Method and search platform apparatus for processing search queries targeting medical sample data and / or databases containing samples.

Abstract

A method for processing search queries directed to a database containing medical sample data and / or samples from a plurality of samples of an individual, comprising: inputting a search query to an AI processor device, the search query including a request profile directed to retrieving at least one sample containing sample data having predetermined queried search parameters; searching at least one selected sample using the AI ​​processor device, wherein the at least one selected sample satisfies the request profile with at least a predetermined probability; and outputting identification data of the at least one selected sample. A search platform device configured to process search queries directed to a database containing medical sample data and / or samples from a plurality of samples of an individual is also described.

Description

Technical Field

[0001] The present disclosure relates to a method for processing a search query for a database containing medical sample data and / or samples from multiple samples of an individual (especially a patient), and a search platform device. The present invention particularly relates to a method for using an intelligent and self-learning platform that provides a mediation function in multiple directions between, on the one hand, databases of, for example, research institutes, biobanks, and / or clinics, and individuals whose sample data and / or samples are stored in the database, and on the other hand, research entities such as pharmaceutical industries and (bio)medical research organizations, and an apparatus for establishing the same. The present invention can be used, for example, for the operation of a database having sample data and / or samples from multiple samples of an individual, and / or particularly for the processing of search queries by machine learning and / or artificial intelligence (AI).

Background Art

[0002] For example, for the purpose of research or investigation, when conducting an investigation, research entities (departments), which are parties such as companies in the pharmaceutical industry and / or (bio)medical research organizations, rely on samples and data of a subject cohort. The subject cohort includes subjects having specific characteristics (parameters) for a specific investigation, such as a profile of a specific general disease and / or specific genetic commonalities.

[0003] Samples and data of subjects that may be suitable for forming a cohort are available in medical research facilities and / or biobanks. However, in the past, entities searching for subjects had to spend a great deal of effort searching various medical research institutes and biobanks for the required data and patients. Legal restrictions also apply, and entities are not permitted to directly access or use, for example, the personal data of patients without authorization.

[0004] Commercial medical laboratories and biobanks are very interested in selling the samples and data they possess to research organizations, for example, those that have made inquiries. However, selling these specific samples requires genetic sequencing of the material, because otherwise, research organizations cannot determine the relevance of the corresponding samples to their own research. However, complete genetic sequencing is costly and time-consuming for the laboratories or biobanks concerned. Therefore, complete genetic sequencing is only valuable if the corresponding samples are important to the research.

[0005] Therefore, there is growing interest in developing predictive tools to help organizations that possess samples and data identify those samples that are worthwhile to perform expensive sequencing and other sample preparation on.

[0006] It is generally known that several search engines with specific filtering systems, such as the commercial provider iSPECIMEN, are available and can be used by industry and research departments to find suitable patient samples. However, such search engines can only operate on standardized data within specific databases and cannot provide predictive commentary. Research organizations making queries can only find samples for which identifiers have been provided (tagged) and for which genome sequencing has already been performed.

[0007] Therefore, conventional search engines cannot assist laboratories / biobanks in selecting samples that may be important for research (and especially worth expensive genome sequencing), nor can they generate information to provide feedback to laboratories / biobanks in response to queries from research organizations regarding these samples that are worth tagging and would facilitate cohort formation for the research organizations. [Prior art documents] [Patent Documents]

[0008] [Patent Document 1] International Publication No. 2020 / 233850 [Patent Document 2] International Publication No. 2020 / 233851 [Patent Document 3] German Patent Application Publication No. 102019135380 Specification [Overview of the project] [Problems that the invention aims to solve]

[0009] The object of the present invention is to avoid the shortcomings of the prior art by providing an improved method for processing search queries targeting databases containing sample data from multiple samples of individuals and / or samples, and / or a corresponding search platform device. Processing search queries targeting databases is intended, in particular, to enable faster and / or less effort in identifying target samples. Processing search queries targeting databases is also intended, in particular, to provide faster and / or easier access to sample data and / or samples, and further / or to provide privacy protection to the individuals to whom the sample data and / or samples belong. [Means for solving the problem]

[0010] The aforementioned objectives are achieved by a method for processing database-targeted search queries having the features of the independent claims, and by a search platform device. Preferred embodiments and applications of the present invention will become apparent from the dependent claims.

[0011] According to a first general aspect of the present invention, the above-mentioned objective is achieved by a method for processing search queries targeting a database containing medical sample data and / or samples from multiple samples of an individual, the method comprising the steps of: inputting a search query into an AI processor device, the query including a request profile targeting the retrieval of at least one sample containing sample data having predetermined query target search parameters; using the AI ​​processor device to retrieve at least one selected sample, wherein the at least one selected sample satisfies the request profile; and outputting identification data for the at least one selected sample. The database is provided, for example, by a medical facility, a laboratory, and / or a sample bank (biobank) of biological samples. The request profile is preferably satisfied with at least a predetermined probability, which is determined by the functionality of the AI ​​processor device.

[0012] According to a second general aspect of the present invention, the above objective is achieved by a search platform device configured to process search queries targeting a database containing medical sample data and / or samples from multiple samples of an individual, the search platform device comprising: an input device adapted to receive search queries including a request profile targeting retrieval of at least one sample containing sample data having predetermined query target search parameters; and an AI processor device connected to the input device, adapted to retrieve at least one selected sample, wherein the at least one selected sample satisfies the request profile; and an output device adapted to output identification data of at least one selected sample. The search platform device or one of its embodiments is preferably adapted to carry out a method according to the first general aspect of the present invention or an embodiment thereof.

[0013] The technology according to the present invention preferably offers more functionality than conventional search engines. It is a self-learning predictive tool with integrated artificial intelligence. This technology according to the present invention can, for example, assist the pharmaceutical industry / (bio)medical research organizations in searching for patient cohorts suitable for research, and on the other hand, assist database organizations such as laboratories / biobanks in evaluating which of their samples are worth investing time and cost in sample preparation, particularly for genome sequencing. This evaluation by the platform's learning artificial intelligence (AI) can then be directly influenced by the research entity through queries.

[0014] Using the present invention, queries from research subjects are preferably digitized and processed by an AI processor device. The AI ​​processor device comprises a computer device adapted to use a machine learning-based program, such as at least one artificial neural network. The expression “AI processor device” will also be used hereafter to refer to artificial intelligence provided by the computer device. Methods for establishing AI are described in International Publication No. 2020 / 233850 ("Recursive coupling of artificial learning units") and International Publication No. 2020 / 233851 ("Coupling multiple artificially learning units with a projection level"), which introduce not only the establishment of a learning system consisting of several AI units, but also the memory and independent use of ethical principles / rules by the AI ​​system.

[0015] The AI ​​processor device learns through individual training and / or preferably by evaluating queries directed at the AI ​​processor device and relevant identification results from the database.

