JP7112660B2 - Electronic distribution of information in personalized medicine - Google Patents

Description

Embodiments described herein relate to electronic delivery of information in personalized medicine.

Healthcare is becoming more and more personalized, meaning that treatments are tailored to a patient's individual health data, including genotypic and phenotypic data. Genotypic data can include selected genetic markers, single nucleotide polymorphisms (SNPs), or the entire genetic sequence. Phenotypic data can include laboratory results such as physical examination data from patients, clinical scores and rating scales, in vitro tests, and in vivo imaging data such as magnetic resonance imaging (MRI) scans. The cost of sequencing is dropping rapidly with new technologies such as next-generation sequencing (NGS), and it is expected that such data will become as universal and inexpensive as MRI scans. Wearable sensors embedded in consumer electronic devices, such as accelerometers and mobile electrocardiograms (ECGs), are emerging that provide a means of continuously measuring phenotypic data in real time over the Internet. , bring “digital health”.

Diagnosis is the first step in defining the precise nature of a patient's disease state and typically involves physical measurements converted into digital information such as MRI scans to files in DICOM image format. Laboratory data (laboratory results) can be converted into Portable Document Format (PDF) files or delivered in structured Health Layer 7 (HL7) format. The patient's disease state is then "stratified" based on general characteristics, and then a treatment regimen tailored to achieve the optimal patient outcome is selected.

Alzheimer's disease (AD) diagnosis is particularly complex in the early stages of the disease (precursor or presymptomatic disease). Diagnosis may include clinical scores (such as cognitive tests) and advanced biomarkers such as quantitative MRI data. Patients with cognitive problems are typically seen initially by a busy non-specialist primary care physician (Primary Care Physician: PCP) who ultimately refers the patient to a specialist memory clinic. However, early diagnosis of Alzheimer's disease is often delayed years after the first cognitive symptoms. Tests are repeated, often because data cannot be easily shared because they have quality issues and lack standardization, or simply because specialists did not have access to previous tests. Occasionally, a costly PET scan is requested to the primary care physician very early in the process without first staging the diagnostic process from low-cost screening to confirmatory diagnosis so as to incrementally increase diagnostic certainty. directed by

The subject matter claimed herein is not limited to embodiments in which any disadvantages are resolved or which operate only in environments such as those described above. Rather, this background is provided merely to illustrate one exemplary technology area in which some embodiments described herein may be practiced.

According to an aspect of an embodiment, a method of delivering information corresponding to personalized healthcare in a clinical non-research setting includes capturing one or more data streams , each of the data streams relating to patient care. The method can further include consolidating the data streams to generate consolidated diagnostic data and analyzing the consolidated diagnostic data to generate analyzed diagnostic data. The method further comprises curating the analyzed diagnostic data and generating an integrated report for presentation to the patient's physician based on the analyzed and curated diagnostic data. can contain.

The objectives and advantages of these embodiments will be realized and attained through at least the elements, features, and combinations particularly pointed out in the claims. It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are not restrictive of the invention as claimed.

Illustrative embodiments will be described and explained with additional specificity and detail through the use of the following accompanying drawings.
FIG. 1 is a diagram illustrating an exemplary social network for personalized medicine. FIG. 2 is a diagram illustrating exemplary data sharing of case information between participants in a social network for personalized medicine. FIG. 3 is a diagram illustrating an exemplary referral data flow between participants of a private network within a social network for personalized medicine. FIG. 4 is a diagram illustrating exemplary network management of a private network within a social network for personalized medicine. FIG. 5 is a diagram illustrating an exemplary data flow for adding new information to a case within a social network for personalized medicine. FIG. 6 is a diagram illustrating an exemplary diagnostic and therapeutic data flow within a social network for personalized medicine. FIG. 7 is a diagram illustrating an example of adding and integrating case information within a social network for personalized medicine. FIG. 8 is a diagram illustrating exemplary components of a system using social networks for personalized medicine. FIG. 9 is a diagram illustrating an exemplary system for personalized medicine. FIG. 10 is a dataflow for an exemplary method of generating a report for presentation. FIG. 11 is a dataflow for an exemplary method of generating a report for presentation. FIG. 12 is a dataflow for an exemplary method of generating a report for presentation. FIG. 13 is a dataflow for an exemplary method of generating a report for presentation. FIG. 14 is a dataflow for an exemplary method of generating a report for presentation. FIG. 15 is a flowchart of an exemplary method for distributing information corresponding to personalized healthcare in a clinical, non-research setting. FIG. 4 shows the scientific rationale for four novel targets for diagnostic (Dx) and/or therapeutic (Rx) use in personalized medicine for AD.

Some embodiments herein describe methods and systems for a digital health platform for personalized medicine, particularly in the field of Alzheimer's disease diagnosis. The present disclosure provides the integration of various data streams captured from physical measurements of a patient's disease state, as well as primary care physicians and physicians to facilitate the delivery of diagnoses and personalized treatments in Alzheimer's disease and other diseases. The electronic routing of such information between specialists and data analysis units is described. The system incorporates a scalable cloud-based social network architecture that complies with the Health Insurance Portability and Accountability Act (HIPAA). The cloud-based social network architecture manages the exchange of personal health information, including encrypted file transfer and messaging, between social network participants. The exchange of diagnostic information is permission-based, with referrals to specialists improving diagnostic certainty and augmented by data analysis units.

Cloud-based social network architectures effectively enable healthcare providers to collaborate on patient cases and share data in a regulatory compliant manner. In addition, the data analysis unit of the cloud-based social network architecture enables optimization in the workflow of diagnostics between healthcare providers, as well as reducing unnecessary examinations, improving data quality and diagnostic utility. improve quality of life and enable non-specialist physicians to take advantage of best practices.

Embodiments of the present invention will be described with reference to the accompanying drawings.

FIG. 1 is a diagram illustrating an exemplary social network 100 for personalized medicine, arranged in accordance with at least some embodiments described herein. The social network illustrates various participants including existing users 110 , new users 120 , agents 122 , consultants 130 , laboratories 140 and data analysis units 150 . Various actions may be performed on social network 100 and participants within social network 100 . For example, participants are added to the social network 100, participants within the social network 100 collaborate on patient medical care, and patient medical care data is aggregated and analyzed. be able to.

To add participants to the social network, participants may first be identified. For example, an existing user 110 , who may be a family doctor, may identify a new user 120 , who may be a specialist neurologist, to be added to social network 100 . Existing users 110 may have the option of searching external provider databases, such as the National Plan and Provider Enumeration System (NPPES), to provide new users 120 with initial contact information. Existing users 110 can also verify correct electronic contact information, for example, via telephone.

When a participant is being added to the social network 100, the participant can designate the participant in the private portion of the social network 100 or the non-private portion of the social network 100. In some embodiments, the private portion of social network 100 may be for participants associated with a hospital network. In some embodiments, social network 100 may have multiple private parts. As an example of the designation of participants in the private portion of social network 100, existing user 110 may have a private portion of social network 100 when new user 120 is part of a hospital network with which existing user 110 is associated. New users 120 can be designated to become part of the part. In these and other embodiments, consultant 130 may be an external participant, such as a nuclear medicine specialist outside the hospital network, who may perform an emission tomography (PET) scan or other diagnostic examination.

New users 120 may be added to social network 100 in a variety of ways. For example, in some embodiments new users 120 may be added to the social network by invitation. An existing user 110 enters basic information such as email and name of the new user 120 into a system used to support the social network 100, such as the system described with respect to FIG. You can initiate contact with After entering basic information or indicating basic information to the new user 120, the existing user 110 sends the new user 120 an invitation to join the social network 100 for personalized medicine. may In some embodiments, an invitation to a new user 120 may be automatically generated from a template after an existing user 110 indicates the new user's 120 identity.

The invitation may be sent by email or other messaging system and may contain a link to the social network 100 login page. An initial password for new user 120 to access social network 100 may also be automatically created. The initial password for security purposes may be sent in a separate message or delivered by calling the new user's 120 office or the new user's 120 mobile phone. In some embodiments, password authentication may be supplemented or substituted with biometric authentication such as fingerprint, voice, facial recognition, and/or another electronic access control system. Alternative electronic access control systems may include, by way of example and without limitation, smart phones with card-based and electronic authentication. In some embodiments, authentication methods include knowledge factors (such as passwords), possession factors (such as special access cards issued by social network providers), and unique factors (biometric factors, such as voice or video authentication). It may be multi-factor authentication, such as two-factor authentication (TFA), which may use the presentation of two or more of the three authentication factors of .

After the user receives an invitation to become part of the social network 100, the new user 120 changes the initial password to a password in a secure format according to the user's choice (e.g. having certain length and characters). and adding information, such as entering professional details, address, and other contact information such as pager, mobile phone, fax numbers, preferences, etc. into the form.

In some embodiments, the registration process may not require the new user 120 to enter a password or add any additional information about the new user 120 . For example, authentication may consist solely of TFA as described above. In these and other embodiments, the TFA verifies the new member's credentials, obtains a biometric, and then distributes a special access card or some other access information for the social network 100. may consist of In some embodiments, biometrics may be obtained by existing users 110 from new users 120, for example, at a medical conference. Additional information about the new user 120 may also be obtained through other sources such as provider databases or certificate authorities. Additional information may be entered into the form regarding the new user 120 to facilitate the registration process for the new user 120 while maintaining security and privacy.

In some embodiments, when a new user 120 is part of the private portion of social network 100, an administrator associated with the private portion of social network 100, through an administration module associated with social network 100, New users 120 may be added. The administrator may be an existing user 110 and/or another participant in the social network 100 not illustrated in FIG. In these and other embodiments, the administrator authorizes participants within the social network 100 regarding their ability to invite, receive, and/or inquire/send new members within the social network 100. may be restricted.

In some embodiments, when a new user 120 is added to the social network 100, the new user 120 is assigned an agent, such as a nurse practitioner or physician's assistant, who acts on behalf of the new user 120. 122. In these and other embodiments, an administrator may add agents 122 through the administration module. In some embodiments, the social network 100 invites participants within the social network 100 to participate within the social network 100 for collaborative purposes, such as to collaborate on patient medical care. person can be searchable.

The social network 100 allows participants within the social network to communicate regarding specific treatments using two-way secure messaging communications and/or video/audio conferencing within the social network 100. It may be further configured. Alternatively or additionally, participants within the social network may communicate using adapted mobile communication applications. Existing communication channels such as text (SMS) messaging, pagers, or email may also be utilized for non-secure messaging/alerts. For example, non-secure messaging may be used to indicate to a physician that a physician's attention is required, or to notify a participant of recent activity such as information changed, added, or updated regarding a patient. may be

As mentioned above, the participants may collaborate on patient medical care. As used herein, patient care (medical care) may be referred to herein as a patient case or case. As shown, participants within the social network 100 may collaborate on patient cases. This cooperation may be based on a referral process. For example, an existing user 110, who may be a primary care physician (PCP), may share a patient case with a specialist, etc. after a new user 120 joins the social network 100 or laboratory 140 for further diagnostic evaluation. of new users 120 may be queried. An existing user 110 can serve as a “patient case owner” during a particular episode of patient care and can provide basic patient case data such as name, gender, date of birth, contact information, and patient details such as insurance information can be entered. Existing users 110 can also enter descriptive information about the case, such as priority, type of referral, expected response to the referral, and case summary information. Descriptive information, such as summary information, can also be imported from another healthcare system, such as an electronic medical record (EMR) system. An existing user 110 can further refer the case to multiple members within the private portion of the social network 100 or within the social network 100 as a whole to collaborate on the case. For example, an existing user 110 can refer a case to a consultant 130 during an episode of care (eg, a stage of treatment) until the episode of care is considered complete by the existing user 110.

