Method for customising a biosensor to measure an analyte with high performance parameters by means of artificial intelligence

Artificial intelligence-based individualization of biosensors addresses manufacturing inconsistencies by determining quality and performance parameters, ensuring reliable analyte detection and efficient production.

WO2024209024A9PCT designated stage expired Publication Date: 2026-11-12DIGID GMBH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/EP2024/059290
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-04-05
Filing Date
2024-04-05
Publication Date
2026-11-12

AI Technical Summary

Technical Problem

Existing biosensors exhibit manufacturing variations leading to inconsistent performance across different application scenarios, necessitating a method to accurately assign them to suitable scenarios for reliable analyte detection.

Method used

A method involving artificial intelligence to determine quality and performance parameters of biosensors, using a trained neural network to individualize biosensors based on self-similarity of resistance bridges, allowing customization for specific applications.

Benefits of technology

Ensures consistent and reliable analyte detection by precisely assigning biosensors to appropriate scenarios, reducing manufacturing waste and improving test reliability through targeted production and logistics management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure EP2024059290_12112026_PF_FP_ABST
    Figure EP2024059290_12112026_PF_FP_ABST
Patent Text Reader

Abstract

The present invention relates to a method (100) for customising a biosensor (1) to measure a specific analyte, comprising: - determining at least one quality parameter of the biosensor (1), - determining at least one performance parameter and an associated threshold value of the biosensor (1) based on the at least one quality parameter using a trained first artificial intelligence (S2), and - customising the biosensor to measure a specific analyte by evaluating at least one of the performance parameters and the associated threshold value for a customised application scenario (S3).
Need to check novelty before this filing date? Find Prior Art

Description

[0001] Applicant: digid GmbH

[0002] Our reference: DID29174PCT

[0003] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0004] Date: April 5, 2024

[0005] Method for individualizing a biosensor for measuring an analyte with high performance parameters using artificial intelligence

[0006] Technical field

[0007] The present invention relates to a method for individualizing a biosensor for measuring a specific analyte in order to derive an improved, preferably best possible, qualitative and / or quantitative statement about the presence and / or concentration of the analyte(s) in the sample.

[0008] State of the art

[0009] A highly sensitive method for the detection and analysis of analytes in a sample consists of using an analyte-induced change in the surface tension of a miniaturized spring element, a so-called cantilever, to measure the presence and / or concentration of an analyte in a sample.

[0010] A sensor device for detecting the presence, concentration, and / or quantity of an analyte in a sample, comprising a sensor, connection electronics, and a housing, wherein the sensor is configured to convert chemical and / or biochemical information of an analyte in a sample into an electrical signal, wherein the sensor comprises a test cantilever having a base and a deformable part, wherein at least on the deformable part a receptor layer for selective uptake of the analyte is applied, and wherein the sensor comprises a reference cantilever having a base and a deformable part, wherein a reference layer for selective non-uptake of the analyte is applied on the deformable part, is known from WO 2022 / 200438 A1.

[0011] US patent 2022 / 0335817 A1 discloses a sensor device for detecting an environmental parameter and a method for determining information about the operating state of a sensor device.

[0012] Page 1 of 49 Applicant: digid GmbH

[0013] Our reference: DID29174PCT

[0014] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0015] Date: April 5, 2024

[0016] Description of the invention

[0017] To further improve the accuracy of analyte detection, a method for individualizing a biosensor for measuring a specific analyte with the features of claim 1 is proposed.

[0018] Accordingly, a method for individualizing a biosensor for measuring a specific analyte is proposed, comprising the steps of determining at least one quality parameter of the biosensor, determining at least one performance parameter and an associated threshold of the biosensor based on the at least one quality parameter using a trained first artificial intelligence, and individualizing the biosensor for measuring a specific analyte by evaluating at least one of the performance parameters and the associated threshold for an individual application scenario.

[0019] In this way, a biosensor can be customized for use in a specific application scenario during or after its manufacture.

[0020] During the production of biosensors, for example, in the manufacture of MEMS biosensors by growing MEMS structures onto a wafer, applying nanosensors, or immobilizing the biochemistry, statistical variations or fluctuations in the manufacturing process can lead to different properties of the biosensors. Thus, some biosensors may be better suited to one application scenario, while others may be better suited to a second.

[0021] The proposed method for individualizing biosensors allows them to be precisely assigned to an application scenario according to their properties, in which the individualized biosensor has good, preferably optimal, measurement accuracy.

[0022] The quality parameter can be determined based on a property of the biosensor, wherein at least one electrical resistance is determined as the property, preferably all electrical resistances of a resistance bridge of the biosensor.

[0023] If the biosensor includes at least one resistance bridge, at least one of the quality parameters can be a self-similarity of the electrical resistances of the resistance bridge.

[0024] Page 2 of 49 Applicant: digid GmbH

[0025] Our reference: DID29174PCT

[0026] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0027] Date: April 5, 2024

[0028] It has been found that the self-similarity of the resistances of the resistance bridge is a good indicator of the performance of the biosensor and can therefore be used as at least one of the quality parameters that can be used to individualize the respective biosensor.

[0029] At least one of the biosensor's performance parameters can include sensitivity, specificity, and / or yield. These performance parameters allow for a good description of a biosensor's performance for a specific analyte.

[0030] For the user of a biosensor, it is of great importance that the respective analysis results are consistent and reliable, or that the user receives an indication of the probability or certainty associated with the respective analysis results.

[0031] In other words, the user wants to know whether the test result is reliable. However, the expected level of reliability can vary depending on the application. For a virus test for humans, very high reliability may be required, for example, to reliably order or lift quarantine measures. For a home test for the virus as a preliminary test before visiting a doctor, or for testing an animal, such as a pet, the expected reliability may be lower.

[0032] To quantify the reliability of tests, parameters such as sensitivity and specificity are known, which can be used to classify a test.

[0033] For example, in a test for the presence of a specific virus, these parameters can be specified as performance parameters to describe the reliability of the test.

[0034] The sensitivity of a test indicates what proportion of the tested individuals are reliably identified as sick ("positive") by the test. Sensitivity is defined as the number of individuals correctly identified as sick by the test divided by the number of actually sick individuals in the test group. With high test sensitivity, very few sick individuals are falsely identified as healthy, meaning that a test with high sensitivity is highly likely to correctly identify a sick individual.

[0035] Page 3 of 49 Applicant: digid GmbH

[0036] Our reference: DID29174PCT

[0037] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0038] Date: April 5, 2024

[0039] The specificity of a test indicates what proportion of the tested individuals are reliably identified as healthy ("negative") by the test. Specificity is defined as the number of individuals correctly identified as healthy by the test divided by the number of truly healthy individuals in the test group. With high test specificity, very few healthy individuals are falsely identified as ill, meaning that a test with high specificity is highly likely to correctly identify a healthy individual as such.

[0040] Accordingly, these performance parameters are well suited to determining the performance of the biosensor.

[0041] The trained first artificial intelligence, for example a neural network, can determine at least one of the performance parameters and an associated threshold based on a feature vector, whereby the feature vector is obtained from a data processing of the at least one quality parameter.

[0042] The first trained artificial intelligence can be trained based on training data obtained through the following steps:

[0043] Determining a time-resolved measurement curve of at least one specific analyte and at least one sample without the specific analyte, each using a large number of biosensors;

[0044] Determining an individual measurement result and a validity parameter for each biosensor using a second artificial intelligence from the time-resolved measurement curve of the respective biosensor;

[0045] Determine at least one quality parameter for each biosensor;

[0046] Statistical evaluation of the individual measurement results based on the assigned validity parameters using threshold values, the assigned quality parameters, and the previously known result for each biosensor;

[0047] Determining performance parameters from statistical analysis;

[0048] Using quality parameters, performance parameters, and thresholds as training data.

