a main CDS pathway described above such that the patient needs to ‘pass’ all the checkpoints in the main pathway that check for hearing concerns and other reversible / treatable causes before s / he is entered into our cognitive impairment pathway. If the patient’s performance on the DCR meets certain specified rules that flag diagnostic indications related to AD, the patient is entered into the DMT pathway, illustrated below. At each step of the pathway, if the assessment of patient’s cognitive, functional, or biomarker status is deemed unsuitable for treatment or if the patient meets any of the exclusion criteria related to safety (described in the text and / or in the diagram), the patient is not considered a suitable candidate for treatment and is referred to neurology for further assessment. If, however, the patient successfully passes the entire DMT decision tree by meeting all the eligibility criteria, and none of the diagnostic steps raise a safety or suitability concern, the patient is deemed high-priority for DMT prescription by Neurology.

[0055] Referring now to FIG. 6, a schematic of an example of a computing node is shown. Computing node 10 is only one example of a suitable computing node and is not intended to suggest any limitation as to the scope of use or functionality of embodiments described herein. Regardless, computing node 10 is capable of being implemented and / or performing any of the functionality set forth hereinabove.

[0056] In computing node 10 there is a computer system / server 12, which is operational with numerous other general purpose or special purpose computing system environments or configurations. Examples of well-known computing systems, environments, and / or configurations that may be suitable for use with computer system / server 12 include, but are not limited to, personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable consumer electronics, network PCs, minicomputer systems, mainframe computer systems, and distributed cloud computing environments that include any of the above systems or devices, and the like.

[0057] Computer system / server 12 may be described in the general context of computer system-executable instructions, such as program modules, being executed by a computer system. Generally, program modules may include routines, programs, objects, components, logic, data structures, and so on that perform particular tasks or implement particular abstract data types. Computer system / server 12 may be practiced in distributed cloud computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed cloud computing environment, program modules may be located in both local and remote computer system storage media including memory storage devices.

[0058] As shown in FIG. 6, computer system / server 12 in computing node 10 is shown in the form of a general-purpose computing device. The components of computer system / server 12 may include, but are not limited to, one or more processors or processing units 16, a system memory 28, and a bus 18 that couples various system components including system memory 28 to processor 16.

[0059] Bus 18 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a processor or local bus using any of a variety of bus architectures. By way of example, and notlimitation, such architectures include Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, Enhanced ISA (EISA) bus, Video Electronics Standards Association (VESA) local bus, Peripheral Component Interconnect (PCI) bus, Peripheral Component Interconnect Express (PCIe), and Advanced Microcontroller Bus Architecture (AMBA).

[0060] Computer system / server 12 typically includes a variety of computer system readable media. Such media may be any available media that is accessible by computer system / server 12, and it includes both volatile and non-volatile media, removable and non-removable media.

[0061] System memory 28 can include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Computer system / server 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 can be provided for reading from and writing to a non-removable, non-volatile magnetic media (not shown and typically called a "hard drive"). Although not shown, a magnetic disk drive for reading from and writing to a removable, non-volatile magnetic disk (e.g., a "floppy disk"), and an optical disk drive for reading from or writing to a removable, non-volatile optical disk such as a CD-ROM, DVD-ROM or other optical media can be provided. In such instances, each can be connected to bus 18 by one or more data media interfaces. As will be further depicted and described below, memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the disclosure.

[0062] Program / utility 40, having a set (at least one) of program modules 42, may be stored in memory 28 by way of example, and not limitation, as well as an operating system, one or more application programs, other program modules, and program data. Each of the operating system, one or more application programs, other program modules, and program data or some combination thereof, may include an implementation of a networking environment. Program modules 42 generally carry out the functions and / or methodologies of embodiments as described herein.

[0063] Computer system / server 12 may also communicate with one or more external devices 14 such as a keyboard, a pointing device, a display 24, etc. ; one or more devices that enable a user to interact with computer system / server 12; and / or any devices (e.g., network card, modem, etc.) that enable computer system / server 12 to communicate with one or more other computing devices. Such communication can occur via Input / Output (I / O) interfaces 22. Still yet,computer system / server 12 can communicate with one or more networks such as a local area network (LAN), a general wide area network (WAN), and / or a public network (e.g., the Internet) via network adapter 20. As depicted, network adapter 20 communicates with the other components of computer system / server 12 via bus 18. It should be understood that although not shown, other hardware and / or software components could be used in conjunction with computer system / server 12. Examples, include, but are not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data archival storage systems, etc.

[0064] The present disclosure may be embodied as a system, a method, and / or a computer program product. The computer program product may include a computer readable storage medium (or media) having computer readable program instructions thereon for causing a processor to carry out aspects of the present disclosure.