Case collaboration between participants in the social network 100 includes sharing information with participants to determine cases (e.g., patient test results, laboratory results, diagnoses, patient medical history, etc.). can contain. In some embodiments, social network 100 may include cloud storage for storing information about cases. Participants in the social network 100 associated with the case may be able to access the cloud storage for the case and add information to or retrieve information from the cloud storage. Adding information to cloud storage or retrieving information from cloud storage for a particular case may be referred to herein as adding information to the case or retrieving, displaying, or accessing information from the case. Because the case information is part of the social network 100, participants in the social network 100 associated with the case can access information from cases added by other participants.

For example, an existing user 110 can add additional case content (along with accompanying metadata/descriptive information) to the case. Additional case content may include results of cognitive screening tests, genetic tests, and/or blood tests for diseases such as Alzheimer's disease. Inspection results may be provided in an electronic format, such as an inspection report in PDF document format. Reports can be added to cases by file upload or directly from another healthcare system, such as an electronic medical record (EMR) system. Medical images can be added to the case as well. For example, medical images can be added to a case by uploading a Digital Imaging and Communications in Medicine (DICOM) file or directly from a Picture Archive Communication System (PACS). Information from tests such as cognitive screening tests on mobile devices performed at the patient's home may be added to the case. Alternatively or additionally, information from examinations from the PCP or specialist's office, augmented or not by a third party service provider, such as a laboratory service or data analysis unit, may be added to the case. can be In some embodiments, the information that can be added to the case in the form of test results and/or reports is normative and/or age-related ranges, normative and/or age-related ranges, patient as well as medical images of the patient or other representative illustrative cases. Information that may be added to the case in the form of test results and/or reports includes contextual and/or interpretive information pointing to URLs or publications, excerpts or summaries of one or more publications. may further contain

Collaboration among participants within the social network 100 is facilitated by participants requesting reviews from other participants. For example, an existing user 110, who may be a PCP, may request a new user 120, such as a specialist, to join the social network 100 and be associated with a case within the social network 100. Alternatively or additionally, existing users 110 may identify new users 120 as part of social network 100 using some other method, such as a social network or a directory website outside social network 100. can do. In these and other embodiments, an existing user 110 can request that a new user 120 collaborate with an existing user 110 on a case. Requests may be issued in similar or different manners as described above. After a new user 120 sets up to start collaborating on a case, an existing user 110 can send the case to the new user 120 for further evaluation and/or review. In some embodiments, an existing user 110 can indicate to a new user 120 that information has been added to a case for the new user 120 to review. As mentioned, presenting new information may be performed manually or automatically, for example by messaging, after the new information has been added to the case.

For example, an existing user 110 can send a patient's case to a new user 120, where the existing user 110 is a PCP and the new user 120 is a consulting neurologist for further evaluation. is. Further assessment may include comprehensive neurological and/or neuropsychological testing with a computerized cognitive battery. Prior to sending a patient's case to a new user 120, an existing user 110 may provide the new user 120 with screening test results and/or other information such as the patient's medical history and medications, as well as case content such as a case summary. may have been added. A new user 120 can review the case summary information and associated messages through secure mobile messaging or a mobile case dashboard application provided by the social network 100 for displaying information about the case. The dashboard application allows previewing of available case content, eg reports, or pre-existing images such as magnetic resonance imaging (MRI) or PET scans.

In some embodiments, the dashboard application may be navigated through a natural user interface (NUI) driven by voice, touch, gestures, eye tracking, and other inputs. Dashboard applications are displayed on flexible or wearable display devices, such as glasses capable of displaying a variety of wearable or projected displays and/or contextually-aware superimposed information (augmented reality). can be In some embodiments, the dashboard application may also be a browser-based application with or without NUI input, case content such as images for further review in a third party viewer, or DICOM It may be possible to download an application such as an image viewer. In these and other embodiments, the browser-based dashboard application resides within a secure corporate computing environment, such as a dedicated workstation or within the firewall of a hospital, medical center, clinic, or some other firewall. access device.

When information (such information is images or laboratory or cognitive test data) is added to a case, the information may be obtained without stringent quality standards in place and/or may be collected between cases or in the same case. It does not have to be a form that is comparable within. As a result, the information may not be suitable for further analysis, which may include quantitative image analysis, next generation sequencing (NGS) genomic analysis or gene expression analysis. To ensure comparability of data elements within and between cases, the information added to the cases may be checked for adherence to quality criteria by the data analysis unit 150 . For example, data may be analyzed using automated analysis, amyloid PET scans, hippocampal volume quantification, or fiber tractography using diffusion tensor imaging (DTI) with MRI imaging.
It can be checked for the specific source data acquisition parameters and equipment used before further processing such as by tractography. When information does not meet certain criteria, e.g., if the information does not allow suitable quantification, a message may be sent back to the participant sending the information or to other participants connected to the case.

In some embodiments, information shared in cases within social network 100 may not be in quantitative form. For example, images produced by MRI or PET scans may not have quantitative information associated with them. However, quantitative information may be obtained from the images. In these and other embodiments, image data and other information that is not quantitative in nature is subject to quantitative image analysis to aid diagnosis and to share with other participants in social networks. good. For example, a radiologist may perform quantitative analyzes for longitudinal comparison and/or treatment decision making. Quantitative image analysis may be performed in addition to the quantitative readout of scans such as MRI to rule out other diseases or aid in differential diagnosis, and summarized in a qualitative report (qualitative report) ( can be summarized easily). Quantitative image analysis may be fully automated or semi-automated with operator interaction (interaction) and may be performed at the site/premises on a workstation or server appliance. Quantified data or results may then be imported into the case. In some embodiments, quantitative image analysis may be performed on demand in a data analysis unit 150 separate from where information is gathered. In some embodiments, quantitative image analysis may generate a report, such as a PDF format. The report includes the normative and/or age-related ranges and the patient's individual quantitative image values (eg hippocampal volume plots) for the normative and/or age-related ranges. The report may also include images of selected patients. Quantitative image reports (quantitative image reports) may be interactive and allow the display of real medical images in 2D, 3D or 4D (3D over time) and data points are selected. Advanced visualization features may be included, such as quantitative values plotted overlaid with their respective source images as they occur.

It shall be understood that the data in known medical reports (medical reports) and even current EMRs are qualitative and text-based (often free text), and are highly tolerant of variability and ambiguity in language. . However, it is desirable to have the data in quantitative form or to utilize a standardized vocabulary (ontology) such as the Systematized Nomenclature Of Medicine Clinical Terms (SNOMED). However, requiring specific data types may limit the use of all existing medical information available within a patient's case. To solve these problems, content curation may be used as an intermediate step for advanced capabilities such as integrated reporting, searching, semantic integration, and data mining/advanced analysis. good. Existing quantitative data are concise, presenting one or several biomarkers along with contextual information such as relevant excerpts from medical guidelines and/or medical literature, or links to original references. May be compiled (compiled) into an integrated reporting format. In some embodiments, key original references may be included in their entirety in the integrated report. Existing non-standard information, such as text data, may be annotated using standardized vocabularies for subsequent processing such as retrieval, semantic integration, and data mining purposes. Content curation can be done using fully automated or semi-automated auto-curation software tools and databases in the data analysis unit 150 or, after data anonymization, in third-party on-demand services such as Amazon Mechanical Turk. may be used.

In some embodiments, data analysis unit 150 may be configured to generate an integrated digital diagnostic report (digital diagnostic report) after performing quality control, quantitative image analysis, and curation processes. good. Integrated digital diagnostic reports are generated from laboratories 140, new users 120, existing users 110, agents 122, consultants 130, and/or other participants in social network 100 collaborating on cases. , one or several biomarkers or aggregated efficacy indicators may be combined. Data analysis unit 150 may then curate the shared information into an organized display for evaluation by physicians, such as existing users 110 . In some embodiments, reports may be embodied in mobile applications. In some embodiments, reports may further provide longitudinal information on biomarkers or efficacy measures in the form of plots or other advanced visualizations described above. In some embodiments, an existing user 110 or any other participant in the social network 100 reading the report will be contacted at the call center by some of the messaging and value-added diagnostic services within the social network. , a specialist physician (a person skilled in interpreting the pooled data in the report) may be consulted.

In some embodiments, personally identifiable information such as name, date of birth, address and other patient-identifying information may be stripped from information within social network 100 in the anonymization process. In some embodiments, a third party or open source data anonymization software tool may be employed at the data analysis unit 150 to provide the anonymization process. An exemplary anonymization process may include retrieving patient identification information within the DICOM image header. Anonymization may be fully automated (such as when the data is in a standardized format) or semi-automated. In some embodiments, patient data may be anonymized within social network 100 . The social network 100 may assign a temporary or permanent unique identification number (UID) to the patient data. The social network 100 may allow physicians interacting with patients associated with patient data to associate patient data with a particular case or patient. In some embodiments, temporary or permanent unique identification numbers (UIDs) assigned to patient data may be used in conjunction with diagnostic services such as home screening. For example, a patient may obtain a prepaid card containing a UID for a pharmacy. The patient may then visit a PCP that is part of social network 100 . PCPs may request anonymized and consolidated screening reports (screening reports) from data analysis unit 150, by messaging within social network 100, and at call centers as part of value-added diagnostic services. A specialist physician (one skilled in interpreting biomarker combinations/patterns) may also be consulted. In these embodiments, the specialist physician and data analysis unit 150 may not know the patient's identity. Rather, the specialist physician and data analysis unit 150 may simply associate the data with the associated UID. As such, the patient's identity may remain confidential within the social network 100 even when the patient's case is being worked on by various participants within the social network 100 .

In some embodiments, social network 100 may aggregate data collected in various cases. For example, after quality control, curation and anonymization steps performed in data analysis unit 150, data from multiple cases may be aggregated in a centralized or federated database and advanced analysis may be performed on the database. executed. Non-image and/or non-sequencing data may be stored in SQL or NoSQL databases (such as Cassandra), while media-rich content such as images or NGS source data may be stored in the file system for performance reasons. A database may support semantic data integration to correlate data from other databases and datasets. File system storage may be distributed, such as in a Hadoop Distributed File System (HDFS). Image files may reside in external image repositories referenced by URL links or other pointers stored in a database and optimized for performance.

After data from one or more cases is aggregated into a database, various sophisticated "big data" analyzes may be performed on the aggregated data. Advanced analysis may involve the process of examining large amounts of data of various types to uncover hidden patterns, unknown correlations and other useful information. Advanced analysis may be performed in data analysis unit 150 . For example, advanced analysis may include predictive analysis based on machine learning algorithms and/or data mining, or statistical analysis techniques (eg, using R). Advanced analysis may be distributed (such as in Map Reduce) or parallelized. For example, advanced analysis may be based on combinations of biomarker patterns and/or genetic profiles to predict treatment response or future onset of disease in presymptomatic Alzheimer's disease patients, or factors contained in case information in databases. ), the probability of the current disease given a certain combination of . Advanced analysis may also be used to perform advanced predictive analysis based on PET scans (eg, fully automated amyloid PET or tau tracer quantification). Alternatively or additionally, the advanced predictive analysis may be based on diffusion tensor imaging (DTI) MRI or functional MRI-based predictive brain network "connectome" analysis. Advanced analysis may be performed on a particular case to assist in determining diagnosis, treatment, or other aspects associated with a particular case within social network 100 .