[0049] Page 4 of 49 Applicant: digid GmbH

[0050] Our reference: DID29174PCT

[0051] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0052] Date: April 5, 2024

[0053] Time-resolved measurement curves can preferably be recorded at one or more defined and therefore known concentrations of the analyte. Training data can also be obtained for different specific analytes, for example, for different virus groups or bacteria.

[0054] Preferably, measurements are recorded using at least one defined measurement routine, for example, the measurement routine that is also used in the subsequent application of the respective type of biosensor. Different measurement routines, which may vary depending on the application scenario or the measuring apparatus used, can also be applied.

[0055] The time-resolved measurement curves can also be recorded in the event of at least one defined measurement disturbance, such as premature termination of the measurement, excessive temperature change during the measurement, or displacement of the biosensor during the measurement. Training data for various disturbances, such as those caused by temperature or interfering proteins, can also be generated by appropriately adjusting the samples and / or measurement routine.

[0056] This allows optimized training data to be obtained in order to have a representative data set for training a first trained artificial intelligence, so that in turn the trained first artificial intelligence can perform a precise individualization of a biosensor.

[0057] In particular, the trained first artificial intelligence can be trained on the basis of training data which includes, as quality parameters, the self-similarity of the electrical resistances of a resistance bridge of each biosensor.

[0058] The first trained artificial intelligence can be trained on the basis of training data that includes, as performance parameters, the sensitivity and / or the specificity and / or the yield of each biosensor.

[0059] By training the first artificial intelligence, a particularly good prediction of the suitability of the individual biosensor for a specific application scenario can be made, thus reliably enabling the individualization of the biosensor.

[0060] Page 5 of 49 Applicant: digid GmbH

[0061] Our reference: DID29174PCT

[0062] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0063] Date: April 5, 2024

[0064] The measurement of the analyte can be reliably achieved, or the most economically viable application of the biosensor can be determined.

[0065] Based on its individualization, the biosensor may undergo no or at least one further manufacturing step, wherein the no or at least one further manufacturing step preferably includes the storage of quality parameters and / or performance parameters and / or threshold values ​​on the biosensor or a database and / or the at least one further manufacturing step includes the functionalization of the biosensor, wherein the functionalization of the biosensor preferably includes the application of a receptor layer and / or reference layer.

[0066] The customization of biosensors makes it possible to assign biosensors to at least one suitable application scenario early in the manufacturing process or at any other time. Furthermore, a biosensor whose customization does not yield a suitable application scenario and / or whose performance parameters are insufficient or inconsistent can be rejected early and / or reworked accordingly. This saves manufacturing resources and / or materials. It also ensures that every biosensor assigned to an application scenario during the customization process is indeed suitable for that scenario, and preferably, ideally, optimally suited.

[0067] Furthermore, the individualization of biosensors, which are, for example, temporarily stored, enables a targeted and rapid continuation and / or reintegration of the individual biosensors into their manufacturing process. Individualizing the biosensors allows for the determination of a suitable application scenario for each one. Subsequently, each biosensor can potentially undergo at least one further manufacturing step, depending on its individual application scenario.

[0068] Data from the customization of biosensors, such as their performance parameters, application scenario, and / or threshold, can also be stored in a memory or even stored in the memory before intermediate storage and used to continue and / or reintegrate the individual biosensors into their manufacturing process. Thus, the customization of biosensors enables targeted and efficient control of biosensor production and logistics, including inventory management. A production and logistics management system, which, among other things,

[0069] Page 6 of 49 Applicant: digid GmbH

[0070] Our reference: DID29174PCT

[0071] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0072] Date: April 5, 2024

[0073] The manufacturing of biosensors can thus include and utilize the process for the individualization of biosensors. Furthermore, biosensors can be re-tested for performance after transport, if necessary, based on a prior individualization. Consequently, biosensors damaged during transport can then be sorted out and / or reworked through this testing.

[0074] The biosensor customization method is also suitable as a tool for reusing previously used biosensors to measure a specific analyte. Based on the application of this customization method, the biosensor's performance after use is assessed. If the biosensor's performance is still sufficiently reliable, it can undergo at least one further manufacturing or neutralization step before being used again to measure a specific analyte.

[0075] Furthermore, a method for generating training data for training a trained first artificial intelligence for use in the individualization of biosensors is proposed, with the following steps:

[0076] Determining a time-resolved measurement curve of at least one specific analyte, as well as at least one sample without the specific analyte, each using a large number of biosensors;

[0077] Determining an individual measurement result and a validity parameter for each biosensor using a second artificial intelligence from the time-resolved measurement curve of the respective biosensor;

[0078] Determine at least one quality parameter for each biosensor;

[0079] Statistical evaluation of the individual measurement results based on the assigned validity parameters using threshold values, the assigned quality parameters, and the previously known result for each biosensor;

[0080] Determining performance parameters from statistical analysis;

[0081] Page 7 of 49 Applicant: digid GmbH

[0082] Our reference: DID29174PCT

[0083] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0084] Date: April 5, 2024

[0085] Using quality parameters, performance parameters, and thresholds as training data.

[0086] Time-resolved measurement curves can preferably be recorded at one or more defined and therefore known concentrations of the analyte. Training data can also be obtained for different specific analytes, for example, for different virus groups or bacteria.

[0087] Preferably, measurements are recorded using at least one defined measurement routine, for example, the measurement routine that is also used in the subsequent application of the respective type of biosensor. Different measurement routines, which may vary depending on the application scenario or the measuring apparatus used, can also be applied.

[0088] The time-resolved measurement curves can also be recorded in the event of at least one defined measurement disturbance, such as premature termination of the measurement, excessive temperature change during the measurement, or displacement of the biosensor during the measurement. Training data for various disturbances, such as those caused by temperature or interfering proteins, can also be generated by appropriately adjusting the samples and / or measurement routine.

[0089] This allows optimized training data to be obtained in order to have a representative data set for training a first trained artificial intelligence, so that in turn the trained first artificial intelligence can perform a precise individualization of a biosensor.

[0090] This specific training data subsequently allows for the training of a trained first artificial intelligence in such a way that the individualization of biosensors can be reliably carried out.

[0091] The quality parameter determined for each biosensor can include the self-similarity of the resistors in a resistor bridge of each biosensor.

[0092] Furthermore, a biosensor for measuring a specific analyte is proposed, comprising at least one resistance bridge, wherein the biosensor is based on a result derived from the

[0093] Page 8 of 49 Applicant: digid GmbH

[0094] Our reference: DID29174PCT

[0095] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0096] Date: April 5, 2024

[0097] The resistance bridge is individualized for a specific quality parameter for a specific application scenario to measure the specific analyte.

[0098] The quality parameter can be the self-similarity of the resistors in the resistor bridge.

[0099] The biosensor envisioned here can be a biosensor for converting the chemical and / or biochemical information of an analyte into an electrical signal in a sample. The biosensor can comprise a test cantilever having a base and a deformable part, wherein at least on the deformable part a receptor layer for selective uptake of the analyte is applied, and a reference cantilever having a base and a deformable part, wherein on the deformable part a reference layer for selective non-uptake of the analyte is applied.