[0016] According to a preferred embodiment of the present invention, an AI processor device is used to search for a group of selected samples, all of which satisfy a required profile with at least a predetermined probability. The group of selected samples preferably provides a limited number of samples in relation to a wide range of stored data, and these multiple samples may be further processed or inspected or output for each query, significantly reducing the effort required for database operation.

[0017] Each method or search platform apparatus according to the present invention preferably satisfies at least one of the following two functions. a) Independently generate a ranking of samples according to their relevance to the research. b) Independently evaluate (rate) the parameters that are important to the study (which patient data are relevant to the study).

[0018] According to the first function, the identification data of the selected sample group is output preferably accompanied by ranking information (positioning information) that assigns a probability to each selected sample that the requirements profile is well satisfied.

[0019] According to the second function, additional rating search parameters are preferably output, which are matched to the search query so that a group of samples with a high probability of satisfying the required profile can be selected. The rating search parameters particularly preferably include genetic profiles, clinical data, medical history, familial predisposition to disease, personal diagnostic results, lifestyle characteristics, dietary habits, consumer behavior, exercise and physical activity characteristics, data on drug or other toxic substance intake, drug use data, radiation exposure data, epigenetic data, geographical information, age, sex, ethnicity, allergies, and / or mental illness.

[0020] Compared with conventional search engines that cannot be affected by queries for specific sample profiles from entities such as research institutions and cannot transmit this information to research institutes / biobanks, the AI processor device can be affected by queries and can be specifically trained. Furthermore, conventional search engines cannot independently output an evaluation (rating) regarding which parameters of patient data are important for a study.

[0021] According to a preferred embodiment of the present invention, a request profile from a research entity is intended to retrieve a group of samples including sample data, and individuals associated with the group of samples form a subject cohort. The subject cohort can thus preferably be directly generated in response to a query.

[0022] According to a more preferred embodiment of the present invention, the AI processor device processes specialized information, particularly commercial market information from industrial sectors related to pharmaceutical products, and / or information from an approval database related to pharmaceutical products.

[0023] Thus, the platform preferably forms a search portal for research entities, which is of high practical value for searching for suitable patient cohorts, and the platform transmits search queries to the research institute / biobank. For example, queries from the pharmaceutical industry are preferably used by the platform's AI to generate sample rankings. To generate this ranking, the platform's AI may monitor not only queries from the pharmaceutical industry but also scientific literature. Not only can pharmaceutical companies be placed in a position to have access to appropriate sample materials via this platform, but academic non-profit research organizations can also use this platform. Furthermore, they provide the platform with information regarding which research fields are in vogue. Not only scientific publications but also information from the industrial sector may be important here. For example, which new drugs are entering the market or what trends are observed in pharmaceutical research can be observed from approval databases. All such entities (pharmaceutical companies, non-profit research organizations, the status of relevant publications, the development of new drugs, etc.) all influence the weighting in the ranking of patient samples according to their relevance within the AI of the device (see Figure 1 below). Supported by this information, the platform can successfully advise the research institute / biobank regarding which patient samples are worthy of being subject to expensive genomic sequencing (based on its independently generated ranking). The device's AI gives advice based on the sample ranking generated based on current research trends and queries from research organizations. Supported by this analysis, the research institute / biobank reduces the initial cost of sequencing.

[0024] Preferably, the search query is defined to be converted into a fragmented search query using an encoding function. The conversion is preferably performed by an encoding device connected to the AI processor device and the input device.

[0025] To consider data protection requirements, it is preferable that personal information of individuals, such as clinical data or genome sequencing data, is not freely accessible. This is subject to data protection. Nevertheless, to ensure that the platform can function optimally and mediate between the two, it may be desirable to search for specific features of the patient profile. This search may be performed by using an encoding function, for example, using a hash function, so that the platform itself does not reveal the patient's genetic data, but the hash allows for searching encrypted data in the laboratory database for a genetic profile appropriate to the research subject without knowing specific genetic information. The encoding function may be used in particular by the technique described in German Patent Application No. 102019135380.7 ("Verfahren und Datanverarbeitungsvorrichtung zur Bearbeitung von genetischen Datan", which was not published on the priority date of this disclosure), in which it is not necessary for all information of the data to be disclosed, and research can be conducted on restricted data (such as personal genetic information) using an encrypted method (e.g., hashing). German Patent Application No. 102019135380.7 is incorporated into this disclosure by reference, particularly with respect to the processing of genetic data.

[0026] For this type of application, it is preferable to install a "secure enclave" in the database, through which the platform can transmit queries from the pharmaceutical industry / research organizations regarding specific samples using encoding functions such as hashing. Databases such as laboratories / biobanks can confirm or reject the existence of appropriate samples without disclosing underlying personal data. Thus, the platform does not need to have personal patient data for individual samples / patients, nor a central ID system. Instead, each currently relevant sample is stored in a distributed and temporary (hashed) encrypted format (as in the case of a disease alert app). This device does not store specific genetic data for specific samples, clinical patient data, or other patient profile information. Such personal data remains with laboratories / biobanks, or with physicians and patients. The platform transmits searches in the form of queries transformed using encoding functions, particularly "hashing," so that it does not need to directly access such personal data.

[0027] In a preferred embodiment of the present invention, consent from the relevant parties is required for the use of data and / or samples. The request for consent is particularly preferably made by using questionnaires and / or informed consent documents and / or by an avatar device.

[0028] Preferably, it is even possible to output information about pharmaceutical products, diseases, research results and / or recommended actions, particularly to individuals for whom sample data is stored in the database.

[0029] Features disclosed in conjunction with the search platform device and its embodiments also constitute preferred features of the method according to the present invention, and vice versa. The above-described embodiments and preferred features of the present invention relate particularly to the establishment of the search platform device and the functions and configurations of the individual components described in conjunction with the search platform device, and therefore also apply to the method. The preferred embodiments, modifications, and features of the present invention described above are combinable with each other.

[0030] Further details and advantages of the present invention will be described below with reference to the accompanying drawings. [Brief explanation of the drawing]

[0031] [Figure 1] This is a schematic diagram of a search platform device and search method having features of a preferred embodiment of the present invention. [Figure 2] This describes the functions of a search platform device and search method according to a preferred embodiment of the present invention. [Figure 3] This figure shows the advantages of actually using the present invention. [Figure 4] This is a schematic diagram of an avatar device having features of a preferred embodiment of the present invention. [Figure 5] This diagram illustrates communication between the avatar device and its surroundings (the outside world), as well as the monitoring of that communication. [Figure 6] This is a diagram illustrating the usage of the first and second storage devices. [Figure 7] This is a schematic diagram of an avatar method having features of a preferred embodiment of the present invention. [Modes for carrying out the invention]

[0032] Features of embodiments of the present invention are described below with reference to the configuration and functions of a search platform device and an avatar device optionally connected to the search platform device. Details of the AI ​​processor device are not described to the extent known in the prior art. The AI ​​processor device may be configured in particular as described in International Publication No. 2020 / 233850 ("Recursive coupling of artificial learning units") and International Publication No. 2020 / 233851 ("Coupling multiple artificially learning units with a projection level"). International Publication Nos. 2020 / 233850 and 2020 / 233851 are incorporated into this disclosure by reference, in particular with respect to the configuration of the AI ​​processor device and its peripheral devices such as input / output devices and memory devices.