In some embodiments, the advanced analysis finds and ranks cases with known treatment outcomes similar to the specific case to provide personalized treatment in the specific case. As noted, it may further include using semantic search against the aggregated data repository. The results of advanced analysis may be summarized or presented in reports. In some embodiments, reporting is done by participants within the social network 100 through an adapted Personalized Healthcare (PHC) dashboard driven by a Natural User Interface (NUI), as described above. may be accessed.

In some embodiments, advanced analysis may also be performed by data scientists. Data scientists may derive new knowledge from aggregated databases that can support diagnostic certainty and/or provide treatment stratification in specific cases.

Modifications, additions, or omissions may be made to social network 100 without departing from the disclosure of the present disclosure. For example, social network 100 may include participants other than those described above. Further, social network 100 may include various aspects other than those described above. For example, other aspects of social networks may be described herein with respect to other figures.

FIG. 2 is a diagram illustrating an example of data sharing of case information among participants of a social network 200 for personalized medicine, arranged in accordance with at least some embodiments described herein. be. Specifically, FIG. 2 illustrates data sharing between a primary care physician (PCP) 210 , a specialist neurologist 212 , a radiologist 214 , and a data analysis unit 216 . Participants in social network 200 can further collaborate on shared case information 226 using messages 218, for example.

FIG. 3 is a diagram illustrating an exemplary referral data flow between participants of a private network 320 within a social network 300 for personalized medicine, including at least some of the data flows described herein. Arranged according to the embodiment. Specifically, FIG. 3 illustrates query data flow between participants of a private network 320 that is part of a social network 300, such as within a hospital, that includes existing users 310, Includes physician 312 and MRI facility 316 . FIG. 3 also illustrates an external network 322 that may be part of social network 300 . External network 322 may include PET facilities 314 configured to share data with private network 320 participants.

FIG. 4 is a diagram illustrating exemplary network management of a private network 420 within a social network 400 for personalized medicine, arranged in accordance with at least some embodiments described herein. be done. Specifically, FIG. 4 illustrates network management of a private network 420, such as in a hospital, through a private network administrator 410, which includes a specialist neurologist 412 and a surrogate. New private members such as person 416 and their representatives can be added. FIG. 4 also illustrates external network 422 that is part of social network 400 . External network 422 may include PET facilities 414 configured to share data with private network 420 participants.

FIG. 5 is a diagram illustrating data flow for adding new information 516 to a case within a social network 500 for personalized medicine, arranged according to at least some embodiments described herein. is set. Specifically, FIG. 5 illustrates information 516 being added by a neurologist 512 associated with PCP 510 and radiologist 514 in social network 500 . PCP 510 and/or neurologist 512 may further query/consult data analysis unit 518, for example, to generate an integrated report on patient care. A consolidated report may be added to the case by the data analysis unit 518 .

FIG. 6 is a diagram illustrating exemplary diagnostic and therapeutic data flow within a social network 600 for personalized medicine, arranged in accordance with at least some embodiments described herein. . Specifically, FIG. 6 illustrates diagnostic and therapeutic data flow in a personalized medicine platform for Alzheimer's disease among participants of social network 600 . For example, FIG. 6 illustrates that the PCP 610 may perform a patient cognitive screening test 620, which prescribes an APOE test and/or a blood screening test for Alzheimer's disease performed in the laboratory 616. can include After PCP 610 receives a laboratory report (laboratory report), it may consult data analysis unit 624 to consolidate the screening data into a consolidated report. PCP 610 may also refer the patient's case to a specialist neurologist 612 for further evaluation of the patient. These steps with PCP610 can conclude a case screening episode for a patient. A neurologist 612 may order another IVD test from the lab 616 . The IVD test may be accompanied by a CSF Abeta/Tau test. The specialist neurologist 612 may also complete a comprehensive examination 618 in the patient, which in some embodiments may include a fully computerized cognitive battery. Specialist neurologist 612 may also refer/order MRI and/or PET scans 622 from radiologist 614, which may include image quantification. The specialist neurologist 612 may then consult the data analysis unit 624 to consolidate the comprehensive assessment data into a consolidated report. These steps by a specialist neurologist 612 can conclude an overall diagnostic episode of care. In some embodiments, data analysis unit 624 may provide actionable information for specialist neurologist 612 or PCP 610 to prescribe personalized medications for patients associated with Alzheimer's disease, mentioned above. Prognostic reports may be generated for treatment stratification, such as those described above. In some embodiments, the PCP 610 is augmented by the data analysis unit 624, prior to prescribing a personalized medication, for the function of comprehensive assessment and/or subsequent treatment stratification as mentioned above. Some or all of them may be implemented.

FIG. 7 is a diagram illustrating an example of adding and integrating case information within a social network 700 for personalized medicine, arranged in accordance with at least some embodiments described herein. be. Specifically, FIG. 7 illustrates adding 714 cognitive scores 720 from a cognitive screening test directed or performed by a separate piece of information, eg, PCP 710 , to data analysis unit 724 . In some embodiments, cognitive score 720 may be further supplemented by APOE test and/or blood screening test results, which may be obtained from laboratory 718 . Other test results, such as image data 726 and/or radiological report 728, may also be added to the case. Alternatively or additionally, the data analysis unit 724 may also add an integrated report 716 to the case.

FIG. 8 is a diagram illustrating exemplary components of a system 800 using social network infrastructure 837 for personalized medicine, arranged according to at least some embodiments described herein. is set. The system 800 may include a social network infrastructure 837 that can interact with a local case management layer 829 , an anonymizer, a data aggregation and analysis layer 838 , and a physician presentation layer 826 . Social network infrastructure 837 may, in some embodiments, be configured to enable interaction between participants of the social network, as described with respect to FIGS. 1-7.

The social network infrastructure 837 may include an extensible cloud-based case content delivery network 839 that manages the HIPAA-compliant exchange of personal health information, which network 839 provides full audit trail, encrypted Including the transfer of files and messaging between participants in social networks hosted by the social network infrastructure 837 . The social network infrastructure 837 is system software such as front-end (client) and server-side application software code (e.g., web-enabled) or desktop virtualization applications that can be used to implement the core functionality of the social network. It may further include software components. The social network infrastructure 837 also includes APIs for external applications to interface with social networks; application frameworks such as the web and rich media application frameworks, database server software, and web server software; server virtualization software, load It may include balancers, networking devices, and servers and storage devices. The social network infrastructure 837 may include server devices that may include diskless server nodes with solid state drives (SSDs). The social network infrastructure 837 may include storage devices including flash array storage. Social network infrastructure 837 may also include cloud-networking equipment such as low-latency network switches, and may further include network security and encryption appliances.

The case content delivery cloud 839 is further deployed in private cloud infrastructure, such as facilities owned or rented by social network operators, and/or in dedicated cloud facilities with appropriate security implemented and managed by third parties. good too. For example, the case content delivery cloud 839 can be hosted in a public cloud such as Amazon or in combination with a private cloud, e.g. a hybrid with sensitive, non-anonymized information such as protected health information. It may be deployed in a cloud architecture. In some embodiments, case content delivery cloud 839 is a Force. It may also be deployed in a PaaS (platform as a service) environment such as PaaS.com.

In some embodiments, the case content delivery cloud 839 is further linked to specialized (such as data transfer rates and certain viewers) third-party cloud-based repositories, such as image or genome sequencing data repositories, through the use of APIs. may be connected. In these and other embodiments, pointers such as URLs or XDS may be used to link content stored in case content delivery cloud 839 to corresponding case content in third party repositories.

Social network infrastructure 837 may be configured to communicate with local case management layer 829 . The local case management layer 829 may be configured to enable local case management by administrative personnel such as physician assistants, nurses or technicians, and by corporate or healthcare provider (hospital or clinic) etc.) may reside within the firewall 860 . Patient cases within the social network infrastructure 837 can be installed on a web browser interface 832 (such as Firefox, Internet Explorer, Chrome, or Safari) or device 834, although the social network infrastructure 837 can be cloud-based. It may be accessed and managed through a device 834, such as a hospital desktop, laptop computer or mobile device, via a non-browser-based native application (eg, iOS, Android, Windows, or Mac OS software applications). Case content such as images, reports, or other content may be uploaded via file uploader module 830 accessed by case composer module 831 . The case composer module 831 provides functionality such as entering new patients, finding existing patients, or identifying recipients for cases such as expert neurologists and/or data analysis units within the social network infrastructure 837. may The case composer module 831 may further allow metadata to be added, such as content description information (category and description of the uploaded file, etc.). Metadata may be imported directly from a hospital EMR or an EMR such as the cloud-based EMR840. Case content may be organized into folders or directories. Actual case content may also be uploaded directly as images, etc. from a local DICOM server 846 (node) or cloud-based remote image repository. Similarly, case content can be downloaded onto a secure local device 834 for image display, such as in a third party image viewer 836 . As explained above, the network management module provides fully HIPAA compliant (and time-stamped) data in all patient data within the social network infrastructure 837, e.g., which users have accessed a given patient. It may further be possible to generate audit trails.

In some embodiments, the local case management layer 829 may include a locally installed cache server 835 . Caching servers 835, like third-party image viewers 836, may be configured to replicate and/or pre-fetch content from social network infrastructure 837 for faster access. In some embodiments, an SSL connection is established between caching server 835 and case content delivery cloud 839 to allow caching server 835 to access information from social network infrastructure 837. may Similarly, content such as images may be batch uploaded using cache server 835 . Cache server 835 may be a software application installed locally on a local computer within the hospital's firewall 860, or a network appliance such as one found in a data center. Cache server 835 may be an embedded system appliance with flash memory, integrated firewall, and wireless networking capabilities. Alternatively or additionally, cache server 835 may be directly connected with physician interfacing applications and devices 834 such as tablets and wearable computing devices. Wearable computing devices may include, for example, eyeglasses with display capabilities.

In some embodiments, cache server 835 encrypts or decrypts data stored on cache server 835 before sending the data to case content delivery cloud 839 or after receiving encrypted data therefrom. may be further configured to In these and other embodiments, properly encrypted data is stored in case content delivery cloud 839 and clouds hosted, for example, in public clouds.

In some embodiments, local case management layer 829 may be configured to communicate with enterprise plug-in layer 850 . Specifically, the case composer module 831 is configured to provide interoperability between the local case management layer 829 and other local healthcare systems within the healthcare provider's firewall 860. may be configured to communicate with, or include, a plug-in 841 capable of Other local healthcare systems may include EMRs 840, PACS 846, or other systems 842 that support the HL7 messaging protocol, such as Laboratory Information Management Systems (LIMS). Plug-ins 841 may allow local case management layer 829 to import (or export) data from/to the system.

For example, case composer module 831 may connect with local EMR 840 and invoke a plug-in application to import case metadata for a given case. As another example, medical images may be imported into cases through the case composer module 831 after they are received from a local or cloud-based PACS 846 . As another example, laboratory results indicate that they are HL7 v2. may be imported into a case through the case composer module 831 after being received from the LIMS via x messaging. In some embodiments, case composer module 831 may use plug-ins 841 to access information from a personal health record (PHR) such as HealthVault. For example, the case composer module 831 also uses plug-ins 841 to provide patient demographics, insurance information, medications, allergies, and treatment plans, among other information, via continuity of care (CCR) or direct protocols 848. may be imported. In some embodiments, plug-in 841 may be pre-installed on cache server 835 .

In some embodiments, plug-in 841 may be configured to provide information from mobile data capture device 844 to case composer module 831 . Mobile data capture device 844 may use the device used for cognitive screen test 620 . In these and other embodiments, case composer module 831 may connect directly with mobile data capture device 844 to a server device that stores data captured from home screening devices or the like that are connected via the Internet. Similarly, server device 845 may be used to perform quantitative analysis on images or on gene sequence data and to export analysis data to cases using case composer module 831 . Server device 845 may also be a specialized or compact supercomputer for medical image analysis or NGS genome analysis, an NGS genome sequencing desktop device, or a portable USB sequencing device. Server device 845 may also enable automatic backup of data within social network interface 837 and/or may function as cache server 835 as described above.