[0100] The base of the test cantilever and / or the base of the reference cantilever can be designed as a rigid base. A rigid base means that the respective cantilever—that is, the test cantilever and / or the reference cantilever—does not deform, or deforms only minimally, compared to the deformable part of the respective cantilever. The rigid base is, for example, connected to a substrate, supported by a substrate, or machined from the substrate. The deformable part of the test cantilever and / or the reference cantilever, on the other hand, is not supported by the substrate but rather protrudes beyond an edge of the substrate and is thus freely formed.

[0101] The deformable part of the test cantilever and / or the reference cantilever can, for example, be designed to be deflectable. Deflection of the respective cantilever can be achieved, for instance, around a flexural edge formed in a transition zone between the base and the deflectable area. The flexural edge is, for example, the edge of the substrate along which the cantilever is divided into the base and the deformable part.

[0102] However, the deformation of the respective cantilever in its deformable part is not limited to an upward or downward deformation; the cantilever itself can also be deformed, for example, bulged, wavy, or distorted.

[0103] In this context, a sample refers to a limited quantity of a substance that has been taken from a larger quantity of the substance, for example from a reservoir, whereby the composition of the sample is representative of the composition of the substance in the reservoir and accordingly consists of

[0104] Page 9 of 49 Applicant: digid GmbH

[0105] Our reference: DID29174PCT

[0106] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0107] Date: April 5, 2024

[0108] The presence and composition of the substance in the sample can be used to infer the corresponding presence in the reservoir.

[0109] For example, a sample can be a saliva sample, a blood sample, lymph, urine, sweat, interstitial fluid, or a swab, in particular a throat swab, a nasal swab, or a sinus swab, or a tissue sample. A sample includes, in particular, any type of biological sample, including samples from animals.

[0110] A sample can also be a non-biological sample, for example, a sample of a chemical substance. A gaseous sample is also possible.

[0111] An analyte is the substance whose presence in the sample is to be qualitatively and / or quantitatively detected, or detected by the biosensor. The analyte can be directly present in the sample, dissolved in the sample, or adhering to the sample or a part of the sample, particularly a sample particle. The analyte can also undergo a chemical, biological, and / or physical interaction with the sample, meaning that the analyte can only be detected indirectly via such an interaction.

[0112] In particular, one sample form can be converted into another sample form, allowing the analyte, or its presence, to be detected in a simple and reliable manner. For example, a swab can be dissolved in a liquid, so that the dissolved swab then becomes the actual sample, which is analyzed for the analyte.

[0113] The analyte in the sample can also be chemically pretreated, for example, by disrupting the viral envelope (if the analyte is a virus) to access nucleocapsid antigens. Furthermore, such pretreatment can also "label" the analyte to enhance the measurement signal. For this purpose, conjugated antibodies can bind to antigens of the analyte to generate maximum deformation on the cantilever system.

[0114] The sample then contains chemical and / or biochemical information about the analyte. Chemical information can include, for example, the type of analyte, its concentration, its occurrence, its weight, its reactivity, its density, etc. Biochemical information encompasses the same properties as chemical information; however, these substances can be further specified, for example, by...

[0115] Page 10 of 49 Applicant: digid GmbH

[0116] Our reference: DID29174PCT

[0117] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0118] Date: April 5, 2024

[0119] Biological processes occur. In particular, one speaks of biochemical information when the analyte has a special influence on the biological cycle, for example, metabolism or the immune system.

[0120] The test cantilever includes a passive and an active test transducer, and the reference cantilever includes a passive and an active reference transducer.

[0121] The chemical and / or biochemical information is converted into an electrical signal. This can mean that the chemical composition of the analyte can alter or generate an electrical signal. This can, for example, relate to the conductivity of a circuit. For instance, a first piece of biochemical information might be present if the circuit conducts, and a second piece of biochemical information might be present if the circuit does not conduct or exhibits reduced conductivity.

[0122] Besides the direct accessibility of the information, such as via conductivity, it is also possible to infer the biochemical information via a physical and / or chemical process and / or an interaction.

[0123] For this purpose, the proposed biosensor comprises a reference and a test cantilever. A cantilever is a spring element comprising a base and a deformable portion. The base is therefore a fixed part of the cantilever, which is typically mounted in a stationary position on a substrate. The deformable portion of the cantilever is located at the base and extends beyond it.

[0124] In particular, the base and the cantilever can be formed as a single unit. In other words, the deformable part of the cantilever is suspended from the base on one side. Because the deformable part extends beyond the substrate, it can be bent, deflected, and stretched. The spatial boundary at which the cantilever becomes bendable, or where the cantilever transitions from the base to the deformable part, is called the bending edge. The bending edge is typically an edge of the substrate when the cantilever extends beyond the base.

[0125] If the cantilever is bent, material stresses and forces occur in or on the cantilever material, which can be measured. If such material stress and / or force can be measured, it can be concluded that the cantilever is bent.

[0126] Page 11 of 49 Applicant: digid GmbH

[0127] Our reference: DID29174PCT

[0128] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0129] Date: April 5, 2024

[0130] The purpose of transducers is to determine or measure the deformation of the cantilevers. Active transducers are located on the deformable parts of the cantilevers, whereas passive transducers are located on the bases, for example, the cantilever bases. In particular, the electrical properties of a circuit can be influenced via the transducers.

[0131] For example, deformation of the cantilever can cause the resistance of a transducer, such as an active transducer, to increase, while no deformation of the cantilever results in no change in the transducer's resistance. This can be achieved, for instance, by designing the transducer according to the principle of a strain gauge, whereby deformation of the respective cantilever manifests itself as a change in the length of the strain gauge applied to it, and thus a deformation of the cantilever can be detected directly by a change in the resistance of the strain gauge.

[0132] Thus, the chemical and / or biochemical information of the analyte can be detected via a deformation of the cantilever, a subsequent registration via a transducer, and finally via a change in an electrical property of a circuit.

[0133] The deformation of the cantilever can be induced by a suitable coating specific to the material. For this reason, the reference cantilever has a reference layer for the selective non-uptake of the analyte, while the test cantilever has a receptor layer for analyte uptake.

[0134] A receptor layer is a substance that can interact with the analyte. This means that the receptor layer is specifically chosen for each analyte. Similarly, a reference layer is a substance that cannot interact with the analyte. The reference layer is also therefore specifically chosen for the analyte.

[0135] In this context, interaction means that the analyte is in chemical, biochemical, and / or physical interaction with the receptor layer. In particular, the interaction can consist of the analyte binding to the receptor layer. An interaction can also consist of the absorption, adsorption, or non-specific adhesion of the analyte to the receptor layer.

[0136] Page 12 of 49 Applicant: digid GmbH

[0137] Our reference: DID29174PCT

[0138] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0139] Date: April 5, 2024

[0140] The receptor and reference layers are preferably chemically identical with respect to potential interfering influences and preferably differ only in their interaction with the analyte. A substance that is not the analyte therefore interacts with the receptor layer with the same strength or weakness as with the reference layer.

[0141] The selective uptake of the analyte by the test cantilever causes a force to act upon it, making the test cantilever highly sensitive to the analyte. Consequently, other substances in the sample, excluding the analyte, contribute only to background noise in the form of a baseline deflection of the test cantilever. The force on the test cantilever increases, for example, more rapidly with increasing analyte concentration in the sample or with the rapid accumulation of analyte on the cantilever surface. The maximum force achievable for a given configuration is reached when the cantilever is completely covered.

[0142] The selective non-uptake of the analyte at the reference cantilever, on the other hand, means that no force is exerted by the analyte on the reference cantilever, so that only the substances that are not the analyte contribute to background noise in the form of a basic bending of the reference cantilever.