[0033] (Search platform device and its operating method) Figure 1 illustrates the interaction between the AI ​​processor device of the search platform and research organizations, approval agencies, and scientific institutions. The AI ​​processor device, connected to an input / output device (not shown), learns from queries from the pharmaceutical industry or other research organizations, and by accessing information on current scientific publications, medical innovations, and / or samples of particular importance to the research, and weights the information accordingly. The AI ​​processor device learns further in a similar manner to assess which parameters are important to the research for each individual, such as genomic features, age, sex, lifestyle, and medical history. These queries influence the weighting of rankings and ratings based on the querying entity.

[0034] Figure 2 illustrates the interaction between the AI ​​processor device of the search platform and databases such as laboratories / biobanks and physicians / patients. By evaluating queries from research subjects (Figure 1), the AI ​​processor device can provide laboratories and biobanks with information (identification information) about which samples are worthy of genome sequencing and which parameters are important to the research subjects. Patients or physicians can also gather information directly from the search platform device about where the current focus of the research lies. Communication between the search platform device and the databases can be conducted completely anonymously and / or using encoding functions, particularly hash functions, via a secure enclave within the laboratory / biobank. Ethical avatars, as described with reference to Figures 4-7, can be used as digital proxyes for individuals to assist the search platform device in responding to queries (predicting appropriate samples), for example, to verify whether a particular patient is eligible to participate in the study.

[0035] The search platform device can preferably develop a ranking of samples (e.g., from "important" to "unimportant," or based on appropriate probability) and identify parameters important to the study based on positive or negative responses from the database (laboratory / biobank) and / or search queries for specific profiles from the research organization. Such parameters may include, for example, not only the patient's genetic profile but also data on the patient's clinical health, pre-existing conditions and familial predisposition to disease, individual diagnostic results, the patient's lifestyle, dietary habits, consumer behavior, exercise and physical activity, drug intake, medication use, any radiation exposure, epigenetic data, geographical information, age, sex, ethnicity, allergies, and / or mental illness. The search platform device assists laboratories / biobanks in both assessing which samples currently have significant value to the study and evaluating which of those parameters will be important to the investigation and whether collecting them is equally valuable to the laboratory / biobank (see Figure 2).

[0036] The use of patient data / samples may require the voluntary consent of the patients concerned. To facilitate obtaining such declarations, ethical avatars may be used. These avatars are intelligent tools that have learned and can simulate the ethical intentions of real people. With the assistance of these avatars, the platform can pre-select the ethical background of specific data / samples for the pharmaceutical industry, and pre-selection can be carried out fairly quickly by surveying their ethical avatars instead of contacting real people to inquire about their intentions regarding specific options for the application and use of the data / samples. Thus, the platform not only assists pharmaceutical companies (or non-profit scientific organizations) in finding patients with appropriate genetic and medical profiles, but also simultaneously pre-selects which of these patients are likely to be more inclined to participate in R&D (research and development). The ethical avatars function as predictive units of patients' willingness to participate in research, which is derived from their personal values ​​or religious intentions or cultural background.

[0037] However, the prediction of whether patients will, in principle, be willing to participate in the study may be made not by an "ethical avatar," but by the platform's AI evaluating the corresponding questionnaire or informed consent document.

[0038] The operating method of the search platform device brings the following advantages to the stakeholders.

[0039] [Benefits for individuals, especially patients, whose samples and sample data are stored.] The platform's operating mechanism satisfies all data protection requirements, and patient-related data does not need to be stored on the platform (see the above explanation regarding hash-based searching). Therefore, optimal security measures are in place to protect the patient's interests regarding data and privacy.

[0040] By efficiently operating the search platform, patients are given the opportunity to participate in research that may be important to them (for example, because it is related to their illness). The platform can also provide patients with links to important sources of information, such as those particularly relevant to the medications they are taking or that offer new research findings regarding their clinical picture.

[0041] If interested, patients may share personal information (such as behavioral and environmental data) by directly contacting the platform, for example, by granting access to data from various fitness apps. Thus, patients may be linked to suitable research projects. While this all requires a high level of patient participation, it offers the benefit of personalized access to medical innovations that are highly relevant to their individual needs and closely related to the latest research findings.

[0042] [Advantages for databases, especially for research institutes / biobanks] The AI ​​processor in the search platform device processes large amounts of data (queries from research organizations, the status of current scientific publications, approval procedures, and current medical innovations) to evaluate the aforementioned data and thus predict which samples are relevant to research. These predictions are highly valuable to laboratories / biobanks because they allow them to assess which of their samples are worth subjecting to costly full genome sequencing. In particular, sequenced samples are valuable to the pharmaceutical industry and (bio)medical research organizations, so it is crucial for laboratories / biobanks to primarily sequence these samples, which will later be purchased by research organizations. Therefore, the platform facilitates intermediation between laboratories / biobanks and potential buyers of samples. However, the platform further assists laboratories / biobanks in assessing which samples should be collected in order to make their parameters important to research organizations (this may include not only clinical and diagnostic data, but also the aforementioned information such as age and sex). The platform's rankings allow laboratories / biobanks to assess which samples are suitable for rough SNP sequencing and which samples are considered relevant enough to warrant whole-genome sequencing (because they are highly likely to be purchased).

[0043] [Benefits for the inquiring party, especially the pharmaceutical industry / (biological) medical research organizations] The platform assists research organizations in collecting appropriate patient cohorts, thereby transmitting queries from research organizations about specific samples to relevant laboratories / biobanks. These queries influence both the weighting within sample rankings and the platform's AI-driven assessment regarding which sample parameters (e.g., patient radiation exposure, medical history, drug abuse, etc.) should be considered important. Thus, the platform allows research organizations to influence laboratory / biobank sequencing and the emphasis placed on specific sample parameters for their own purposes.

[0044] This platform not only protects personal patient data with security measures in place (see explanation above), but also ensures that internal research information from pharmaceutical companies, disclosed to some extent through queries, is treated confidentially. Companies have absolutely no desire to attract unwanted competition by disclosing their current research areas.