Local case management layer 829 and social network infrastructure 837 may be further configured to communicate with physician presentation layer 826 . Physicians are typically time constrained and require a different type of presentation layer than administrative personnel such as technicians or physician assistants. Various systems and applications may be utilized to optimize the delivery of information to busy physicians. For example, a mobile phone messaging application 820 or an integrated report 822 delivered through an application on a mobile device such as a tablet or "phablet" may be used. Alternatively or additionally, a wearable display device such as an intelligent wristband/watch that can wirelessly connect with the patient's wristband/watch (e.g., with a built-in dementia self-test or continuous monitoring application) , invokes presentation of the patient's integrated report within an intelligent wristband/watch. In this way, the physician can receive information about the patient as the patient is seen by the physician.

In some embodiments, the physician presentation layer 826 can use touch, voice, or other natural user interface (NUI) as input and interactions adapted for personalized healthcare (PHC). a target dashboard 824 . The interactive dashboard 824 may be displayed on a flexible or disposable display device contained within pharmaceutical packaging. Alternatively or additionally, the interactive dashboard 824 is displayed on a wall-mounted TV that can be wirelessly connected to the patient's wearable computer wristband/watch, thereby providing patient information to the physician. You may call the offer. The interactive dashboard 824 may also be projected by a wirelessly connected projector device onto the walls of the procedure room or doctor's room.

Image viewing capabilities, such as from a third party image viewer 836, such as a DICOM viewer, may be integrated into dashboards, integrated reports, or run as browser-based or standalone applications. In these and other embodiments, images may be downloaded to a device such as device 834 or imported into a local or remote PACS system 846 for image review purposes. In some embodiments, the image display function may not require the actual image to be downloaded to the device. In these and other embodiments, the application component may transmit rendered pixel data on the device display, while the actual image files are stored on a local server, such as cache server 835 or within social network infrastructure 837. It resides on a remote server, such as a server. In some embodiments, the physician presentation layer 826 may utilize HTML5 web standards, Java, or desktop virtualization technology to render images without downloading the images.

Physician presentation layer 826 , social network infrastructure 837 and local case management layer 829 may be configured to communicate with anonymizer, data aggregation and analysis layer 838 . The anonymizer, data aggregation and analysis layer 838 may include one or more subsystems. One or more subsystems provide the functionality to de-identify patient data and store such de-identified data in a centralized or federated repository for further analysis, such as advanced "big data" analysis. It provides. For example, the anonymizer, data aggregation, and analysis layer 838 may include an anonymizer subsystem, a data aggregation subsystem, and/or an analysis subsystem.

In some embodiments, the anonymizer subsystem that performs patient data de-identification may be a server component installed within the local case management layer 829 . The data aggregation subsystem may consist of a large and/or distributed database and file storage system installed on-premises or at external facilities of a cloud service provider such as Amazon, for example.

The analysis subsystem may consist of a server-based system, such as a compute node or multiple thereof, configured to run application software for automated quantification of images, such as hippocampal volumetric measurements. The analytics subsystem may also be a cloud-based, e.g., private cloud-based, computing resource, and may be connected to the social network infrastructure 837 or other third party infrastructure as described herein. good. The analytics subsystem may consist of a highly automated analytics system and cloud computing resources such as Amazon EC2 installed in a data center or a "data center in the cloud" as described above. . A highly automated analysis subsystem may perform analysis on the aggregated data and may be proximate to them, eg, at the same facility.

The system 800 illustrated in FIG. 8 may be used for electronic distribution of information in personalized medicine and Alzheimer's disease (AD) diagnosis. The system 800 captures multiple patient data streams and then enables electronic distribution of information by utilizing a scalable cloud-based social network architecture to route such information among various healthcare providers. may be implemented. The system also includes a data analysis unit (a fully automated "big data analysis" system component) that analyzes the data and presents the information in extracted and "curated" form to the prescriber (e.g. on a mobile device). etc.), including System 800 is applicable to any personalized medicine application and therapy area, such as neurodegenerative diseases, multiple sclerosis, and cancer, among others.

The system 800 is particularly well suited, without limitation, in the context of biologic drug therapy that requires personalization with respect to risk/benefit and economic considerations. System 800 is also well suited for connecting to mobile/wireless sensors that acquire real-time and continuous data streams. The system 800 can be used for imaging and/or next generation sequencing data that require data analysis prior to using the image and/or next generation sequencing data for personalized healthcare in real-world clinical applications outside of academic research. It is well suited for using sing data. System 800 may further be implemented in computer software, or hardware circuitry, or any combination of software and hardware components, and is not limited to any particular software or hardware implementation.

FIG. 8 illustrates some of the major components of the system described herein and the data flow for providing personalized medicine. In some embodiments, the service provider/operator provides social Components 837, 839, 838, 820, 822, 824, 830, 831, 835, 831, 841, 844, and 845 of system 800 and services provided to certain users and/or members of network infrastructure 837 and devices may be implemented. Members may include physicians, administrative personnel, e.g., physician assistants or nurses, and e.g. , 841, 844, 845 and to securely upload protected health information from hospital or external repositories 840, 842, 846, 848, certain devices and software 826, 834, 832, 860 may control and own the

In some embodiments, the case content delivery cloud 839 component of the social network infrastructure 837 includes AJAX, LAMP stacks, Java, Javascript, XML/XSLT, Python, web application frameworks like Ruby, ASP. NET and/or other proprietary frameworks, server libraries, and GUI components that enable rapid development of custom and AJAX-enabled cross-browser applications. It may be implemented as a base system. Web services may include open source SMART APIs for driving interactive reports (interactive reports) 822 and interactive dashboards 824; RESTful APIs; or third party natural interface (NUI) devices, such as , via an API for driving augmented reality glasses. In some embodiments, the native application resides within the physician presentation layer 826 or local case management layer 829 (such as on device 834) that securely connects to the backend components of the case content delivery cloud 839 system over the Internet. , may be installed. Such security may be implemented for secure Internet communications via 256-bit, 512-bit, 1024-bit AES SSL, or higher bit encryption, or other protocols such as Transport Layer Protocol (TSL) may include security by implementing

In some embodiments, native mobile applications within physician presentation layer 826 may be developed using mobile development tools such as, for example, Android or iOS SDKs, Corona SDKs, Sencha, or Unity, among others. Other development tools include some native interactive charting/reporting functionality for integrated reports 822 and interactive dashboards 824 as described above, and wearable computing It may also be used to implement third party SDKs for the device. Other development tools may also be used to implement gesture-based controllers, for example other types of NUI devices such as Kinect, or LeapMotion, among others.

Access control to applications within the social network infrastructure 837 and/or the physician presentation layer 826 can be implemented, for example, by web-based login pages (with encrypted password submission), single sign-on (SSO) approaches (LDAP, Active Directory, OpenSSO, etc.), or may be implemented using biometric SSO. Biometric authentication uses one or more biometric factors, such as gestures, hand geometry, EEG, eye-tracking, retinal signatures, fingerprints, voice, facial recognition, and/or other biometric characteristics captured from access devices. may further include: Access control may also be implemented using certain SDKs/APIs or other electronic access control approaches, such as card-based approaches, or smart phones with electronic authentication.

In some embodiments, case composer module 831 may be embodied in a web-based application such as Javascript. In some embodiments, file uploader module 830 may be implemented in Java or Flash. In some embodiments, the front-end (client) application on the device (such as on device 834) in the local case management layer 829 or desktop computer in the physician presentation layer 826 is Visual Studio, Xcode, Eclipse, and other software. It may also be implemented using integrated development environments (IDEs) such as the development environment. DICOM representations may be further integrated using OsiriX.

Various interfaces for bidirectional (import/export) interoperability with other systems, e.g., hospital, external, or locally installed equipment/appliances, or other embodiments within the enterprise plug-in layer 850 Plug-ins 841 may be embodied as native desktop or server applications or Java applications. Alternatively or additionally, various plug-ins 841 may be implemented using an IDE, such as the IDEs mentioned above. An open source interface engine, e.g. Mirth Connect and/or SMART API, to implement said plugins into hospital or external repositories such as HL7, CDA, DICOM and/or from PHRs using the CCR/Direct protocol may be further used for Various plug-ins 841 are also embedded and pre-installed in the equipment mentioned above, such as NGS sequencing devices, specialized devices that support local quantitative medical image analysis and backup to social network infrastructure 837. good too. Various plug-ins 841 may also be embedded and pre-installed on cache server 835 .

In some embodiments, cache server 835 may be implemented as a software application and installed on a local computer within the hospital's firewall or on a server residing in a data center. Cache server 835 may further be implemented in embedded system code, eg, running under an Embedded Linux distribution (eg, OpenWrt) or a real-time operating system (RTOS), eg, VxWorks or Neutrino. In some embodiments, cache server application code may be implemented using IDEs such as Eclipse, Tornado, QNX, Visual Studio, Xcode, and other software development tools. Cache server 835 further encrypts or decrypts data stored on cache server 835 before sending data to case content delivery cloud 839 or after receiving encrypted data from case content delivery cloud 839. may In these and other embodiments, encrypted data is stored in case content delivery cloud 839 and clouds hosted, for example, in public clouds. Encryption may be implemented, for example, via the FUSE-based EncFS encrypted file system in Linux or TrueCrypt for other operating systems.

Cache server 835 may, in some embodiments, have a very compact form factor and may be designed for small clinics. For example, cache server 835 may include a turnkey embedded system appliance with flash memory and wireless networking capabilities such as Bluetooth, NFC or Wi-Fi. In these and other embodiments, cache server 835 may further incorporate integrated firewall and network intrusion detection capabilities.

In some embodiments, cache server 835 wirelessly connects with physician interface applications and devices within physician presentation layer 826, such as tablets and wearable computing devices, e.g., smart glasses in an integrated display. and/or may have circuitry for driving other types of natural user interfaces, such as gesture control interfaces. Cache server 835 may be implemented using an RTOS, such as VxWorks, or a secure embedded Linux distribution to ensure high stability and security.

An anonymizer, data aggregation and analysis layer 838 may be embodied in the data analysis unit to de-identify patient data and focus the de-identified data for further analysis, as described above. functionality for storing in a federated repository. System 800 may include one or more of the data analysis units. In some embodiments, the data analysis unit is tailored to certain customer needs, such as proprietary reference data from pharmaceutical companies, health systems, payers, or public and private partnerships. There may be. The adapted data analysis unit may for example be executed in a private cloud context.

The anonymizer subsystem within the data analysis unit may be a server component installed within the firewall of the data analysis unit. Anonymizers may be implemented using third party or open source data anonymization software tools such as XNAT for DICOM data, Mirth Connect for HL7 data, and/or protected data such as patient identification information. A custom Python shell script may be written to pull off the health information.

In some embodiments, the data format of the data in the data analysis unit may be standardized before further analysis into quantitative information in favor of, for example, qualitative data. In these and other embodiments, data curation may be fully automated for standardized data types. In some embodiments, quality control and anonymization of data imported into case content delivery cloud 839 is controlled by one or more of various plug-ins 841 such as mobile data capture device 844 and server device 845 plug-ins. It may be further embodied in a plug-in. For example, these plugins may be developed with open source software components such as RSNA CTP. Plug-ins with quality control (such as for correcting image acquisition parameters in DICOM headers) and anonymization functions may be further complemented by a Python shell script implemented on the server in the data analysis unit, which analyzes the image data for defects. may be checked and defective images rejected prior to storage/aggregation of such data.