[0143] This force can cause deformation in the deformable part of the test cantilever, while the deformable part of the reference cantilever remains unchanged. The deflection of the cantilever is based on a change in surface tension due to interaction with the analyte. This change in surface tension leads to elongation or compression of the upper (or lower) surface of the cantilever. The differential elongation or compression on the upper and lower surfaces induces an internal force or stress in the material, resulting in deformation.

[0144] In the proposed biosensor, the selective non-uptake of the analyte by the reference cantilever drastically simplifies the measurement procedure, since the reference cantilever is not sensitive to the analyte and therefore the analyte does not contribute to the background noise. Only substances other than the analyte contribute to the background noise of the reference cantilever. In a sense, the selective non-uptake of the analyte by the reference cantilever ensures that the reference cantilever is exposed to the same turbulence, thermal drift, and the same influence of all non-analyte substances as in a reference sample. However, the difference is that the reference signal is directly in the

[0145] Page 13 of 49 Applicant: digid GmbH

[0146] Our reference: DID29174PCT

[0147] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0148] Date: April 5, 2024

[0149] The sample liquid, or, if applicable, the gaseous sample at the relevant point, is determined.

[0150] In particular, a reference cantilever with a reference layer and a test cantilever with a receptor layer results in a significantly more specific analysis of the analyte than a reference cantilever without a receptor layer, since both the reference layer and the receptor layer exhibit a specific interaction or non-interaction with the analyte.

[0151] The design of the biosensor with a reference cantilever and a test cantilever offers the advantage that two measurements can be performed simultaneously in the sample, with the measurement of the reference cantilever calibrating the measurement of the test cantilever. This allows environmental influences, such as chemical, thermal, mechanical, electrical, and fluidic interferences, to be reduced in relation to each individual measurement, so that the presence of the analyte can be inferred from a comparison of the measurements taken at the test cantilever and the reference cantilever.

[0152] These forces or material stresses, such as strains or compressions acting on the cantilevers, can ultimately be detected by the transducers, whereby different strains or compressions result in different stresses being detected by the transducers.

[0153] Brief description of the characters

[0154] Preferred further embodiments of the invention are explained in more detail by the following description of the figures.

[0155] This shows:

[0156] Figure 1 shows a first schematic embodiment of an exemplary biosensor for converting chemical and / or biochemical information;

[0157] Figure 2 shows another schematic embodiment of a biosensor;

[0158] Figure 3 shows a schematic way of connecting the electrodes of a biosensor in the form of a resistor bridge;

[0159] Page 14 of 49 Applicant: digid GmbH

[0160] Our reference: DID29174PCT

[0161] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0162] Date: April 5, 2024

[0163] Figure 4 shows a schematic flowchart of a process for individualizing a custom-made biosensor for measuring a specific analyte;

[0164] Figure 5A shows a schematic embodiment for determining voltages and / or electrical resistances at a resistance bridge of a biosensor using a measuring device, here in the form of a schematically shown multimeter;

[0165] Figure 5B shows an embodiment for determining at least one property of the biosensor using a schematically shown camera system;

[0166] Figure 6A shows a schematic neural network as an embodiment of a trained first artificial intelligence for determining performance parameters and an associated threshold;

[0167] Figure 6B shows a schematic flowchart comprising a second artificial intelligence for generating training data;

[0168] Figure 7 shows a schematic representation of a training data set comprising the performance parameters sensitivity, specificity and yield as a function of the self-similarity of a resistance bridge and a threshold;

[0169] Figure 8 shows a schematic flowchart as an example of an evaluation of a biosensor measurement based on its individualization using a second artificial intelligence; and

[0170] Figure 9 shows a schematic flow chart of an exemplary embodiment for going through no or at least one further manufacturing step of a biosensor based on its individualization.

[0171] Detailed description of preferred implementation examples

[0172] Preferred embodiments are described below with reference to the figures. Identical, similar, or equivalent elements in the different figures are designated with identical reference numerals, and repeated descriptions of these elements are sometimes omitted to avoid redundancy.

[0173] Page 15 of 49 Applicant: digid GmbH

[0174] Our reference: DID29174PCT

[0175] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0176] Date: April 5, 2024

[0177] Figure 1 schematically shows a first embodiment of an exemplary biosensor 1 for converting chemical and / or biochemical information. The biosensor 1 comprises a test cantilever 2, which in turn has a base 20 and a deformable part 22. A passive test transducer 200 is arranged on the base 20, while an active test transducer 220 is arranged on the deformable part 22.

[0178] Analogous to the test cantilever 2, the biosensor 1 also has a reference cantilever 3, which in turn has a base 30 with a passive reference transducer 300, as well as a deformable part 32 which has an active reference transducer 320.

[0179] The purpose of transducers is to determine or measure the deformation of the cantilevers. Active transducers are located on the deformable parts of the cantilevers, whereas passive transducers are located on the bases, for example, the cantilever bases. In particular, the electrical properties of a circuit can be influenced via the transducers.

[0180] For example, deformation of the cantilever can cause the resistance of a transducer, such as an active transducer, to increase, while no deformation of the cantilever results in no change in the transducer's resistance. This can be achieved, for instance, by designing the transducer according to the principle of a strain gauge, whereby deformation of the respective cantilever manifests itself as a change in the length of the strain gauge applied to it, and thus a deformation of the cantilever can be detected directly by a change in the resistance of the strain gauge.

[0181] Thus, the chemical and / or biochemical information of the analyte can be detected via a deformation of the cantilever, a subsequent registration via a transducer, and finally via a change in an electrical property of a circuit.

[0182] The transducer can also detect a contraction of the surface on which it is mounted. However, in the illustrated embodiments, the transducers are always mounted on surfaces where strain is expected.

[0183] However, the strain and / or change in surface tension and / or force detected by the transducer can also be a bending force or a shear force, or be caused by a bending force or shear force, or generally be based on the elastic modulus of the

[0184] Page 16 of 49 Applicant: digid GmbH

[0185] Our reference: DID29174PCT

[0186] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0187] Date: April 5, 2024

[0188] The respective cantilevers are based on this. In particular, the attachment of the deformable part 22, 32 to the base 20, 30 results in the deformable part 22, 32 aligning itself along a bending curve due to a force acting upon it, which is caused by a change in the surface tension of the test cantilever. The resulting bending curve is determined in particular by the geometry, especially the area moment of inertia of the cantilever, as well as by the mass of the cantilever and the modulus of elasticity. The bending curve can be described, for example, according to beam theory.

[0189] The transducers 200, 220, 300, 320 are each connected to an electronics unit 4, which is capable of recording or forwarding a measurement signal from the transducers 200, 220, 300, 320, while the electronics unit 4 is also capable of supplying the transducers 200, 220, 300, 320 with current and / or voltage.

[0190] The transducers 300, 320, 200, and 220 are preferably contacted via electrodes 401, 402, 403, and 404. Specifically, the active test transducer 220 is connected to the active reference transducer 320 via electrode 401. Furthermore, the passive test transducer 200 is connected to the passive reference transducer 300 via electrode 403. The active test transducer 220 is also connected to the passive test transducer 200 via electrode 402, while the active reference transducer 320 is connected to the passive reference transducer 300 via electrode 404. Thus, a total of four electrodes are used to electrically contact the transducers.

[0191] Electrical contact can be achieved in particular by applying the transducers to existing electrodes, thus creating a conductive connection.

[0192] The biosensor 1 has the task of indicating the presence and / or concentration and / or quantity of an analyte 90 in a sample 9.