[0045] Figure 3 illustrates the benefits of using the search platform device based on a business plan. Users in the pharmaceutical industry pay a predetermined amount to use the search platform device, for example, EUR 100,000 per user. In return, the platform mediates between users and laboratories / biobanks, providing samples that satisfy the request profile queried by the user, and the weighting of sample ranking is affected. This ranking directly influences which samples are sequenced by the laboratories / biobanks. For specific samples relevant to the research, the search platform device may allocate sequencing costs. Nonprofit research organizations in the scientific field can use the platform free of charge. The AI ​​of the search platform device benefits from the queries, as it can adapt its ranking to the current research interests based on the queries. Patients participating in the study, or those willing to share more personal data, may in return be connected to relevant networks related to their disease, for example.

[0046] For example, genome sequencing costs a laboratory / biobank around 500 euros. Laboratories / biobanks only recoup this cost for about 1% of the samples they possess. Performing relatively rough sequencing methods (such as SNP sequencing) or diagnostic tests costs approximately 50 euros each time. Predictions from search platform devices provide laboratories / biobanks with valuable suggestions regarding which samples are appropriate for their research organizations, potentially leading to faster and more successful sequencing and more efficient use of sequencing costs.

[0047] Pharmaceutical companies and other for-profit organizations making inquiries must pay a certain amount to the fund (in the form of a "participation fee") if their queries target the platform, because their queries influence the platform's ranking and parameter evaluation. Non-profit organizations such as universities and other non-profit (bio)medical research organizations can run queries free of charge. For the platform, the value of such queries lies in the information they provide, and the platform's AI system can learn from each query which samples are currently being used in research and adapt its ranking and predictions accordingly.

[0048] When a research institute / biobank inquiring about a particular sample is ranked as highly relevant to the research and performs certain diagnostic tasks that provide important parameters for the research, it will receive funding from the platform's money fund when sequencing that sample. In this way, the institute / biobank is encouraged to sequence specific samples, which then benefits the pharmaceutical industry / (bio)medical research organization inquiring about the sample, thus providing an interesting set of samples that have been queried. Furthermore, the platform's funds will be used for the operation of the platform (see Figure 3 for an overview of the business plan).

[0049] Payments for individual samples by purchasing research organizations are made directly between the institute and the research organization; the search platform equipment is not involved in this process.

[0050] (Avatar device and method for use with search platform device) The present invention further relates to an avatar device ("ethical avatar") that represents an individual and is adapted to process the individual's personal data, an avatar system having multiple avatar devices, uses of avatar devices, and methods for operating avatar devices. The present invention particularly relates to a method for developing a virtual surrogate of an individual, also referred to herein as an ethical avatar or avatar device, and a device for establishing such surrogate, which can learn, for example, an individual's moral intentions, ethical principles, values, and / or further opinions and interests, and simulate independently. The present invention can be used, for example, for the operation of a database having an individual's personal data, and / or, in particular, for the processing of personal data by machine learning and / or artificial intelligence (AI), and / or for the processing of queries targeting the search platform device described above according to the present invention.

[0051] During (bio)medical research and in the medical care and treatment of individuals (patients and / or subjects), large amounts of personal data about patients or subjects are accumulated through the provision of biological materials (tissue samples, samples from somatic cells or stem cells) or in the course of disease treatment (biopsy collection, tumor removal, etc.). This personal data may include, for example, the patient's / subject's genetic information, medical history, and / or environmental impacts to which they are exposed.

[0052] Managing and protecting such data requires significant investment in information technology, prioritizing adherence to ethical standards when handling this data (see provisions on data protection and "informed consent").

[0053] Traditionally, informed consent has been obtained in writing from individuals, and patient / subject identities have been anonymized or pseudo-anonymized to protect confidential data from unauthorized access by third parties. This necessary method makes it extremely difficult, or even impossible, to interview those involved after a certain period of time. If circumstances change, new research fields emerge, or there is uncertainty in individuals' opinions or interests, it may no longer be possible, or if possible, to resolve the resulting issues with great difficulty.

[0054] In opinion surveys, sociological or demographic analyses, big data research processes (data science), and the industrial / commercial sector, large amounts of personal data must be processed as quickly as possible while ensuring the privacy of stakeholders. This process has historically been extremely time-consuming and costly.

[0055] In the field of personalized medicine, attempts have already been made to digitize human bodily functions and body parts (see, for example, the "Blue Brain Project" and virtual heart simulations). The development of digital versions of body parts is becoming more common than ever before, and it is becoming possible to virtually represent cells, organs, or even an entire individual. This concept is called a "digital twin," and it generates a digital double of the person as a biological replica.

[0056] Previous methods were limited to simulating human physiological processes. In contrast, mental response characteristics such as a person's moral intentions, opinions, individual values, and ethical principles have not been considered.

[0057] An objective of this secondary aspect of the present invention is to provide an improved, data-based digital proxy for an individual and / or a method of using such a proxy, while avoiding the shortcomings of prior art. The digital proxy for an individual is intended to enable the provision of additional information, and / or information with expanded categories, such as non-physiological information relating to the individual, and / or to provide a wider range of uses. The digital proxy for an individual is further intended to provide faster and / or easier access to the individual's comments and / or to provide protection of the individual's privacy.

[0058] According to the present invention, an avatar device, an avatar system having a plurality of avatar devices, an application for an avatar device, and an avatar method for operating an avatar device, having the features described below, are used for this purpose.

[0059] Preferably, an avatar device is used that is adapted to represent an individual and to process the individual's personal data. The avatar device is an AI processor device adapted to machine learning of an individual's individual mental response features by data training, and is configured to respond to questions in a manner dependent on the mental response features and to generate semantic answers (in particular, at least one sentence or individual words such as "yes" or "no"); a first memory device (personal memory) connected to the AI ​​processor device and adapted to store the individual's mental response features; an input device connected to the AI ​​processor device and adapted to interact with the individual and / or receive questions about the individual; and an output device configured to output signals representing the answers generated by the AI ​​processor device.

[0060] In another embodiment, preferably, an avatar device or one of its embodiments is provided, the use of which includes at least one of the following: forming a multidirectional platform for connecting a medical laboratory and / or biobank with a business company and / or scientific research organization; forming a platform for social surveys, opinion surveys and / or big data analysis; forming a platform for providing patient declarations, in particular consent declarations for the use of personal data; forming a data protection platform; generating advance directives for individuals; providing input information for synthetic biology procedures; and controlling a technology system.

[0061] In another embodiment, an avatar method is used to operate an avatar device for an individual, the method comprising the steps of receiving at least one question about the individual, processing the question using an AI processor device, and outputting a semantic response to at least one question.

[0062] According to another aspect of the present invention, an avatar system comprising multiple avatar devices is used, each avatar device being adapted for a different application.

[0063] Using the avatar device, an AI processor device selectively digitizes and / or simulates an individual's mental response characteristics. The AI ​​processor device comprises a computer device adapted to use a machine learning-based program, such as at least one artificial neural network. The term "AI processor device" will hereafter also be used to refer to artificial intelligence provided by the computer device.