Non-image and/or non-sequencing data can be stored in SQL (such as MySQL) or NoSQL databases (such as Cassandra, MongoDB, or Hbase) and/or Hive. On the other hand, media-rich content such as images or NGS source data may be stored in a file system, such as distributed HDFS, for performance reasons. The case content delivery cloud 839 may support semantic data integration to correlate data from other databases and datasets, eg, ADNI or Connectome DB. Image files may also reside in external cloud image repositories that are performance optimized and referenced by URL links or other pointers stored in case content delivery cloud 839 or via APIs such as RESTful APIs. good.

The analysis subsystem within the data analysis unit may be embodied in a server-based system within the firewall of the data analysis unit, such as a compute node or nodes thereof. The analysis subsystem performs automatic quantification of images, e.g., hippocampal volumetric measurements and/or other brain structure quantification using MRI, voxel-based amyloid PET quantification, texture analysis of MRI scans, or DTI fiber tracking. based brain "connectome" analysis, and fMRI. Automatic quantification may be implemented, for example, in MATLAB computer code and may be executable as part of the automatic quantification server code or compiled as a C/C++ shared library. In some embodiments, the server code for the analysis subsystem may also be parallelized. In some embodiments, the automatic quantification code within the analysis subsystem may further exhibit certain numerical instabilities in computation, which may be caused by computer chip high temperatures, cosmic radiation, moisture, and/or manufacturing defects. caused. The automatic quantification code incorporates external real-time data such as sensor data in the computation and/or includes recomputation on a separate computer chip to flag computational errors based on chip manufacturing defects. This numerical instability may be further corrected by Such computational errors are caused by performing random check operations on image data in an externally referenced dataset, such as ADNI or other database, and by combining random check operations from externally referenced datasets with manually tracked volumes in that database. It may be further detected by comparison. Automatic quantification code may additionally store a physical hardware signature of the compute node on which the code is executing. For example, dmidecode in Linux can be used to get detailed information of the chip used in performing the operation, and then said hardware signature is a potential Can be compared against external hardware reference data and given said external sensor data to flag computational errors, then compensated by performing recomputations at a later time/on different hardware can enable

In some embodiments, the analysis subsystem may be configured to perform connectome analysis, which may be based on data from ultra-high resolution DTI MRI or resting state fMRI.

In some embodiments, the analysis subsystem may further include so-called "social network analysis" (SNA) tools, which may be implemented using open source tools such as Cytoscape and R tools for SNA. The SNA tool may be utilized to perform brain connectome analysis in one particular embodiment. In another embodiment, SNA tools may be used to perform analysis on social network infrastructure 837 . For example, the SNA tool may perform analysis on the social network infrastructure 837 to discover user-related (interaction) patterns of patterns among network participants and improve the services provided to the participants. good. Alternatively or additionally, the SNA tool may perform analytics on the social network infrastructure 837 to provide payers with insights to optimize the delivery of care in a cost-effective manner. The SNA tool for brain connectome analysis may, in yet another embodiment, incorporate other graph theory analysis tools such as the Brain Connectivity Toolbox (BCT) or MatlabBGL. The Brain Connectivity Toolbox (BCT) may be used, for example, to compute the "small world network" index and characteristics of patients with suspected Alzheimer's disease. The Brain Connectivity Toolbox (BCT), in another embodiment, is used to assist the data scientist or physician using an interface at the Physician Presentation Layer 826, such as the Brainnet Viewer, Connectome Viewer Toolkit, or Topology Graph Theory Visualization Tool. May be used in visualization.

In some embodiments, the analysis subsystem may execute application software to perform automated "big data" analysis, as described above in another embodiment. Automated analysis may be performed on aggregated data for multiple patients, eg, multiple patients, as analysis on patient data. The analytics subsystem may be embodied in a data analytics unit or cloud computing resource, such as Amazon EC2, Google Compute Engine, Azure, or a data analytics unit within the social network infrastructure 837 .

In some embodiments, the analysis subsystem may also be proximate to the aggregated data, eg, within the same facility. In some embodiments, the analysis subsystem may provide in-memory analysis with large (eg, terabyte-class or larger) data being analyzed residing in memory. In-memory analysis can reduce latency and provide immediate analysis. In some embodiments, the analysis subsystem may also be accessed by physician applications and devices 826, 836, as described above. In these and other embodiments, a first analysis server, eg, an in-memory analysis server, may be used to render certain interface components to a physician or data scientist in physician presentation layer 826 . A second web server may be used for interface components that may be rendered. Alternatively or additionally, desktop virtualization technology may be used to render the interface components (eg, Citrix, or an open source alternative).

In some embodiments, the analysis system uses machine learning or other data mining techniques such as natural language processing, neural networks, Support Vector Machines (SVM), and statistical classifiers such as k-Nearest Neighbor (k -NN), or Linear Discriminant Analysis (LDA). Data mining techniques may be implemented, for example, using NLTK, R, MATLAB, or PLINK, or by using the highly extensible Apache Mahout machine learning library built on top of the Hadoop system. good.

In some embodiments, a physician calculates the probability of current disease given a combination, among others, to predict treatment response, future onset of disease, or in presymptomatic Alzheimer's disease (AD) patients. You may request an analysis for Analysis in these and other embodiments may be based on a combination of patient biomarker patterns. Machine learning algorithms, such as supervised neural networks or SVM classifiers, may be learned on aggregated data from multiple other patients. The learned algorithm may then be applied to the patient's biomarker combinations. A trained algorithm estimates a Bayesian posterior probability (Bayesian posterior probability) of disease at a given set of biomarkers based on known transformations to AD status from the aggregated longitudinal data. may be further trained to estimate the probability of

A physician may also invoke further analysis through the physician presentation layer 826 . Further analysis includes calculating the predictive power of add-on markers, e.g., by calculating the area under the receiver operating characteristic (ROC) curve or the AUC value of the proposed biomarker combination from the aggregated data; Identification of useful add-on markers may also be included. Identification of useful add-on markers may be performed by running simulations adding default (eg, mean) values for new biomarkers. Based on the simulations, the analysis system may compute modified Bayesian posterior probabilities (Bayesian posterior probabilities) using add-on markers to estimate potential gains in diagnostic certainty. Certain statistical classifiers may also use a priori (prior) probabilities of disease as model inputs, which, in another embodiment, may be useful when these variables are not directly included as input variables. It can be estimated based on risk factors such as risk factor and age and epidemiological data.

FIG. 9 is a diagram illustrating an exemplary system 900 for personalized medicine, arranged in accordance with at least some embodiments described herein and described herein. Disposed according to at least some embodiments. The system 900 may include a social network 910, a data system 920, and a natural user interface (NUI) unit 940 communicatively coupled and capable of exchanging information therebetween.

Social network 910 may be configured similarly to social network 100 of FIG. Alternatively or additionally, social network 910 may be similar to social network infrastructure 837 of FIG. In some embodiments, social network 910 may be configured to facilitate interaction between multiple participants regarding a patient's medical diagnosis. In some embodiments, data system 920 may be included within social network 910 or external to social network 910 . Social network 910 may store patient data such as biomarkers from images and other tests performed by one or more patients. In some embodiments, patient data may be imported via the natural user interface unit 940 or some other interface.

Natural user interface unit 940 may be similar to physician presentation layer 826 of FIG. 8 and may be configured to send data to and receive data from social network 910 and data system 920. . Additionally, the natural user interface unit 940 may be configured to interact with a physician associated with the patient. A natural user interface unit 940 may receive commands from the physician and present information to the physician. In some embodiments, the natural user interface may be further configured to instruct the data system to analyze the data based on instructions from the physician.

Data system 920 may include various units, including data anonymizer unit 922 , data aggregation unit 924 , data analysis unit 926 , and reporting unit 930 . In some embodiments, data system 920 may be similar to the data analysis units described herein.

The data anonymizer unit 922 may be configured to receive patient data associated with a patient's medical diagnosis and to retrieve at least a portion of patient identifying information from the received patient data to generate first anonymized data. good. Data anonymizer unit 922 may be similar to the anonymizer subsystem described with respect to FIG. In some embodiments, patient data being transmitted to social network 910 via natural user interface unit 940 may first be received by data anonymizer unit 922, and a portion of the patient-identifying information may be extracted. good. In some embodiments, patient data in social network 910 may be sent to data anonymizer unit 922 and patient identifying information may be retrieved from the patient data. Patient data may then be transmitted back to social network 910 .

The data aggregation unit 924 may be configured to store de-identified patient data and store de-identified data of at least one other patient.

Data analysis unit 926 may be configured to analyze patient data from within social network 910 , specifically using other patient data from data aggregation unit 924 . For example, in some embodiments patient data may include biomarkers. Information about the patient may be determined by analyzing the biomarkers of the patient data in relation to biomarkers from other patients or from healthy individuals. The information may relate to diagnosis of disease, therapy of disease, and/or progression of a diagnosed disease, among other things. In some embodiments, data analysis unit 926 may be configured to curate patient data and/or information determined by data analysis unit 926 . Data anonymizer unit 922 may be similar to the analysis subsystem described with respect to FIG.

Reporting unit 930 may be configured to receive information and patient anonymized data to generate a report that may be presented to a physician through natural user interface unit 940 .

Modifications, additions, or omissions may be made to system 900 without departing from the disclosure of the present disclosure. For example, system 900 may include modules, units, or systems other than those described above. Further, social network 910 may include various aspects other than those described above.

FIG. 10 is an exemplary data flow of a method 1000 of generating reports for presentation, arranged in accordance with at least some embodiments described herein. For example, in method 1000, an integrated screening report may be generated from low-cost tests for early Alzheimer's disease (AD), such as prodromal AD. A report may consist of the patient completing a low-cost test, such as a cognitive test on a mobile device, at the patient's home or at the doctor's office. In some embodiments, cognitive tests may be conducted using a web-based application, such as an HTML5 web application, or using a native application on a mobile device or some other computing device.

After completing the cognitive tests, additional tests may be ordered and/or taken by the patient when the patient receives a score above a threshold indicative of the possibility of AD. For example, the patient may undergo genetic testing for APOE genotype and/or testing for certain genetic variations such as single nucleotide polymorphisms (SNPs). SNPs may further be selected from lists generated by quantitative trait locus genome wide association analysis (QTL GWAS). Alternatively or additionally, the patient may undergo an in vitro diagnostic (IVD) screening test, such as a blood test, or an eye-based screening test configured to detect the presence of intraocular deposits of amyloid beta peptide. . Patients may also undergo other tests, such as new screening tests, which may include continuous wireless monitoring with gait sensors, eye tracking, or wireless sleep monitors such as accelerometers and/or electroencephalogram (EEG) sensors. .

In some embodiments, the patient or the patient's personal physician may order additional tests. In some embodiments, additional tests may be ordered directly through the cognitive test application. For example, the patient may press a prompt button on a web-based application or receive a coupon obtained from within the application. In some embodiments, additional testing may be performed by a personal genomics service online store such as 23andMe. Alternatively, the patient already has an account with a personal genomics service, logs into that account, retrieves the patient's genetic data through an application programming interface, and transfers the data to the case content delivery cloud of FIG. 839 or other social network cloud.

Data from additional tests may be captured in the social network cloud and combined with data from cognitive tests. Combined data from patients may be analyzed and curated in a brief summary form as described above. For example, the combined data may be curated into a report, such as an integrated screening report. A consolidated screening report may suggest a comprehensive diagnostic evaluation. Alternatively or additionally, the integrated screening report may not suggest a comprehensive diagnostic assessment. Whether the combined screening report advocates an overall diagnostic assessment may be based on whether the combined data raises the probability of AD above a threshold. In some embodiments, the combined data may be compared or analyzed in relation to other data from other patients for possible diagnosis of AD. This analysis can help in determining when the combined data may indicate the possibility of AD.