[0193] In Figure 1, sample 9 is a liquid obtained, for example, by taking a swab, in particular a nasal or throat swab, from a test subject. It is also conceivable that the sample comprises a liquid on an object, such as a string or a cotton swab. However, sample 9 could also be saliva, blood, or another bodily fluid. It could also be a gargling solution used by the test subject. Finally, sample 9 could be taken from a tissue sample or from another specimen obtained from the sample.

[0194] Page 17 of 49 Applicant: digid GmbH

[0195] Our reference: DID29174PCT

[0196] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0197] Date: April 5, 2024

[0198] The substance was obtained from the test subject and / or synthesized. Analyte 90 may be dissolved in the sample or present in an undissolved form as a suspension, dispersion, or emulsion.

[0199] However, sample 9 and the analyte of interest 90 can also be of non-biological origin.

[0200] In any case, the biosensor 1 is intended to analyze sample 9 with regard to the presence, concentration, and / or quantity of analyte 90 in the sample 9. For this purpose, a receptor layer 24 is applied to the test cantilever 2, with which analyte 90 can interact. For example, the receptor layer 24 can adsorb or absorb the analyte 90. In the case of adsorption, the analyte 90 would adhere to the surface of the receptor layer 24, while in the case of absorption, the analyte 90 would penetrate into the interior of the receptor layer 24.

[0201] If sample 9 contains a specific analyte 90, this analyte can interact with the receptor layer 24. This can cause the surface tension of the section of the deformable part 22 of the test cantilever 2 covered by the receptor layer 24 to change. This change in surface tension can be registered by the active test transducer 220, which is then interpreted as a measurement signal in the electronics 4.

[0202] The change in the surface tension of the section of the deformable part 22 of the test cantilever 2 covered with the receptor layer 24 can also lead to a deformation of the deformable part 22 of the test cantilever 2. The active test transducer 220 can therefore also register a deformation of the deformable part of the test cantilever 2, which is in turn interpreted as a measurement signal in the electronics 4.

[0203] However, the interaction with the sample liquid of sample 9 can already lead to the registration of a force by the active test transducer 220, for example, if only the surface tension of the liquid acts on the deformable part 22 of the test cantilever 2. Therefore, in this case, the presence of an analyte 90 is not responsible for any resulting deformation or change in surface tension.

[0204] To determine the magnitude of this basic effect of sample 9 on the test cantilever 2, the reference cantilever 3 is simultaneously brought into contact with sample 9 along with the test cantilever 2.

[0205] Page 18 of 49 Applicant: digid GmbH

[0206] Our reference: DID29174PCT

[0207] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0208] Date: April 5, 2024

[0209] For this purpose, the reference cantilever 3 has a reference layer 34 with which an analyte 90 cannot interact, or rather, a reference layer 34 that cannot adsorb or absorb the analyte 90. This avoidance of interaction with the analyte 90 is intended to allow differentiation from the measurement signal of the test cantilever 2.

[0210] Since both the test cantilever 2 and the reference cantilever 3 interact with the sample 9, both cantilevers 2 and 3 interact with the sample 9 in a similar manner. The difference here, however, is that the test cantilever 2 can additionally interact with any analyte 90 that may be present via its reference layer 24.

[0211] Accordingly, the measurement signals of the active transducers 220 and 320 differ if analyte 90 is present in sample 9. In the simplest case, the magnitude of the difference in the measurement signals can therefore be used to determine the amount of analyte 90 present in sample 9.

[0212] The deformation of the cantilever can be induced by a suitable coating specific to the material. For this reason, the reference cantilever has a reference layer for the selective non-uptake of the analyte, while the test cantilever has a receptor layer for analyte uptake.

[0213] A receptor layer is a substance that can interact with the analyte. This means that the receptor layer is specifically chosen for each analyte. Similarly, a reference layer is a substance that cannot interact with the analyte. The reference layer is also therefore specifically chosen for the analyte.

[0214] In this context, interaction means that the analyte is in chemical, biochemical, and / or physical interaction with the receptor layer. In particular, the interaction can consist of the analyte binding to the receptor layer. An interaction can also consist of the absorption, adsorption, or non-specific adhesion of the analyte to the receptor layer.

[0215] The receptor and reference layers are preferably chemically identical with respect to potential interfering influences and preferably differ only in their interaction with the analyte. A substance that is not the analyte therefore interacts with the receptor layer with the same strength or weakness as with the reference layer.

[0216] Page 19 of 49 Applicant: digid GmbH

[0217] Our reference: DID29174PCT

[0218] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0219] Date: April 5, 2024

[0220] The selective uptake of the analyte by the test cantilever causes a force to act upon it, making the test cantilever highly sensitive to the analyte. Consequently, other substances in the sample, excluding the analyte, contribute only to background noise in the form of a baseline deflection of the test cantilever. The force on the test cantilever increases, for example, more rapidly with increasing analyte concentration in the sample or with the rapid accumulation of analyte on the cantilever surface. The maximum force achievable for a given configuration is reached when the cantilever is completely covered.

[0221] The selective non-uptake of the analyte at the reference cantilever, on the other hand, means that no force is exerted by the analyte on the reference cantilever, so that only the substances that are not the analyte contribute to background noise in the form of a basic bending of the reference cantilever.

[0222] This force can cause deformation in the deformable part of the test cantilever, while the deformable part of the reference cantilever remains unchanged. The deflection of the cantilever is based on a change in surface tension due to interaction with the analyte. This change in surface tension leads to elongation or compression of the upper (or lower) surface of the cantilever. The differential elongation or compression on the upper and lower surfaces induces an internal force or stress in the material, resulting in deformation.

[0223] State-of-the-art reference cantilevers lack a receptor layer that is sensitive to the analyte. While this allows for the determination of effects such as sample turbulence and the thermal drift of the sensor system, the analyte can bind to the reference layer of such a cantilever through non-specific binding. This means the analyte itself contributes to background noise. Therefore, with a state-of-the-art sensor, reference measurements in a reference sample—that is, a sample without analyte—are necessary. Only in this way can the effect of non-analyte binding be detected.

[0224] In this biosensor, the selective non-uptake of the analyte by the reference cantilever drastically simplifies the measurement procedure, since the reference cantilever is not sensitive to the analyte and therefore the analyte does not contribute to the background noise. Only the

[0225] Page 20 of 49 Applicant: digid GmbH

[0226] Our reference: DID29174PCT

[0227] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0228] Date: April 5, 2024

[0229] Substances other than the analyte contribute to the background noise of the reference cantilever. In a sense, by selectively excluding the analyte from the reference cantilever, it can be ensured that the reference cantilever is exposed to the same turbulence, thermal drift, and the same influence of all non-analyte substances as in a reference sample. However, the difference is that the reference signal is determined directly in the sample liquid.

[0230] In particular, a reference cantilever with a reference layer and a test cantilever with a receptor layer results in a significantly more specific analysis of the analyte than a reference cantilever without a receptor layer, since both the reference layer and the receptor layer exhibit a specific interaction or non-interaction with the analyte.

[0231] The sensor design with a reference cantilever and a test cantilever offers the advantage of allowing two measurements to be taken simultaneously in the sample, with the measurement from the reference cantilever calibrating the measurement from the test cantilever. This reduces the influence of environmental factors, such as chemical, thermal, mechanical, electrical, and fluidic disturbances, on the respective measurement, enabling the presence of the analyte to be determined by comparing the measurements at the test cantilever and the reference cantilever.

[0232] These forces or material stresses, such as strains or compressions acting on the cantilevers, can ultimately be detected by the transducers, whereby different strains or compressions result in different stresses being detected by the transducers.