[0064] Preferably, the avatar device provides a digital twin of an individual to whom the avatar device is associated, and the avatar device represents, for example, the individual opinions, interests, and / or moral intentions of a real person. Preferably, the avatar device enables the simplification and / or acceleration of the management of personal data. At the same time, the avatar device ensures that the privacy of the parties concerned is protected and that their interests are given maximum consideration.

[0065] The avatar device according to the present invention preferably constitutes an intelligent, digital, and predictively usable support system that acts as a proxy for a real person. The avatar device can be used, for example, as an advisory link between an individual and an inquiring entity such as a research organization, research institution, or company, while protecting the anonymity of the parties involved (eliminating the need for direct contact with the individual) and simultaneously representing the individual's interests in transactions with others. These advantages will be described below with particular reference to the drawings.

[0066] Preferably, the avatar device is a tool that can predict the comments, especially opinions, of the actual person, and can be used in any form of database / analysis that previously required surveying thousands of individuals in a time-consuming and costly manner.

[0067] An ethical avatar constitutes artificial intelligence (AI) that can be trained by an individual, particularly regarding their moral intentions, and which, based on this training, can substantially predict answers to newly posed questions. During this training, the avatar's AI learns the individual's moral intentions, values, interests, and beliefs. Using the learning associative AI, the avatar can utilize these beliefs / values / intentions when answering new (ethical) questions. The avatar device is therefore capable not only of expressing what it has learned, but also of intelligently and creatively applying what it has learned to new contexts. Based on its knowledge of the person involved, it can intelligently predict what opinions and judgments that individual will make in response to questions in a particular new context.

[0068] The AI ​​processor device is preferably adapted to machine learning of individual mental response features, including the individual's moral intentions, opinions, values, interests, and / or ethical principles. Response features include the individual's subjective characteristics that characterize how the individual responds to questions, facts, experiences, situations, others, social groups, and / or social concepts, and responses may include clarifying, for example, the form of comments, emotions, and / or behaviors.

[0069] In a preferred embodiment of the present invention, the first memory device is configured to store mental response characteristics in the form of a list of rules, a microcontract (Mikrovertraegen), and / or a blockchain. However, the implementation of the present invention is not limited to these variations, and other digitizable forms are also possible.

[0070] Preferably, the first memory device is adapted to store additional data on which the mental response characteristics depend, and in particular, including the individual's religious affiliation, group affiliation, association membership, and / or political party membership.

[0071] According to another preferred embodiment of the present invention, the AI ​​processor device is adapted to understand and use ethical expressions.

[0072] Preferably, a second storage device (public storage, shared storage) is provided and adapted to store general-purpose additional data representing knowledge about mental response characteristics. The second storage device is particularly preferably connected to an update device, which can update the general-purpose additional data.

[0073] According to a more preferred embodiment of the present invention, the avatar device, particularly the input device, has a communication channel adapted for communication between the avatar device and an individual. This preferably allows the individual to access the avatar device for purposes such as checking its functionality, training an AI processor device, or modifying stored data.

[0074] Therefore, not only is the avatar device trained by the relevant parties, but the actual human being whose moral intentions the avatar device simulates also has the ability to monitor, at any time, comments made by the ethical avatar, especially decisions, correct them as necessary, further train the ethical avatar, or change its stored values / interests / moral principles. However, individuals also have the option of transferring all decision-making authority to the avatar, so that they are represented entirely by their own avatar.

[0075] Since avatars are always under the individual's control and function merely as auxiliary representatives that provide advice based on the individual's mental response characteristics, such as their personal opinions and interests, they preferably do not restrict the autonomy of the actual human being.

[0076] According to another preferred modification of the present invention, the input device comprises an authorization device configured to restrict the reception of questions concerning the individual to a predetermined authorized entity, such as a pharmaceutical company and / or a research organization, and / or to predetermined acceptable content. The authorization device particularly preferably comprises a communication authentication storage and standard communication protocol functions.

[0077] In this embodiment, the protection of individuals' personal data is of particular importance, and the avatar does not permit outsiders to access the data, and when answering questions, it provides personal data only to authorized subjects and within a limited scope.

[0078] The avatar device preferably further includes cryptographic storage adapted for storing access keys. This preferably allows access to the avatar device to be restricted to a highly reliable manner.

[0079] According to a more preferred embodiment of the present invention, the avatar device has a log storage area adapted to store a history of changes to the contents of a first storage device and / or a second storage device.

[0080] Furthermore, preferably, a timer device may be provided that is adapted to provide a time scale for the functions of the avatar device.

[0081] The avatar device preferably includes a communication device adapted to contact and / or communicate with an individual. This preferably allows for verification of semantic responses provided by, for example, an AI processor device.

[0082] Features disclosed in conjunction with the avatar device and its embodiments also constitute preferred features of the method according to the present invention, and vice versa. The above-described embodiments and preferred features of the present invention relate particularly to the establishment of the avatar device and the function and configuration of the individual components described in conjunction with the avatar device, and therefore also apply to the present method. The above-described preferred embodiments, modifications, and features of the present invention are combinable with each other.

[0083] The features of exemplary embodiments of the avatar device are described below with reference to the configuration and function of the avatar device. Details of the AI ​​processor device will not be described to the extent that they are known in the prior art.

[0084] Figure 4 shows an embodiment of an avatar device and its functions, and sections (steps and / or components) a to g shown in Figure 4 have the following meanings: Section a includes the step of storing learned materials and / or accessing learned materials; Section b includes the step of recording the time of the training unit; Section c includes the step of accessing a second storage device so that the investigation can be performed more intelligently and compared with a timer device (clock system); Section d includes the interaction between key management and authentication storage; Section e includes the step of specifying who is authorized to ask questions and / or receive information; Section f includes the step of influencing communication with the surroundings, in particular communication protocols, to enable or prevent their use; Section g relates to channels for AI trained by an individual, AI being investigated, and / or AI providing information.

[0085] Communication between the avatar device and its surroundings (the outside world, particularly the inquiring entity or the actual human training the avatar) (see Figure 7) and monitoring of such communication involves the following steps, as shown in Figure 5: Step a is communication with the actual human. In step b, the AI ​​processor device uses its cryptographic storage to verify whether the individual is authorized to make changes to and / or communicate with the personal storage. In step c, a yes / no decision is made regarding whether to allow training and / or whether to provide information. Step d is a detailed query from an entity, such as a research organization, regarding a specific item. In step e, the avatar's AI processor device uses the directory of all authorized entities in its communication authentication storage, and in particular, the communication protocol in the communication authentication storage, to check whether the AI ​​processor device is authorized to provide information to the entity and for what items the AI ​​processor device is authorized to provide information to the entity. In step f, the results of the check are provided.