[0065] The computer readable storage medium can be a tangible device that can retain and store instructions for use by an instruction execution device. The computer readable storage medium may be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of the computer readable storage medium includes the following: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanically encoded device such as punch-cards or raised structures in a groove having instructions recorded thereon, and any suitable combination of the foregoing. A computer readable storage medium, as used herein, is not to be construed as being transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission media (e.g., light pulses passing through a fiber-optic cable), or electrical signals transmitted through a wire.

[0066] Computer readable program instructions described herein can be downloaded to respective computing / processing devices from a computer readable storage medium or to an external computer or external storage device via a network, for example, the Internet, a local areanetwork, a wide area network and / or a wireless network. The network may comprise copper transmission cables, optical transmission fibers, wireless transmission, routers, firewalls, switches, gateway computers and / or edge servers. A network adapter card or network interface in each computing / processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium within the respective computing / processing device.

[0067] Computer readable program instructions for carrying out operations of the present disclosure may be assembler instructions, instruction-set-architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, or either source code or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C++ or the like, and conventional procedural programming languages, such as the “C” programming language or similar programming languages. The computer readable program instructions may execute entirely on the user’s computer, partly on the user’s computer, as a stand-alone software package, partly on the user’s computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer may be connected to the user’s computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate arrays (FPGA), or programmable logic arrays (PLA) may execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present disclosure.

[0068] Aspects of the present disclosure are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer readable program instructions.

[0069] These computer readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processingapparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks. These computer readable program instructions may also be stored in a computer readable storage medium that can direct a computer, a programmable data processing apparatus, and / or other devices to function in a particular manner, such that the computer readable storage medium having instructions stored therein comprises an article of manufacture including instructions which implement aspects of the function / act specified in the flowchart and / or block diagram block or blocks.

[0070] The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process, such that the instructions which execute on the computer, other programmable apparatus, or other device implement the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0071] The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and / or flowchart illustration, and combinations of blocks in the block diagrams and / or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts or carry out combinations of special purpose hardware and computer instructions.

[0072] The descriptions of the various embodiments of the present disclosure have been presented for purposes of illustration, but are not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinaryskill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein was chosen to best explain the principles of the embodiments, the practical application or technical improvement over technologies found in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

CLAIMS1. A method, comprising: receiving a result of one or more multimodal assessments from a user interaction with a computing device; extracting one or more cognitive features from the result; applying a machine learning model to the one or more cognitive features to generate a composite score for the user; normalizing the composite score by age-category values; and predicting a cognitive impairment status based on the normalized composite score.

2. The method of claim 1 , wherein the machine learning model is an artificial neural network.

3. The method of claim 1, wherein the machine learning model is a logistic regression classifier.

4. The method of claim 1, further comprising outputting a recommendation based on the composite score.

5. The method of claim 1, wherein the result for one or more multimodal assessments comprises data collected from one or more of a touchscreen, a stylus, a webcam, and / or a microphone.

6. The method of claim 1 , wherein applying the machine learning model comprises applying a series of stacked logistic regressions producing a series of composite scores associated with features of cognitive function.

7. The method of claim 6, wherein the series of composite scores correspond to one or more of drawing efficiency, simple and complex motor function, speed of information processing, and / or visuospatial reasoning.

8. The method of claim 1, wherein the composite score is a weighted combination of the one or more cognitive features.

9. The method of claim 1, further comprising: completing a populated brain health questionnaire based on the composite score; and based on answers to the questionnaire, providing a healthcare provider with one or more assessments to administer to the user.

10. The method of claim 9, further comprising providing the answers to a large language model to thereby optimize a final set of recommendations for the healthcare provider.

11. The method of claim 1, wherein the result of the multimodal assessment comprises data collected from one or more of a touchscreen, a stylus, a webcam, and / or a microphone.

12. The method of claim 1, wherein the cognitive impairment status comprises a likelihood of verbal memory impairment.

13. A system comprising: at least one input device; a computing node coupled to the at least one input device and comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor of the computing node to perform a method comprising: receiving a result for one or more multimodal assessments of a user interaction with a computing device; extracting one or more cognitive features from the result; applying a machine learning model to the one or more cognitive features to generate a composite score for the user;normalizing the composite score by age-category values; and predicting a cognitive impairment status based on the normalized composite score.

14. A computer program product for determining a cognitive impairment status, the computer program product comprising a computer readable storage medium having program instructions embedded therewith, the program instructions executable by a processor to cause the processor to perform a method comprising: receiving a result for one or more multimodal assessments of a user interaction with a computing device; extracting one or more cognitive features from the result; applying a machine learning model to the one or more cognitive features to generate a composite score for the user; normalizing the composite score by age-category values; and predicting a cognitive impairment status based on the normalized composite score.

Images