[0233] The test cantilever 2 and the reference cantilever 3 thus measure the presence of analyte 90 in sample 9, but at different positions. Different properties of the sample can occur at different positions, such as temperature fluctuations or concentration gradients. These different environmental conditions can be measured with the passive transducers 200 and 300.

[0234] The passive transducers 200, 300 are arranged on the base and preferably do not detect a measurement signal when there is a deformation or change in the surface tension of the deformable part 22, 32 of the reference or test cantilever 2, 3. However, the base level of the

[0235] Page 21 of 49 Applicant: digid GmbH

[0236] Our reference: DID29174PCT

[0237] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0238] Date: April 5, 2024

[0239] The measurement signals of the passive transducers 200 and 300 are affected by these different environmental conditions. By providing a reference value for each measurement of the active transducers 220 and 230 via the passive transducers 200 and 300, which considers the environmental conditions in isolation, the influence of the environmental conditions on the measurement signals of the active transducers 220 and 320 can be determined and reduced, factored out, or isolated.

[0240] Accordingly, the presence of analyte 90 in a sample 9 can be isolated and analyzed using the biosensor 1 by reducing and isolating the influence of interactions not attributable to analyte 90 through a large number of measurement points on the reference and test cantilevers 3, 2. This enables a high measurement accuracy of the presence of analyte 90 in the sample 9.

[0241] Figure 2 shows another schematic embodiment of a biosensor 1. The transducers 300, 320, 200, 220 are contacted via the electrodes 401, 402, 403, 404.

[0242] In particular, the active test transducer 220 is connected to the active reference transducer 320 via electrode 401. Furthermore, the passive test transducer 200 is connected to the passive reference transducer 300 via electrode 403. The active test transducer 220 is also connected to the passive test transducer 200 via electrode 402, whereas the active reference transducer 320 is connected to the passive reference transducer 300 via electrode 404.

[0243] This results in a total of four electrodes via which the transducers 200, 220, 300, and 320 are electrically contacted. Electrical contact can be achieved, in particular, by placing the transducers onto the electrodes, thus creating a conductive connection.

[0244] Figure 3 shows one possible connection of the electrodes. The electrodes that contact transducers 200, 220, 300, and 320 are all arranged in a mirror-symmetrical configuration. Currents flow through the electrodes, or voltages are applied, so that an asymmetrical design of these electrodes could lead to asymmetrical crosstalk of electrical signals to the other electrodes. This mutual interference can generate a stray signal between the electrodes, which, however, can be avoided by the symmetrical design. Consequently, voltage drops occur across the electrical resistances of transducers 200, 220, 300, and 320, such as...

[0245] Page 22 of 49 Applicant: digid GmbH

[0246] Our reference: DID29174PCT

[0247] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0248] Date: April 5, 2024

[0249] a single-ended voltage at the passive test transducer Utest passive (200), as well as a single-ended voltage at the passive reference transducer Ureterence passive poo).

[0250] The transducers 200, 220, 300, and 320 are electrically connected in a so-called resistive bridge. In this resistive bridge, an excitation voltage VO, for example, DC or AC, is applied between electrodes 403 and 401. Between these electrodes, the passive and active transducers act as voltage dividers due to their electrical resistances. A resistive bridge in the form shown has the advantage that no voltage is generated between electrodes 402 and 404, provided that the ratio of the electrical resistances of the passive transducer 200 to the active transducer 220 of the test cantilever 2 is equal to the ratio of the electrical resistances of the passive transducer 300 to the active transducer 320 of the reference cantilever 3. Therefore, a deviation of just one electrical resistance is sufficient to change the electrical resistance ratios and thus generate a voltage between electrodes 402 and 404.

[0251] When the reference cantilever 3 and the test cantilever 2 interact with the sample 9 and the analyte 90, both deformable parts 22, 32 experience, for example, a change in surface tension, which is greater for the deformable part 22 of the test cantilever 2 than for the deformable part 32 of the reference cantilever 3. Consequently, the electrical resistance of the active test transducer 220 of the deformable part 22 of the test cantilever 2 will vary more than that of the active reference transducer 320 of the deformable part 32 of the reference cantilever 3. If the electrical resistances of the passive transducers 200, 300 do not change or at least change in the same way, a change in the electrical resistance ratios results from the deformation of the deformable part 22 of the test cantilever 2 due to the interaction with the analyte 90 of the sample 9, which is specifically related to the reference layer 24 of the Test cantilevers 2 interacts.In such an interaction, a voltage is accordingly established between the electrodes 402 and 404, so that a change in the surface tension of the active test transducer 220 relative to the active reference transducer 320 can be displayed as the bridge cross-voltage VB. Preferably, the bridge cross-voltage VB scales with the abundance of analyte 90 in the sample 9, thus enabling a quantitative evaluation of the measurement signal.

[0252] The bridge cross-voltage VB corresponds to the difference between the single-ended voltage at the passive test transducer Utest passive (200) and the single-ended voltage at the passive reference transducer Ureterence passive pooj.

[0253] Page 23 of 49 Applicant: digid GmbH

[0254] Our reference: DID29174PCT

[0255] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0256] Date: April 5, 2024

[0257] A bridge cross-voltage detector 44 can display or transmit the bridge cross-voltage VB externally, so that the user of the biosensor 1 can see that a bridge cross-voltage VB is present. In particular, such a bridge cross-voltage detector 44 can also be provided by an analog-to-digital converter 440, wherein the analog-to-digital converter 440 converts the bridge cross-voltage VB into a digital signal that can be transmitted to the external measuring device.

[0258] In particular, the AD converter 440 can be operated in two different measurement modes.

[0259] The first measurement mode is the differential measurement mode in which the bridge cross-sectional stress VB is measured, thus generating a relative measurement value for the deformation of the two reference and test cantilevers 3, 2. In this differential measurement mode, the measurement signal of all transducers 200, 220, 300, 320 is taken into account, so that the output signal of the A / D converter 440 is a measurement signal cleaned of environmental influences, from which the relative deformation of the deformable parts 32, 22 and thus the presence of an analyte 90 can be inferred.

[0260] The second measurement mode is the so-called absolute measurement mode. In the absolute measurement mode, the bridge transverse voltage VB is not detected; instead, the signals at electrodes 402 and 404 are isolated from each other, allowing a statement to be made about the respective deflections of the deformable parts 32 and 22. This information is unavailable to the user in the differential measurement mode, in which the bridge transverse voltage VB is measured.

[0261] The biosensors actually produced may exhibit a certain degree of variation with regard to the applied structures and thus, in particular, with regard to the properties of the transducers, especially the active and passive test transducers 220, 200, as well as the active and passive reference transducers 320, 300. These differences in properties, for example, in the electrical resistances of the transducers 200, 220, 300, 320, are due to the manufacturing process and therefore cannot be completely ruled out.

[0262] However, for the user of a biosensor 1, it is of great importance that the respective analysis results are consistent and reliable, or that the user receives an indication of the probability or certainty associated with the respective analysis results.

[0263] Page 24 of 49 Applicant: digid GmbH

[0264] Our reference: DID29174PCT

[0265] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0266] Date: April 5, 2024

[0267] In other words, the user wants to know whether the test result is reliable. However, the expected level of reliability can vary depending on the application. For a virus test for humans, very high reliability may be required, for example, to reliably order or lift quarantine measures. For a lactate measurement for a recreational athlete, the expected reliability may be lower.

[0268] To quantify the reliability of tests, parameters such as sensitivity and specificity are known, which can be used to classify a test.