[0086] Figure 6 illustrates the storage and retrieval of data in the first memory device (personal memory) in combination with the use of a second memory device (shared public memory) for the purpose of arranging training content in chronological order. The ethical avatar's AI processor device can not only store and access (for authorized queries; see Figure 2) the current mental response characteristics of a real human being, such as moral intentions / opinions / interests, in personal memory, but can also store past versions of the mental response characteristics of a real human being in a history log and arrange them in absolute chronological order using a clock system. This chronological assignment also enables alignment with the second memory device. For example, the AI ​​processor device can access the second memory device to ask about the individual's denomination and understand the response "Christian," so the second memory device can be further used to make the AI ​​processor device more intelligent in its training with real humans, because what "being Christian" means in terms of moral intentions at a particular point in time is recorded in the second memory device.

[0087] The avatar device has the following features as shown in Figure 4. ● The avatar device includes (1) a learning component that can learn moral intentions through individual training by real people, ●(2) The results of this training are stored in a first memory device, and the avatar's AI can access this information (the first memory device is private, i.e., access from outside the avatar device is blocked). Mental response characteristics such as an individual's moral beliefs may be stored in the avatar's first memory device in the form of a list of rules, micro-contracts, and / or blockchains. Information regarding the individual's religion or affiliation with a group / association / political party may also be stored in the first memory device, from which the ethical avatar can draw conclusions about the individual's opinions (by accessing another memory area (the second memory device, see (4))). ●Furthermore, (3) the avatar device is adapted to independently generate new answers (outputs) to new questions (inputs) based on its stored knowledge of the mental response characteristics of the individual, such as their intentions. Such an avatar is preferably able to understand and use expressions related to ethics, which is made possible by a corresponding ontology (in particular, a systematic and formalized representation of a subject, such as physical and abstract objects, features, facts, events, and methods). Apart from the first memory for an individual's moral intentions, the avatar may (4) preferably have access to another second memory, which stores more fundamental knowledge about response characteristics, such as moral knowledge (from the aforementioned professional training). This may be, for example, knowledge of various ethical theories and their implementation, knowledge of religions or other groups (such as political parties or associations) and their rules and regulations, or ethical expressions known and discussed in moral philosophy, and other knowledge of ethical dilemma situations. The second memory (a public memory, shared, for example, with other avatar systems) may be used by the avatar device to enhance the individual's awareness of specific questions and categorize their answers during training, in order to more intelligently investigate the individual. The avatar uses this general knowledge only during training, but not when answering ethical questions on behalf of the individual after a certain period of time. In this case, the avatar makes decisions only with respect to mental response characteristics, in particular what it knows about the individual's personal values / interests / beliefs. The second memory device may be updated periodically, and for this purpose, the avatar's clock system (see (9)) is used. ●(5) A communication channel is provided between the ethical avatar and the actual individual, as a result, the individual can train the avatar, preferably monitor the behavior of its digital proxy at any time, and / or resume training if so desired. Communication with the inquiring entity is also possible, but preferably restricted. Preferably, only queries from authorized entities on authorized item areas are permitted. Authorization is performed when, ●(6) By a communication authentication storage that also incorporates standard communication protocol functions. This is a directory of entities with which the avatar can communicate, and is subject to conditions regarding the context in which it may be relevant to answer questions from the inquiring entity. Authorization can be designated by either an actual person or a central authority. Only authorized entities may receive information within a clearly defined and limited scope; for example, an authorized medical organization may investigate the avatar, but only medical-related questions are permitted (Figure 5 illustrates communication between the avatar and the outside world). ●(7) Using the cryptographic storage where the access key is managed, the avatar may preferably check which entity has what rights to receive the information and to what extent. Using this key further ensures that only real people are authorized to make changes to the avatar's personal storage. ●Since ethical avatars can be trained by real people not just once, and the said individuals can preferably change the contents of their personal memory at any time, the avatar (8) preferably includes a memory (log memory) in which the history of changes is stored (e.g., using a blockchain). Here, the avatar is, ●(9) A clock system is provided to allow for the absolute evaluation of the execution of various training exercises in chronological order and to align this with a second memory device with respect to general moral knowledge. In this way, the time t of a particular training sequence x However, it can be assigned to the general state of knowledge at precisely this time, as it is stored in public shared memory (see (4)). This can be of great significance in understanding the content of each training unit, for example, when a real person is t x At any given time, you indicated that you felt bound by the norms of a particular community, or that the community had changed its norms. x+n Whether or not this was shown could make a difference. Both standards (older version t x and revised version t x+n(4) The training units are stored in a second memory device in a manner linked to their respective times, similar to how the training units are stored in personal memory, and (2) in a manner linked to the training time by the clock system. The individual's personal intentions are therefore arranged in absolute chronological order (as shown in Figure 6). ●Ethical avatars may further (10) have units relating to the basic needs of actual human beings, which may provide links to means of contact or make comments relating to the ease of contacting the individual.

[0088] In carrying out the present invention, the following possibilities are particularly relevant. a) Each real person has only one corresponding avatar device (ethical avatar), preferably including a corresponding identification tag, to which various entities can query. All information about the said individual is processed by this single avatar device. b) A real person may be assigned to multiple avatar devices (avatar systems) generated in various areas of the application. Different avatars can be synchronized with a single person by any of the following: an avatar identification number (provided by the avatar information subject), a universally unique identifier (such as a patient ID or biosample ID), or a (hashed) personal identifier (such as an image [for example, the concept of "gravatar" has been developed, which is an image that functions as an identifier that follows the user from website to website (see https: / / de.gravatar.com)]), or a genetic "fingerprint" such as a single nucleotide polymorphism (SNP) or short ray repeat (STR). If a person has multiple avatars, these avatars are preferably synchronized with one another so that all avatars involved remain up-to-date with respect to common mental response characteristics of the real person, such as intentions and opinions. Consistency between the various avatars of a real person must be guaranteed.

[0089] (Details regarding specific exemplary embodiments and uses) [1. Exemplary embodiments of big data analytics / data science] Avatars, as part of a multi-directional platform acting as an integrated link between medical research institutes / biobanks and pharmaceutical companies and / or scientific organizations, allow medical-related (commercial) laboratories to access vast amounts of data on patient clinical information, diagnostic outcomes, and genetic information. They are interested in making this data available to the research and industry sectors. Meanwhile, the pharmaceutical industry and scientific organizations rely on obtaining data from highly specialized patient groups in their research. Using conventional technologies, mediating between these sectors is extremely complex, but this can be facilitated by using a platform that includes at least one integrated avatar device. The platform aggregates information from both sides, analyzes which patients are being tested in the laboratory, and gains insights into which of these data / samples from the laboratory are important to the research sector, based on queries for patient profiles from the research sector. Thus, the platform can support both sides and advise laboratories on which samples / data are worth subjecting to expensive full genome sequencing (which is necessary for the research sector) (because the platform knows which samples / data are important to the research / industry sector), thereby enabling them to collect appropriate cohorts and support the research sector. Ethical avatars of patients assist in this pre-selection process, particularly because they are aware of the moral intentions of the actual person they are interacting with. The use of patient data / samples requires the consent of the patient concerned. Ethical avatars may be used to facilitate obtaining such declarations (see below). ○ Sociological research: Avatars can essentially help pre-select who might be interested in which research topics and survey content, or who might immediately object. →Advantages: Rapid and appropriate pre-selection of sociological cohorts / research groups. ○ Opinion Surveys / Big Data Analysis: To enable the easy and rapid matching and analysis of the opinions of a very large number of individuals, ethical avatars can be advantageous not only by assisting in responses to specific questions from opinion survey organizations, but also by assisting in assessing the suitability of the corresponding survey subjects at an initial stage (by assisting with pre-selection).