If the report indicates the possibility of AD or indicates that a comprehensive diagnostic evaluation should be performed, the patient's primary care physician may send/share with the specialist on social networks.

FIG. 11 is an example dataflow of another method 1100 of generating a report for presentation, arranged in accordance with at least some embodiments described herein. In some embodiments, the method 1100 may begin with the specialist receiving a report that a screening diagnostic evaluation will be performed based on data previously collected by the attending physician. In some embodiments, the report received by the specialist may be similar to the report received by the specialist in method 1000 .

Specifically, method 1100 is configured to generate a consolidated "baseline" diagnostic report (baseline diagnostic report). A baseline diagnostic report may be a report containing baseline data for a patient. Baseline data may be data collected from a patient and used to determine the patient's health in comparison to the patient's future data. The baseline data determines baseline health characteristics of the patient.

The specialist may order additional tests for the patient, such as an MRI and/or a spinal tap and cerebrospinal fluid (CSF) assay for tau protein or amyloid beta peptide. Additional tests on the patient, such as a spinal tap and cerebrospinal fluid (CSF) assay, may be diagnostic markers of neurodegeneration and amyloid accumulation, respectively. Data from additional exams may be uploaded to the social network cloud and may or may not be combined with previous data collected about the patient. Data analysis, such as automated quantitative MRI analysis, or central quality-controlled expert scan readings, may be performed on the combined data. The combined data and/or results from the analysis may be curated in brief summary reports (reports), as described above. In some embodiments, the report may be a consolidated baseline diagnostic report. The consolidated baseline diagnostic report may be shared with another physician in the social network, eg, the patient's primary physician. In some embodiments, the specialist further consults with a data interpretation expert, such as in a call center, via voice, video, or other messaging in a social network, regarding the data in the integrated report. You may

FIG. 12 is an exemplary data flow of another method 1200 for generating reports, arranged in accordance with at least some embodiments described herein, and according to at least some embodiments. For example, in method 1200, an integrated "companion diagnostic" report (companion diagnostic report) may be generated from a subset of the patient's diagnostic assessments. Diagnostic assessment may include, for example, MRI-based hippocampal volume quantification. Diagnostic assessment may also include other biomarkers such as cerebrospinal fluid (CSF) assays. Results from diagnostic assessments may be captured and inserted into the social network's cloud. Diagnostic assessments, once in the cloud, may be analyzed. Analyzes such as stratification based on machine learning or statistical algorithms as described above can be applied to diagnostic information from patients and implemented in the form of integrated companion diagnostic reports for pharmaceutical prescribers. may be summarized in specific personalized treatment information. A medicament may include, for example, a disease-modifying drug for the treatment of early AD in the pre-disease stage. The integrated companion diagnostic report will calculate the values of the individual components of the diagnosis as well as the label information of the approved personalized medicines that will be used with the combination biomarkers as a "companion diagnostic". A combined score may be presented. Medication may be directed directly from a device such as a tablet or wearable computer presenting an integrated companion diagnostic report. The integrated companion diagnostic report may also be shared with other physicians in a social network, such as specialist physicians.

FIG. 13 is an exemplary dataflow of another method 1300 of generating a report for presentation, arranged in accordance with at least some embodiments described herein. In some embodiments, the method 1300 may be initiated by a specialist who receives a report indicating a need for an integrated companion diagnostic that includes actionable, personalized treatment information for the patient. In some embodiments, the report received by the specialist may be similar to the report received by the specialist in method 1200 .

Specifically, method 1300 may be configured to generate integrated longitudinal reports for safety and efficacy monitoring. Reports may be presented to prescribers of disease-modifying AD medications or other AD therapies. Long-term monitoring over time includes determining the effect of therapy on the patient according to the treatment prescribed for the patient and whether therapy is associated with microhemorrhage or vasogenic edema (referred to as ARIA-H or ARIA-E). and determining whether it has any adverse effects on the patient, such as.

Longitudinal monitoring may involve a specialist performing and/or ordering additional examinations of the patient over a period of time. Testing may include CSF assays or PET scans for efficacy monitoring such as measuring levels of amyloid beta in CSF or amyloid burden in the brain. Data analysis may be performed on data with centralized quality control, such as automated quantitative image analysis or safe read MRIs, such as ARIA-H or ARIA-E. Based on the data analysis, when potential safety concerns of drug therapy are detected, then an alert may be triggered. The alert may trigger messaging within the social network to indicate to the specialist and other participants within the social network that the alert has been triggered.

Longitudinal profiles from patients are also presented in a concisely summarized form as described above in the Integrated Longitudinal Efficacy/Safety Profile Report (Efficacy/Safety Profile Report) for therapy monitoring. profile report). In some embodiments, the integrated long-term efficacy/safety profile reporting includes social May be shared among participants in the network. In some embodiments, a subject paying for a patient's medical expenses may pay for aspects of the patient's treatment based on the success of the treatment within a reasonable risk/benefit ratio. A patient prescriber payer may communicate with another physician expert, e.g., in a call center, via voice, video, or other messaging in a social network regarding integrated efficacy/safety profile data. You can consult further via

In some embodiments, a physician prescribing treatment for a patient may switch treatment based on the patient's response to the treatment. In some embodiments, the subject paying for a patient's medical expenses may mandate a change in treatment based on a review of the data.

The data analyzed to generate an integrated longitudinal efficacy/safety profile report may include other information about the patient. For example, analyzed data may include patient individual characteristics including next generation genome sequencing information. Other patient characteristics collected from continuous wireless sensing devices such as activity in multiple sclerosis (using mobile accelerometers) or microinvasive mobile blood sampling/analysis in cancer therapy are novel biomarkers. may be incorporated to capture In some embodiments, therapy monitoring may include therapy monitoring of personalized drug therapy, such as antibody therapy, which may be applied in other diseases such as multiple sclerosis and cancer.

In some embodiments, data from patients may be anonymized and aggregated with anonymized data from other patients. Anonymized and aggregated data may be used to generate pre- or post-approval data and to demonstrate real-world evidence of favorable risk/benefit ratios, e.g. for reimbursement purposes. , may be further utilized by pharmaceutical companies.

FIG. 14 is an example dataflow of another method 1400 of generating a report for presentation, arranged in accordance with at least some embodiments described herein. The method 1400 may be configured to generate a predictive analysis report (predictive analysis report) based on the analysis of individual biomarkers or the integration of multiple biomarkers of the patient. A predictive analysis report may be generated for the physician, the patient's payee, or the patient before the patient receives a disease-modifying AD medication or other AD therapy. Prognostic reporting may predict individual response to specific preventive strategies, such as prescribing biologic drugs, in the early stages before the onset of AD. Prognostic reporting may further identify at-risk individuals (eg, older than their family members, etc.). Prognostic reporting may further stratify patients at specific stages of the AD disease continuum, such as presymptomatic stages, or earlier, and guide treatments that may be approved for these stages of the disease. .

To generate a predictive analysis report, method 1400 may include a first physician, such as a specialist, ordering a set of predictive biomarker tests for a patient. Prognostic biomarker studies may include MRI scans, PET scans, structural MRI analysis, DTI MRI tractography, brain connectivity map analysis, voxel-based amyloid PET analysis, or other advanced brain imaging studies. The physician may also order lab-based testing such as IVDs or whole genome sequencing, among others. In some embodiments, whole genome sequencing may be performed on a semiconductor-based nanopore sequencing device. Data collected from lab-based tests and predictive biomarker tests may be captured and combined together in a cloud within a social network, eg, social network 100 of FIG.

Automatic data analysis may be performed on the combined data as described above. Automated data analysis may be performed at a data analysis unit within the physician's building or office, at a physical data analysis unit, at a physical supercomputer facility, at a cloud-based on-demand computing resource such as Amazon EC2, or at a dedicated facility in the cloud. may be performed in the data analysis unit of In these and other embodiments, the data analysis unit may be communicatively coupled to the social network to receive combined data from the social network. The data analysis unit may further utilize quantum computers, nodes with integrated quantum chips, or nodes optimized for in-memory analysis such as tera- or petabyte-class memory.

In some embodiments, advanced analysis includes real-time automated quantitative analysis of hippocampal volume or other brain structures, DTI MRI-based fiber tract network analysis, or ultra-high resolution resting state, among other analyses. It may also include fMRI-based brain connectivity map analysis. Advanced analysis may further include NGS genomic analysis, analysis of mass spectrometry data, and/or analysis of biomarker combinations.

Once the analysis has been performed, the physician, such as a specialist, may access advanced analytical capabilities residing in the data analysis unit or computing resources connected to social networks on a tablet or wearable with a natural user interface (NUI). Any computer device may be used. In some embodiments, the physician has a set of useful additional predictive biomarker tests to be performed based on the information already available for the patient and the data present in the aggregated data repository. The NUI may be guided to identify the In some embodiments, the NUI may invoke advanced analysis to be performed against the data repository, eg, to identify such useful additional markers. After identifying additional markers, the NUI may then send the results back to the physician or other participants in the social network. In some embodiments, advanced analysis reports may include data presented in quantitative formats such as probabilities, likelihoods, and/or scores, along with contextual information from the literature reporting predictive analysis results. . In these and other embodiments, physicians may order these predictive tests directly from social networks. As such, they can be performed on data from patients and summarized in predictive analysis reports.

Figures 10-14 illustrate a particular method for generating reports related to Alzheimer's disease. In general, the methods and systems described herein are applicable to any personalized healthcare application and therapy area, such as other neurodegenerative diseases, multiple sclerosis, and cancer. The methods and systems described may also be implemented for applications in diagnosing post-traumatic stress disorder (PTSD) or traumatic brain injury (TBI). PTSD and TBI may have common elements in diagnostic approaches and have been hypothesized to contribute to the subsequent development of AD dementia. Other embodiments may incorporate other imaging modalities such as FDP-PET, molecular imaging such as nanoparticle-based MRI, DTI MRI, ASL MRI. Other non-imaging biomarkers via continuous wireless monitoring and/or “self-tracking” consumer devices, such as by walking sensors, eye-tracking, accelerometers and/or wireless sleep monitors such as EEG, are being explored by the Continua Alliance. /ISO/IEEE 11073 Personal Health Data (PHD) standard.

FIG. 15 is a flow chart of an exemplary method 1500 for distributing information corresponding to personalized health care in clinical, non-research settings, arranged according to at least some embodiments described herein. is set. Method 1500 may be implemented by a system, such as system 800 of FIG. 8, in some embodiments. Although shown as separate blocks, various blocks may be divided into additional blocks, combined into fewer blocks, or deleted depending on the desired implementation.

Method 1500 may begin at block 1502 where one or more data streams may be captured. Each of the data streams may relate to patient care. In some embodiments, one of the data streams may be captured from an application running on a mobile device related to screening for early Alzheimer's disease. In some embodiments, one of the data streams may be captured from a mobile cognitive testing application running on a patient-enabled mobile device. In these and other embodiments, when the mobile cognitive test application indicates a positive diagnosis, another of the data streams is captured from genetic testing directed to the patient from within the mobile cognitive test application. In some embodiments, data streams may come from other patient examination results. In some embodiments, the test results may be test results from the same type of test in which the tests are performed separately over time or over time. In some embodiments, test results may come from a patient, laboratory, specialist, or primary care physician. In some embodiments, the data stream may include images or written text, among other types of data.