[0269] For example, in a test for the presence of a specific virus, these performance parameters can be specified to describe the reliability of the test.

[0270] The sensitivity of a test indicates what proportion of the tested individuals are reliably identified as sick ("positive") by the test. Sensitivity is defined as the number of individuals correctly identified as sick by the test divided by the number of actually sick individuals in the test group. With high test sensitivity, very few sick individuals are falsely identified as healthy, meaning that a test with high sensitivity is highly likely to correctly identify a sick individual.

[0271] The specificity of a test indicates what proportion of the tested individuals are reliably identified as healthy ("negative") by the test. Specificity is defined as the number of individuals correctly identified as healthy by the test divided by the number of truly healthy individuals in the test group. With high test specificity, very few healthy individuals are falsely identified as ill, meaning that a test with high specificity is highly likely to correctly identify a healthy individual as such.

[0272] The following describes a method by which the biosensors described above can be customized to measure a specific analyte, and the measurement method is then designed in such a way that the test results always provide a measurement result with high sensitivity and high specificity, despite possible variations in the production of the biosensors.

[0273] Figure 4 shows a schematic flow chart of an embodiment of a method 100 for individualizing a custom-made biosensor 1 for measuring a

[0274] Page 25 of 49 Applicant: digid GmbH

[0275] Our reference: DID29174PCT

[0276] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0277] Date: April 5, 2024

[0278] specific analytes 90. The individualization of the manufactured biosensor 1 is to be carried out in such a way that, after successful individualization, the user can use the biosensor 1 for a specific application scenario. In the corresponding application scenario, the biosensor 1 is then designed to perform the corresponding analysis in such a way that the reliability requirements expected by the user for the specific application scenario, for example, sensitivity and specificity, are always met, or the test is rejected. In other words, a reliable test result can always be obtained in this way.

[0279] In procedure 100, in a first step S1, at least one quality parameter of the individual biosensor 1, whose individualization is to be achieved, is determined. This at least one quality parameter can be used as an indicator of the performance of the biosensor 1. At a minimum, this at least one quality parameter serves to individualize the biosensor in the further course of procedure 100.

[0280] At least one quality parameter of the individual biosensor 1 is derived, for example, from the production data of the biosensor 1, which were recorded during its production, or from a measurement. The quality parameters can be determined directly from the production data or the measurement data, or derived from them. An exemplary determination of measurement data or of quality parameters from measurement data or production data is described below, for example, in Figures 5A and 5B. Reference is made to these figures.

[0281] In a second step S2, at least one performance parameter and a corresponding threshold are determined for the individual biosensor 1 based on the provided quality parameters. The threshold preferably corresponds to a probability used to assess the accuracy of a measurement result obtained with the biosensor. The accuracy can be either correct or invalid. The corresponding threshold depends on the performance parameters of the biosensor 1. Performance parameters can include, for example, the sensitivity, specificity, and / or yield of the respective biosensor 1 for measuring a specific analyte 90. Sensitivity and specificity indicate how reliably a medical diagnostic procedure, such as the measurement of a specific analyte 90 in a user, detects whether or not the user is ill. The yield, in turn, indicates the extent to which a medical...

[0282] Page 26 of 49 Applicant: digid GmbH

[0283] Our reference: DID29174PCT

[0284] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0285] Date: April 5, 2024

[0286] Diagnostic procedures, such as measuring a specific analyte in a user, provide a correct result.

[0287] The performance parameters are determined using a trained first artificial intelligence (CA), which calculates the performance parameters and thresholds of each individual biosensor 1 based on its quality parameters. After inputting at least one quality parameter, the trained CA will output a list of performance parameters and their corresponding thresholds.

[0288] For example, the determination in step S2 results in a list of tuples for each biosensor 1, containing performance parameters and corresponding threshold values ​​based on at least one quality parameter of the respective biosensor 1. Each tuple in the list of tuples includes a value and a corresponding threshold value for each performance parameter. With a suitable choice of performance parameters, the tuple allows for an assessment of the accuracy of a measurement by the biosensor 1 based on its corresponding threshold value.

[0289] As shown in embodiment 6A, a neural network 80 can be used as the first trained artificial intelligence for determining performance parameters. The training of the first artificial intelligence is described below following Figure 6A. Reference is made to this section.

[0290] The training of the first artificial intelligence specifies its capabilities for the task in step S2. This allows the first artificial intelligence used in step S2 to determine lists for one or more performance parameters with their respective thresholds for each, or a specific, analyte. Performance parameters such as the sensitivity, specificity, and / or yield of the respective biosensor 1 for measuring a specific analyte 90 can be determined by the trained first artificial intelligence.

[0291] In a third step S3, based on the determined performance parameters and the assigned threshold value, the individual biosensor 1 is then individualized in such a way that the performance parameters and their threshold values ​​are analyzed for different application scenarios and especially analytes 90 to determine whether the individual

[0292] Page 27 of 49 Applicant: digid GmbH

[0293] Our reference: DID29174PCT

[0294] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0295] Date: April 5, 2024

[0296] Is Biosensor 1 suitable for a specific application scenario, or rather for a different application scenario?

[0297] For example, in an initial application scenario, biosensor 1 is to be used to detect a virus to contain a pandemic. In this case, analyte 90 is clearly specified, as this is the specific virus present in the sample liquid of sample 9. Furthermore, to be applicable in this case, the test must achieve high sensitivity and high specificity. For this initial application scenario, the requirements profile is such that analyte 90, in the form of the virus, must be determined with high specificity and high sensitivity, while also considering the yield. High sensitivity and high specificity are of paramount importance here, among other reasons, to both contain the pandemic and to reliably implement official measures restricting individual freedom, such as quarantine measures.The test typically exhibits both a specificity and a sensitivity of 100%. A high yield is important for resource conservation and user satisfaction, ensuring a sufficient number of tests are available whose accuracy in measuring the specific analyte is not invalidated.

[0298] In a second exemplary application scenario, for instance, the determination of the lactate level in the blood of a recreational athlete might be required. Here, too, the analyte 90 is clearly specified: lactate in the athlete's blood. However, the requirements for the sensitivity and specificity of the corresponding biosensor are not as high in this case, since the recreational athlete needs the data for individual training management, but this does not involve any substantial restrictions on personal freedom, such as those imposed by quarantine measures. Accordingly, in this second application scenario, the analyte would be clear: lactate in the blood, but the requirements for specificity and sensitivity are significantly lower.

[0299] Specifically, in step S3, for example, it can be determined, based on the performance parameters determined in step S2 and the assigned threshold values, whether an individual biosensor 1 is suitable for a virus test to contain a pandemic, which requires high sensitivity and high specificity, or whether the individual biosensor 1 is more suitable for an application scenario where high specificity and high sensitivity are not required.

[0300] Page 28 of 49 Applicant: digid GmbH

[0301] Our reference: DID29174PCT

[0302] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0303] Date: April 5, 2024

[0304] If a list of tuples is available, one tuple can be selected from the list. The selection of the tuple can preferably be based on the values ​​of the performance parameters in the list of tuples, taking the application scenario into account, in order to customize the biosensor.

[0305] The threshold value in the selected tuple can later be used to assess the accuracy of the measurement of the specific analyte by the biosensor. The threshold value depends on the performance parameters. This is also discussed in the description of Figure 7 below. Reference is made to this section. Accordingly, the tuple, and in particular the threshold value for the respective performance parameter, can also be advantageously stored with the biosensor 1.

[0306] After completion of the individualization, the process ends and the individual, now individualized biosensor can be marked and used according to the respective defined application scenario.