[0090] [2. Possible applications in clinical settings and biomedical research] ○ Providing advice on informed consent questions: The avatar learns the individual's answers to various consent declaration questions and becomes substantially capable of anticipating the individual's intentions to new questions and answering them on their behalf. The avatar acts as the individual's advisory agent and learns to answer on their behalf. → Advantages: It eliminates the need to contact the individual directly for questions regarding (one or more) consent declarations, ensures the individual's privacy is protected, and takes into account the individual's interests and instructions. Standardization of the informed consent process: The process of obtaining informed consent can be standardized by using ethical avatars. By using such internationally applicable, standardized equipment as ethical avatars, the existence and complexity of multiple different informed consent documents can be eliminated. →Advantages: International cooperation becomes much easier because the international circulation and understanding of such documents are simplified, and their use as an international vocabulary and international approaches become possible. Dynamic patient advance directives: Avatars can perform support functions when an individual is unable to express their own will. If an individual is in a coma or in a similar situation where they are unable to express their opinion, it would be extremely helpful for relatives or legal guardians if an ethical avatar trained to correspond to that individual could investigate the avatar on behalf of the individual regarding their wishes, interests, and moral intentions. →Advantages: Stakeholders can rely on the fact that even if they themselves lack the ability, the ethical avatars they have trained can anticipate decisions that are in their best interest. Unforeseen future research / application areas: As avatars become more broadly trained, they can learn more about the general moral intentions / values / beliefs of the corresponding individual, in addition to specific "informed consent" questions, and can even anticipate answers to questions beyond those related to "informed consent." →Advantages: It becomes possible to make the most of existing cell donations and enter new research fields rapidly. ○ Stratification of cohorts for biomedical research: The avatars provided here can act on behalf of patients / subjects without disclosing sensitive data. Here, the technology described in German Patent Application No. 102019135380.7 ("Verfahren und Datanverarbeitungsvorrichtung zur Bearbeitung von genetischen Datan", which was not published on the priority date of this disclosure) may be used, and here, not all information of the data needs to be disclosed, and research can be conducted with restricted data (such as personal genetic information) using an encrypted method (by hashing). German Patent Application No. 102019135380.7 is incorporated into this disclosure by reference, particularly with respect to the processing of genetic data. In this case, the avatar can access the corresponding hash and answer questions with yes / no without knowing the encrypted information. This service is particularly advantageous in collecting patient / subject cohorts with specific common characteristics. For example, research often requires a donor with a specific, for instance, genetically similar, cell line. In this case, the corresponding avatar can respond to a query and approve or reject the suitability of the donor cell line indicated by the avatar without revealing the actual genetic information of the other party. The search for suitable patients / subjects can be performed very quickly, preferably by answering questions about an individual's specific genetic characteristics, and avatars can further assist in performing pre-selection without disclosing the underlying sensitive data of the corresponding individual. →Advantages: While ensuring the security of confidential data, corresponding questions are answered simultaneously, and rapid stratification of research cohorts is possible, which previously required considerable effort and time. Synthetic Biology: In the future, it may be possible to develop synthetic biological units, cells, or organoids solely from genetic information. In this case, too, an ethical avatar is assumed to represent the relationship between the actual human being (with the associated genetic information) and the research organization. The avatar can, on behalf of the individual, indicate consent or prohibition regarding what synthetic products may or may not be produced. →Advantages: In new research fields and application areas, individuals can be represented while security measures are in place to protect their identity, and questions can be answered quickly and reliably by advisory avatars on behalf of those involved.

[0091] [3. Use in situations where the law applies] ○ Support for assessing an individual's legal capacity: The question of whether an individual has adequate legal capacity depends particularly on the consistency of their opinions and whether those opinions have changed frequently and in contradictory ways recently. All of this can be done by an avatar that knows not only an individual's current opinions but also the "history" of those opinions, and can examine the logical, and sometimes contradictory, relationships between differing opinions. Avatars can assist judicial and psychological professionals in assessing an individual's mental state. This may also be relevant to psychological assessments in a medical context, where inconsistency in opinions may indicate that the individual is experiencing progressive confusion / dementia. →Advantages: Experts can consider not only the individual's current situation for assessment, but also the history of their opinions over time, or the dynamics of their opinions over time.

[0092] [4. Exemplary Embodiments of Control of Technical Equipment] 〇 Use in technical equipment: Avatars can support technical equipment by implementing user preferences and interests, for example, by controlling the filtering systems of various internet search engines so that only desirable content is displayed to the individual and other content is excluded. Avatars can also perform management functions for minors by ensuring that content inappropriate for minors is not displayed to them. → Promoting the use of targeted technological equipment in a manner that aligns with user wishes / interests and ethical intentions. 〇Implementing Morality in Self-Driving Vehicles: In the search for an appropriate moral system for self-driving vehicles, conventional solutions based on common understanding (implementation of average moral intentions) have been sought. However, with the support of ethical avatars in particular, it will become possible to implement the driver's own moral intentions into their self-driving vehicle. Therefore, the avatar will function as the individual conscience of the vehicle, just like a human driver in a non-self-driving car. →Advantages: Drivers are not required to adhere to average moral views in road traffic situations and can implement their own personal ethical values ​​(as they would when driving manually) (within obvious but legal limits).