At block 1504, the data streams may be aggregated to generate aggregated diagnostic data. In some embodiments, data streams may be consolidated in a cloud environment. In these and other embodiments, data streams may be associated based on the patient in which the data stream occurred. In some embodiments, data streams may be accessed through social networks.

At block 1506, the integrated diagnostic data may be analyzed to generate analyzed diagnostic data. Analysis of the integrated diagnostic data is performed, in some embodiments, by comparing the integrated diagnostic data to the integrated diagnostic data of the patient or one or more other patients that are different from the other data. may be

At block 1508, the analyzed diagnostic data may be curated. Curating analyzed diagnostic data includes presenting one or several biomarkers along with contextual information such as relevant excerpts from medical guidelines and/or medical literature or links to original references. It's okay. Biomarkers are further defined as normative and/or age-related ranges, plots of patient individual values on normative and/or age-related ranges, and medical treatment of patients or other representative illustrative cases. May be presented with the image.

At block 1510, a consolidated report for presentation to the patient's physician may be generated based on the curated and analyzed diagnostic data. In some embodiments, the report may provide information regarding drug therapy to the patient based on the analyzed diagnostic data.

In some embodiments, method 1500 may be implemented in a cloud-based digital health platform for personalized healthcare for Alzheimer's disease. In some embodiments, the patient's medical diagnosis may be related to a baseline diagnosis of Alzheimer's disease, such that the consolidated report is the baseline consolidated report. In some embodiments, the medical diagnosis of the patient is associated with longitudinal monitoring of Alzheimer's disease within the patient such that the integrated report is an integrated longitudinal safety/efficacy monitoring report. good too.

In some embodiments, the patient's medical diagnosis is relevant to therapy monitoring of a disease-modifying multiple sclerosis therapy such that the integrated report is an integrated longitudinal safety/efficacy monitoring report. You may

In some embodiments, the patient's medical diagnosis may relate to the patient's biomarkers, including gene sequence data, and analysis of the integrated diagnostic data is predictive such that the report is a predictive analysis report. Including analytics. In these and other embodiments, the predictive analysis may predict Alzheimer's disease in presymptomatic stages. Alternatively or additionally, predictive analysis may predict a patient's response to a particular therapy.

Tables 1-3 are protective and risk generated by the Quantitative Trait Locus Genome-Wide Association Study (QTL GWAS) and may be used for diagnostic and/or therapeutic use in personalized medicine. A group of single nucleotide polymorphisms (SNPs) containing four novel Alzheimer's disease (AD)-specific disease targets with (risk) properties is shown.

Table 1 is an example of a group of SNPs identified by QTL GWAS using PLINK, where the quantitative trait is a voxel-based amyloid PET quantification that reflects Alzheimer's disease activity. Genotyping was performed using an IlluminaOmni 2.5 microarray. Table 1 presents the results of the association studies and their respective SNP ranking (ID), chromosome (Chr), SNP (related number, also called rsid that identifies the mutation), p-value, number of minor allele carriers (total found Samples n=334) are shown with the mutated allele/base (A1 represents the minor allele and A2 the major allele). Minor alleles can be considered SNPs for purposes of determining SNPs that show a decreased or increased likelihood of disease. The major allele is the common allele that produces symptoms opposite to those produced by the minor allele.

Four novel Alzheimer's disease (AD)-specific disease targets (see Table 3), i. Identified and highlighted within the DYNLL1 gene (Table 1), minor alleles are protective to reduce the likelihood of Alzheimer's disease, or are at risk, or are likely to have Alzheimer's disease is increasing. The following SNPs: rs4891826 (RTTN), rs2030515 (ALCAM) (both of which are protective traits, which are indicated by * for each minor allele as shown in Table 1) , may indicate a reduced likelihood of developing Alzheimer's disease, whereas the presence of these two SNPs indicates no Alzheimer's disease or a low risk of its progression. The presence of two other SNPs, rs7164265 (DMXL2), rs993900 (DYNLL1), indicate Alzheimer's disease risk for each minor allele.

Table 2 shows the same relevant studies, but modified for APOE4 carrier status. In the latter, the ranking was changed to reveal the RTTN mutation minor allele rs4891826 to be the top SNP after correction for APOE4 carrier status. In Tables 1 and 2, bases corresponding to risk or protective (*) traits revealed by QTL GWAS can be found in column A1 for minor alleles. One skilled in the art understands that even though A1 is the minor allele in this QTL GWAS sample, the minor allele (MA) designation depends on the population minor allele frequency (MAF) used for the association study. and the actual base letter replacement may need to be varied depending on the genotyping method and coding/reporting standards used during analysis. For example, the base known to be associated with risk for rs993900 is A, but when reported in the reverse strand orientation it is T.

Figure 16 shows the scientific rationale for four novel targets for diagnostic (Dx) and/or therapeutic (Rx) use in personalized medicine for AD. Each gene/protein and downstream molecular pathway when used in relation to said SNP (e.g. minor allele), rs4891826 (G), rs2030515 (G), rs7164265 (C), and/or rs993900 (A) , has resulted in several therapeutic measures. Therapeutic interventions targeting hippocampal hyperexcitability, hippocampal neurogenesis, neuroinflammation, neural stem cell development, and axonal transport, individually or in combination, are of particular interest.

The four novel Alzheimer's disease-associated mutant alleles (e.g., associated with SNPs with minor alleles), rs4891826 (G), for diagnostic and/or therapeutic use in personalized medicine, A combination of at least one of rs2030515 (G), rs7164265 (C), and/or rs993900 (A), as described in more detail above, is information-enabled in a clinical, non-research setting for early Alzheimer's disease. A computer-implemented method of delivering personalized healthcare, comprising a social network, and can include capturing one or more data streams, each of which is associated with a patient's healthcare. , may be embodied in a method of delivering personalized healthcare. The method may further include integrating the data streams to generate integrated diagnostic data and analyzing the integrated diagnostic data to generate analyzed diagnostic data. The method curates (value-adds) analyzed diagnostic data and generates an integrated report for presentation to a patient's physician based on the curated analyzed diagnostic data. may further include:

In one embodiment, said one having genetic data obtained from genetic testing for one or more specific genetic mutations comprising one or more single nucleotide polymorphism (SNP) minor alleles selected from a group of SNPs. At least one of the one or more data streams is captured. In one aspect, the selected one or more SNPs are the following SNP minor alleles within the individual patient's gene RTTN, ALCAM, DMXL2 or DYNLL1: rs4891826 (G), rs2030515 (G), rs7164265 (C), or at least one of rs993900 (A).

At least one data stream in the computer-implemented method comprises: may include a diagnostic test for the following minor allele (A1), i.e.
rs769449, rs4420638, rs56131196, rs2075650, rs71352238, rs10956245, rs157582, rs283815, rs34095326, rs75627662, rs114798373, rs4689137, rs4698481, rs963281, rs2463471, rs10234008, rs17680408, rs10029820, rs2357394, rs2033296, rs114979482, rs62444137, rs6499632, rs62532372 rs1001029, rs1001026 and/or
rs55867246, rs1930458, rs1999505, rs1427780 rs13071744, rs6576798, rs13222318, rs11006004, rs4357923, rs12589674, rs2174147, rs8009420, rs17815373, rs11161719, rs10515601, rs1874848, rs78323632, rs111851441, rs9920618, rs11740072, rs78252085 rs4845054, rs1086013, rs4508236, rs9458512, rs17147461, rs78341108, rs6866169, rs72681708, rs4912453, rs12120406, rs8056050
may further include one or more of

In one embodiment, the computer-implemented method may be applied to provide personalized therapy, wherein the individual patient is found not to test positive for the rs4891826 (G) SNP ( ie non-carriers), the personalized therapy is an antibody against the BMP-4 protein (anti-BMP-4 antibody) or a functional fragment thereof.

In one embodiment, the computer-implemented method may further comprise predictive analysis, wherein the predictive analysis is based on social network analysis (SNA) of gene/protein molecular networks with respect to the included/selected SNPs. and the personalized therapy is a small molecule oncology kinase inhibitor selected based on the SNA.

In one embodiment, the computer-implemented method may be further applied to provide personalized therapy, wherein the individual patient is found not to test positive for the rs4891826 (G) SNP. When (ie non-carrier), the personalized therapy is a small molecule tumor kinase inhibitor.

In one embodiment, the computer-implemented method may be further applied to provide personalized therapy, wherein the individual patient is found not to test positive for the rs2030515 (G) SNP. When (ie non-carrier), the personalized therapy is an antibody against the ALCAM/CD166 protein (anti-ALCAM/CD166 antibody) or a functional fragment thereof.

In one embodiment, the computer-implemented method may be further applied to provide personalized therapy, wherein the individual patient is found not to test positive for the rs2030515 (G) SNP. (ie non-carrier), the personalized therapy is an mTOR inhibitor.

In one embodiment, the computer-implemented method may be further applied to provide personalized therapy, wherein the individual patient has: If the test is found to be non-positive (ie, non-carrier), the personalized therapy is combination therapy with an anti-BMP-4 antibody or functional fragment thereof and an mTOR inhibitor.

In one embodiment, the computer-implemented method may be further applied to provide personalized therapy, wherein the personalized therapy comprises: rs4891826 (G), rs2030515 (G), rs7164265 ( C), or rs993900 (A) CRISPR-based gene editing or trans-epigenetic regulation within the RTTN, ALCAM, DMXL2, DYNLL1 genes based on testing for SNPs with the following minor alleles (A1):
rs769449, rs4420638, rs56131196, rs2075650, rs71352238, rs10956245, rs157582, rs283815, rs34095326, rs75627662, rs114798373, rs4689137, rs4698481, rs963281, rs2463471, rs10234008, rs17680408, rs10029820, rs2357394, rs2033296, rs114979482, rs62444137, rs6499632, rs62532372 rs1001029, rs1001026 and/or
rs55867246, rs1930458, rs1999505, rs1427780 rs13071744, rs6576798, rs13222318, rs11006004, rs4357923, rs12589674, rs2174147, rs8009420, rs17815373, rs11161719, rs10515601, rs1874848, rs78323632, rs111851441, rs9920618, rs11740072, rs78252085 rs4845054, rs1086013, rs4508236, rs9458512, rs17147461, rs78341108, rs6866169, rs72681708, rs4912453, rs12120406, rs8056050
may further include one or more of

Other nucleolytic enzyme systems such as ZFNs or TALENs may be used for gene editing. Gene editing is a minimally invasive surgical procedure, e.g., image-guided (such as MRI-guided) delivery directly to targeted brain tissues commonly affected by Alzheimer's disease, such as the hippocampus or anterior femur. can be applied to an individual patient's human neural stem cells, which can be delivered to an individual patient's brain using . Image-guided delivery of gene-edited, individual patient neural stem cell-like cells was performed in the same session by hippocampal quantification, voxel-based quantitation, and magnetic resonance imaging (MRI). It can be combined with the imaging quantitation described above, such as texture analysis of scans, or brain connectome analysis based on diffusion tensor imaging (DTI) fiber tracking. Image quantification can also be used for targeted delivery of gene-edited cells, eg, by localizing disease activity. Neural stem cells can also be delivered by injection into the spinal fluid space of an individual patient.

In one embodiment, said gene editing can be performed by Lab 718 (FIG. 7), which is part of a social network for personalized medicine, said rs4891826 (G), rs2030515 (G) , rs7164265 (C), or rs993900 (A) SNPs. Testing for SNPs can be done by the same lab that performs gene editing, or other labs connected to social networks for personalized medicine. Gene editing may be performed in an automated and multiplexed fashion on an integrated device and/or may be integrated into a next-generation genome sequencing device, enabling personalized medicine as described above. Combined with social networks for.