[0093] [5. Other Exemplary Embodiments] ○ Management and Security Measures for Access Keys to Pseudo-Anonymized Personal Data: To protect individual rights and privacy, personal data is protected by encryption in many areas of use. Here, the management of these keys is of paramount importance. These access keys can be lost due to the individual's forgetfulness or negligence. Here, ethical avatars can assist in the security measures and protection of these keys for the corresponding individual against unauthorized access. →Advantages: Support for ethical avatars enhances personal responsibility, keys are professionally managed and secured, and the likelihood of loss is reduced. ○ Data protection monitoring: Avatars constitute a link to communications between individuals and (medical) organizations and / or research institutions. →Advantages: There is no need to contact individuals directly, their identities remain protected, and at the same time, the opportunity to participate in questions relevant to the individual and their interests is maximized. Furthermore, avatars can answer questions without disclosing confidential data (see above). ○ Providing further moral and philosophical capabilities: Ethical avatars not only have access to the values / moral intentions / beliefs of natural persons they have learned, but also possess a broad range of general moral and philosophical knowledge (in a separate memory area shared with other avatars). During training conversations with individuals, ethical avatars can use this knowledge to raise the individual's awareness of a wider range of moral questions and to investigate the individual more intelligently. In this way, the avatar becomes more than just a receiving subject that learns the individual's intentions. Furthermore, while becoming an understanding person using its own moral capabilities, these capabilities are used only during the training process with the individual. After training, when answering questions about the moral intentions of that actual person, the avatar follows the comments made by the individual without accessing its own general moral and philosophical knowledge. →Advantages: Many individuals are initially unaware of the importance of their own moral intentions, and the moral capabilities of the ethical avatar allow individuals to be intelligently investigated during training conversations without compromising their autonomy in decision-making. Furthermore, stakeholders can be motivated in this way (through conversations with their avatars) to consider reaction features such as moral reaction features, thereby training their own moral judgments.

[0094] The features of the present invention disclosed in the above description, drawings, and claims may be important, both individually and in combination or subcombinations, for realizing the invention in its various embodiments.

Claims

1. A method for processing search queries from a research entity using a learning-type AI processor device, targeting a database containing sample data from multiple individual samples, A step in which the AI ​​processor device acquires a search query that includes at least one search parameter, wherein the search query includes a request profile that targets the extraction of at least one sample that includes sample data having the at least one search parameter. A step of searching for sample data of a selected sample from the sample data contained in the database, wherein the AI ​​processor device learns the importance of samples and parameters according to the research entity by undergoing individual training and evaluating the search query targeting the AI ​​processor device, A step in which the AI ​​processor device generates a ranking of the selected samples, wherein the AI ​​processor device generates the ranking of the selected samples based on the appropriate probability of the search query from the research entity, ranging from important samples that are relevant to the search query from the research entity to unimportant samples that are not relevant to the search query from the research entity. The AI ​​processor device determines, based on the responses from the database and / or the search queries from the research entity, which of the additional parameters of the important sample are important to the research entity. The steps include: evaluating the search queries from the research entity, thereby providing identification data regarding which of the selected samples is worth considering for genome sequencing; The output device outputs the identification data of the sample data of at least one of the selected samples, along with the ranking of the selected samples and additional parameters for the selected samples, wherein the additional parameters match the search query so that a group of samples with a high probability of satisfying the required profile can be selected. Methods that include...

2. The AI ​​processor device searches for the group of sample data of the selected sample based on the search parameters included in the search query. The method according to claim 1.

3. The AI ​​processor device obtains consent from the avatar device, which functions as the digital twin of the individual, when using the individual's data or samples. The method according to claim 1.

4. The aforementioned additional parameters include at least one of the following: genetic profile, clinical medical data, medical history, familial predisposition to disease, personal diagnostic results, lifestyle characteristics, dietary habits, consumer behavior, exercise and physical activity characteristics, data on drug or other toxic substance intake, drug use data, radiation exposure data, geographical information, age, sex, ethnicity, allergies, and mental illness. The method according to any one of claims 1 to 3.

5. The search query aims to retrieve a group of samples containing sample data, and the individuals associated with the group of samples form a subject cohort. The method according to claim 2.

6. The AI ​​processor device processes specialized information including at least one of the following: commercial market information from the industrial sector relating to pharmaceutical products, and information from an approval database relating to pharmaceutical products. The method according to any one of claims 1 to 5.

7. The step of converting the search query into a fragmented search query using an encoding function, The method according to any one of claims 1 to 6.

8. The use of at least one of the data and samples includes a step of obtaining consent from the relevant parties. The method according to any one of claims 1 to 7.

9. The aforementioned request for consent may be made by using at least one of a questionnaire and / or an informed consent document, or by an avatar device. The method according to claim 8.

10. The step includes outputting information to individuals whose sample data is stored in the aforementioned database. The aforementioned information pertains to at least one of the following: pharmaceutical products, diseases, research findings, and recommended actions. The method according to any one of claims 1 to 9.

11. A search platform device configured to process search queries from a research entity targeting a database containing sample data from multiple individual samples, An input device that receives a search query containing at least one search parameter, A learning-type AI processor device connected to the input device, which searches for sample data of a selected sample from the sample data contained in the database, satisfying at least one of the search parameters included in the search query, An output device that outputs identification data of the sample data of at least one of the selected samples. Equipped with, The search query includes a request profile aimed at retrieving at least one sample containing sample data having at least one search parameter, The aforementioned AI processor device is Through individual training and evaluation of the search queries targeting the AI ​​processor device, the importance of samples and parameters according to the research entity is learned. A ranking of the selected samples is generated, ranging from important samples relevant to the search query to unimportant samples unrelated to the search query, based on the appropriate probability of the search query from the research entity. Based on the responses from the database and / or the search queries from the research entity, it is determined which of the additional parameters of the important sample are important to the research entity. By evaluating the search queries from the research entity, the identification data regarding which of the selected samples is worth considering for genome sequencing is provided. The output device further outputs the ranking of the selected samples and additional parameters for the selected samples. The aforementioned additional parameters match the search query so that a group of samples with a high probability of satisfying the requirements profile can be selected. Search platform device.

12. The AI ​​processor device searches for the group of sample data of the selected sample based on the search parameters included in the search query. The search platform device according to claim 11.

13. The AI ​​processor device obtains consent from the avatar device, which functions as the digital twin of the individual, when using the individual's data or samples. The search platform device according to claim 11.

14. The aforementioned additional parameters include at least one of the following: genetic profile, clinical medical data, medical history, familial predisposition to disease, personal diagnostic results, lifestyle characteristics, dietary habits, consumer behavior, exercise and physical activity characteristics, data on drug or other toxic substance intake, drug use data, radiation exposure data, geographical information, age, sex, ethnicity, allergies, and mental illness. A search platform device according to any one of claims 11 to 13.

15. The AI ​​processor device processes specialized information including at least one of the following: commercial market information from the industrial sector relating to pharmaceutical products, and information from an approval database relating to pharmaceutical products. A search platform device according to any one of claims 11 to 14.

16. The system includes an encoding device that uses an encoding function to convert the search query into a fragmented search query. A search platform device according to any one of claims 11 to 15.

17. The AI ​​processor device seeks the consent of the relevant parties regarding the use of at least one of the data and samples. A search platform device according to any one of claims 11 to 16.

18. The AI ​​processor device seeks the consent using at least one of a questionnaire and an informed consent document, or by an avatar device. The search platform device according to claim 17.

Images