In one embodiment, the computer-implemented method may be further applied to provide personalized therapy, wherein the individual patient does not test positive for the protective rs2030515 (G) SNP. When found (ie non-carrier), the personalized therapy is antisense oligonucleotide (ASO) or short interfering RNA (siRNA) therapy targeting the ALCAM/CD166 gene.

Here, said gene editing may further target the individual patient's ALCAM gene such that the corrected gene-edited stem cells have two copies of the protective rs2030515 mutation (i.e. are homozygous). , individual patients may particularly benefit from the effects of the protective rs2030515 mutation.

In one embodiment, the analysis of at least one of the four novel Alzheimer's disease-associated mutant alleles rs4891826 (G), rs2030515 (G), rs7164265 (C), or rs993900 (A) is personalized embodied in a method for a SNP or combination of SNPs in a human subject for diagnostic and/or therapeutic use in medicine, the method comprising:
a) obtaining a nucleic acid sample from said human subject;
b) genotyping a sample for a combination of the genes RTTN, ALCAM/CD166, DMXL2, or DYNLL1 associated with Alzheimer's disease, wherein the alleles of said combination of SNPs are rs4891826 (G), rs2030515 (G), examining the genotype, which is rs7164265 (C), or rs993900 (A);
c) detecting the presence of at least one of G at rs4891826, G at rs2030515, C at rs7164265, and A at rs993900 in said nucleic acid sample.

In one embodiment, the combination is rs4891826 (G) and rs2030515 (G), which indicates that the human subject has a protective trait that inhibits the onset or progression of Alzheimer's disease.

In one embodiment, the combination is rs7164265 (C) and rs993900 (A), which indicates that the human subject is at risk for developing or developing Alzheimer's disease.

In one embodiment, the method includes testing for combinations of SNPs, wherein the presence of any one of the SNPs or combination of SNPs is indicative of a disease state, such as Alzheimer's disease. Thus, it can be tested whether a combination of at least 2 or 3 of the 4 SNPs are present in an individual patient's genome. Thus, the presence of at least one SNP from a group of SNPs can be used for the computational methods of personalized therapy described herein.

In one embodiment, a method for detecting at least one single nucleotide polymorphism (SNP) in a human subject from analysis of SNP combinations comprises: a) obtaining a nucleic acid sample from said human subject; To genotype a sample of a combination of disease-associated genes RTTN, ALCAM/CD166, DMXL2, and DYNLL1, wherein the alleles of said combination of SNPs are rs4891826 (G), rs2030515 (G), rs7164265 (C ), and rs993900 (A); and c) detecting the presence of at least one of rs4891826 G, rs2030515 G, rs7164265 C, or rs993900 A in said nucleic acid sample. and In one aspect, the analysis can be any test, protocol, test, sequencing, or genotyping of any kind, such as NGS, microarray, PCR, or the like.

Those skilled in the art will appreciate that with respect to the processes and other processes and methods disclosed herein, the functions performed in the processes and methods may be implemented in different orders. Additionally, the outlined steps and operations are provided as examples only, and some of the steps and operations may be optional or fewer steps may be performed without detracting from the essence of the disclosed embodiments. and operations or expanded to additional steps and operations.

Embodiments described herein may involve the use of special purpose or general purpose computers, including various computer hardware or software modules, as described in more detail below.

Embodiments described herein can be implemented using computer-readable media for carrying or having computer-executable instructions or data structures stored thereon. Such computer-readable media can be any available media that can be accessed by a general purpose or special purpose computer. By way of example, and not limitation, such computer readable media may include tangible (non-transitory) computer media including RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage, or other magnetic storage devices. A readable storage medium or any other storage medium that can be used for carrying or storing desired program code in the form of computer-executable instructions or data structures and that can be accessed by a general purpose or special purpose computer. may contain. Combinations of the above can also be included within the scope of computer-readable media.

Computer-executable instructions include, for example, instructions and data that cause a general purpose computer, special purpose computer, or special purpose processing device to perform a function or group of functions. Although subject matter has been described in terms specific to structural features and/or methodological acts, the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Please understand. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

As used herein, the terms "module," "subsystem," or "component" can refer to software objects or routines that execute on the computing system. The different components, modules, engines and services described herein may be implemented as objects or processes (eg, as separate threads) executing on the computing system. Although the systems and methods described herein are preferably implemented in software, implementations in hardware or a combination of software and hardware are also possible and contemplated. In this description, a "computing entity" can be any computing system, as previously defined herein, or any module or combination of modules executing on a computing system.

All examples and contingent language recited in this specification are for educational purposes to assist the reader in understanding the inventions and concepts to which the inventors contribute in furthering the art. and should not be construed as limited to such specifically recited examples and conditions. Although embodiments of the invention have been described in detail, it should be understood that various changes, substitutions, and modifications can be made thereto without departing from the spirit and scope of the invention. be understood.

Claims (18)

1. A computer-implemented method of delivering information-enabled personalized healthcare in a clinical, non-research setting for early Alzheimer's disease, said computer-implemented method comprising:
Using one or more servers to provide a healthcare social network specific to providing personalized healthcare for individual patients, the healthcare social network comprising:
accepting an individual patient as a new patient into said healthcare social network;
adding content data for the individual patient, including one or more physical measurements from a diagnostic test performed on the individual patient using a test device coupled to the healthcare social network;
designating a recipient within the healthcare social network specific to the individual patient for delivery of the content data;
providing a plurality of human participants with a secure interface specific to the individual patient to send two-way secure communications regarding the personalized healthcare of the individual patient within the healthcare social network; and transfer said content data between said participants based on permissions within said healthcare social network. providing a healthcare social network of
capturing one or more data streams, each of said data streams relating to said individual patient's healthcare provided by one or more of said plurality of human participants, said data streams capturing one or more data streams, one or more of which includes the content data including results from the diagnostic tests of the individual patient;
using one or more processors to analyze the one or more data streams to generate a diagnostic profile for the individual patient;
curating the diagnostic profile by presenting the diagnostic profile with contextual medical information using the one or more processors;
generating, using the one or more processors, an integrated report for presentation to a physician of the individual patient using the curated diagnostic profile, the integrated report comprising: generating an integrated report for providing personalized therapy for the individual patient with respect to diagnostic profile;
voxel-based image quantification , wherein one of said data streams is captured from testing for specific genetic variants, including single nucleotide polymorphisms (SNPs), said testing performed on said individual patient ; cerebrospinal fluid (CSF) analysis for tau protein or amyloid beta peptide of said individual patient obtained from a spinal tap procedure , a blood test for Alzheimer's disease, and one or more of
Said one or more data streams having genetic data obtained from genetic testing for one or more specific genetic variants comprising one or more single nucleotide polymorphism (SNP) minor alleles selected from a group of SNPs. is captured at least one of
The selected one or more SNPs are associated with the following SNP minor alleles within the individual patient's gene RTTN, ALCAM, DMXL2, or DYNLL1: rs4891826 (G), rs2030515 (G), rs7164265 (C) , or rs993900 (A),
A computer-implemented method. one of the data streams is captured from an application running on a mobile or wearable device related to screening for early Alzheimer's disease;
The computer-implemented method of claim 1 . A single nucleotide polymorphism selected from the group of single nucleotide polymorphisms,
rs769449, rs4420638, rs56131196, rs2075650, rs71352238, rs10956245, rs157582, rs283815, rs34095326, rs75627662, rs114798373, rs4689137, rs4698481, rs963281, rs2463471, rs10234008, rs17680408, rs10029820, rs2357394, rs2033296, rs114979482, rs62444137, rs6499632, rs62532372 rs1001029, rs1001026 , rs1551466, rs368475, rs5031052, rs917321, rs2463472, rs2049670, rs75300677, rs4842247, rs12916234, rs10007765, rs17165129, and
rs55867246, rs1930458, rs1999505, rs1427780 rs13071744, rs6576798, rs13222318, rs11006004, rs4357923, rs12589674, rs2174147, rs8009420, rs17815373, rs11161719, rs10515601, rs1874848, rs78323632, rs111851441, rs9920618, rs11740072, rs78252085 rs4845054, rs1086013, rs4508236, rs9458512, rs17147461, rs78341108, rs6866169, rs72681708, rs4912453, rs12120406, or rs8056050
further comprising one or more of
The computer-implemented method of claim 1 . the personalized health care of the individual patient relates to a baseline diagnosis of Alzheimer's disease for the individual patient, such that the integrated report is a baseline integrated report for the individual patient;
The computer-implemented method of claim 1 . said personalized health care of said individual patient relates to longitudinal monitoring of Alzheimer's disease within said individual patient such that said integrated report is an integrated longitudinal safety/efficacy monitoring report ,
The computer-implemented method of claim 1 . The personalized health care of the individual patient relates to biomarkers of the individual patient, and analyzing the one or more data streams is predictive analytics, such that the integrated report is a predictive analytics report. including,
4. The computer-implemented method of claim 3 . the predictive analytics is based on a neural network machine learning algorithm;
7. The computer-implemented method of claim 6 . the predictive analysis predicts Alzheimer's disease at a presymptomatic stage;
7. The computer-implemented method of claim 6 . the predictive analysis predicts the individual patient's response to a particular therapy;
7. The computer-implemented method of claim 6 . When the individual patient is found not to test positive for the rs4891826 (G) SNP, or is determined to be a non-carrier, the personalized therapy includes antibody to BMP-4 protein or a functional fragment thereof,
The computer-implemented method of claim 1 . The predictive analysis is based on social network analysis (SNA) of gene/protein molecular networks with respect to the one or more selected SNPs, and the personalized therapy is selected small molecules based on the SNA. a tumor kinase inhibitor,
7. The computer-implemented method of claim 6 . When the individual patient is found not to test positive for the rs4891826 (G) SNP or is determined to be a non-carrier, the personalized therapy includes a small molecule tumor kinase inhibitor is
The computer-implemented method of claim 1 . When the individual patient is found not to test positive for the rs2030515 (G) SNP, or is determined to be a non-carrier, the personalized therapy includes antibodies to the ALCAM/CD166 protein or a functional fragment thereof,
The computer-implemented method of claim 1 . when said individual patient is found not to test positive for said rs2030515 (G) SNP or is determined to be a non-carrier, said personalized therapy is an mTOR inhibitor;
The computer-implemented method of claim 1 . When the individual patient is found not to test positive for both the rs4891826 (G) and rs2030515 (G) SNPs or is determined to be a non-carrier, the personalized therapy is , combination therapy of an anti-BMP-4 antibody or a functional fragment thereof and an mTOR inhibitor;
The computer-implemented method of claim 1 . The personalized therapy is CRISPR-mediated within the RTTN, ALCAM, DMXL2, DYNLL1 genes based on testing for the rs4891826 (G), rs2030515 (G), rs7164265 (C), or rs993900 (A) SNPs. is gene editing or trans-epigenetic regulation;
5. The computer-implemented method of claim 4 . When said individual patient is found by testing to be non-homozygous for said rs2030515 (G) SNP, said personalized therapy is CRISPR-mediated gene editing in the ALCAM gene, said gene editing is said individual patient to contain two copies of the rs2030515 allele such that the gene-edited stem cells are homozygous, and the gene-edited stem cells are imaged into targeted brain tissue further delivered to the individual patient's brain using guided delivery;
The computer-implemented method of claim 1 . When the individual patient is found not to test positive for the rs2030515 (G) SNP or is determined to be a non-carrier, the personalized therapy targets the ALCAM/CD166 gene. is an antisense oligonucleotide (ASO) or short interfering RNA (siRNA) therapy that
The computer-implemented method of claim 1 .

Images