[0307] In addition, the application scenario and the associated threshold value can be stored in the respective individual, now individualized biosensor.

[0308] Of course, with a corresponding evaluation of the individual biosensor, several different application scenarios with their associated threshold values ​​can be determined and stored, which can then be selected and stored in the biosensor.

[0309] The data, including the relevant application scenario and its associated performance parameters and thresholds, can be stored in non-volatile memory within the biosensor, such as an EEPROM. Storing this data allows for the customization of each individual biosensor.

[0310] The individualization of the biosensor for measuring a specific analyte can also take place in step S3 using the trained first artificial intelligence from step S2, taking into account a specific application scenario. The trained first artificial intelligence thus determines the performance parameter values ​​and the corresponding threshold for biosensor 1 based on at least one quality parameter, also taking into account the specific application scenario. For example, the trained first artificial intelligence selects a tuple from a list of tuples based on the specific application scenario.

[0311] Page 29 of 49 Applicant: digid GmbH

[0312] Our reference: DID29174PCT

[0313] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0314] Date: April 5, 2024

[0315] Application scenarios are selected. The performance parameter values ​​and the corresponding threshold in the selected tuple depend on the specific application scenario, and the trained first artificial intelligence (AI) customizes biosensor 1 accordingly. The specific application scenario can, for example, be defined by a user or determined based on current stock levels and demand and provided to the trained AI.

[0316] If the trained first artificial intelligence cannot select a suitable tuple because, for example, no performance parameter values ​​exist for biosensor 1 in the specific application scenario, the first artificial intelligence can issue an error message. In this case, the trained artificial intelligence can also independently select a different specific application scenario, for example, based on current inventory levels and / or demand. The first artificial intelligence incorporates the specific application scenario to define a performance parameter with a corresponding threshold value for biosensor 1. Accordingly, the training of the first artificial intelligence takes the specific application scenario into account. The consideration of the specific application scenario during the training of the first artificial intelligence is explained in more detail at the end of Figure 6B. Reference is made to this section.

[0317] The determination of at least one quality parameter by measuring measured values ​​or evaluating production data is shown below in Figures 5A and 5B as an example.

[0318] Figure 5A schematically shows an embodiment for determining a property of the biosensor 1 based on an embodiment of a biosensor 1 comprising a test cantilever 2 and a reference cantilever 3 with deformable parts 22, 32. Using a measuring device, here exemplified by a multimeter 50, the voltages across the electrical resistances of the transducers 300, 320, 200, 220 are measured, such as the single-ended voltage across the passive test transducer Utest passive (200) and the single-ended voltage across the passive reference transducer Ureferenz passive (poo). The transducers 300, 320, 200, 220 are contacted via the electrodes 401, 402, 403, 404. For this purpose, measuring probes 52 of the multimeter 50 are positioned accordingly at the electrodes of the biosensor 1. In this embodiment, the measured voltage value across the electrical resistance is displayed on a screen of the multimeter 50.The multimeter 50 can also be used to measure other properties of the biosensor 1, such as electrical properties.

[0319] Page 30 of 49 Applicant: digid GmbH

[0320] Our reference: DID29174PCT

[0321] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0322] Date: April 5, 2024

[0323] Resistances, capacitances, or impedances. For example, the multimeter can be used to measure the electrical resistances of transducers 300, 320, 200, and 220. The Multimeter 50 can also be part of a measuring apparatus, particularly during or after the manufacturing process of the biosensor, allowing the measurement of at least one property to be automated.

[0324] Figure 5B shows a further embodiment for determining at least one property of the subcomponents or all subcomponents of the biosensor 1 using a camera system 60. The camera system 60 can use an imaging technique to capture a section of the biosensor 1, generate a spectrum of the image section, perform a time-resolved analysis, and evaluate spectrally filtered excitation and detection of signals using the camera and, for example, a computer 62. The evaluation can be performed at any stage of production and may include the use of electrons, light, ions, and / or near-field analysis.

[0325] The image analysis can, for example, reveal the electrical resistances of transducers 300, 320, 200, and 220 as a property of biosensor 1. Transducers 300, 320, 200, and 220 are connected via electrodes 401, 402, 403, and 404, respectively. The image analysis can also reveal the coating density and homogeneity of a test cantilever 2 and a reference cantilever 3 as properties of biosensor 1. Furthermore, the image analysis can provide the individual, high-resolution geometric configuration of the cantilever, the electrodes, or the transducer, and may also include spectroscopic information.

[0326] From the previously determined properties of the biosensor, at least one quality parameter can then be determined.

[0327] For example, the proportion and distribution of the functionalized surface of biosensor 1 can be determined as a quality parameter.

[0328] From the previously determined single-ended voltages at the electrical resistances of the passive test transducer Utest passive (200), the passive reference transducer Ureferenz passive poo) and the excitation voltage V0, the self-similarity of the electrical resistances in the resistor bridge of the biosensor 1 can also be determined as a quality parameter.

[0329] Page 31 of 49 Applicant: digid GmbH

[0330] Our reference: DID29174PCT

[0331] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0332] Date: April 5, 2024

[0333] The self-similarity of biosensor 1 is determined here in a first consideration, for example, on the basis of the following calculation rule, taking into account the excitation voltage V0 at the resistance bridge, the single-ended voltage at the passive test transducer Utest passive (200) and the single-ended voltage at the passive reference transducer Ureferenz passive poo) as follows:

[0334] Self-similarity = min(1 - | Utest passive (200) / (V0 / 2) - 11 , 1 - | Ureference passive (300) / (VO / 2) - 11),

[0335] where min( , ) returns the smallest value of the arguments. The single-ended voltages across the electrical resistances of the passive test transducer Utest passive (200) and the passive reference transducer Ureferenz passive (poo) can be determined from the electrical resistances Rreterenz passive (200), Rreterenz active (220), Rtest passive (poo), Rtest active (320) of the transducers 200, 220, 300, 320 in the resistor bridge and the excitation voltage VO across the resistor bridge as follows:

[0336] Utest passive (200) = VO * Rtest passive (200) / ( Rtest passive (200) + Rtest active (220)) ,

[0337] passive ureference (300) = VO * Reterence passive (300) / ( Reterence passive (300) + Reterence active (320)).

[0338] From this, the self-similarity as a function of the electrical resistances Rreterence passive (200), Rreterence active (220), Rtest passive (300), Rtest active (320) of the transducers 200, 220, 300, 320 is obtained as follows:

[0339] Self-similarity = min(1 - I (Rtest passive <200) - Rtest active (220)) / (Rtest passive (200) + Rtest active (220)) | , 1 - | (Rreterence passive (300) - Rreterence active (320)) / (Rreterence passive (300) + Rreterence active (320)) |) ,

[0340] Self-similarity can possibly also be determined by including the bridge cross-stress VB.

[0341] As described above, the first trained artificial intelligence used in step S2 can be a trained neural network 80. Figure 6A shows the trained neural network 80 from step S2 as an example of a possible first artificial intelligence. The first trained artificial intelligence, for example, the neural network 80 from Figure 6A, is trained with specific training data to enable the determination of performance parameters from the quality parameters for the respective biosensor 1. The following section will further discuss the generation and application of the training data used to train the first artificial intelligence for use in step S2.

[0342] Page 32 of 49 Applicant: digid GmbH

[0343] Our reference: DID29174PCT

[0344] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0345] Date: April 5, 2024

[0346] The neural network 80 shown in Figure 6A determines the performance parameters of the biosensor 1 and corresponding threshold values ​​for an analyte 90 in step S2 using a feature vector. The feature vector is derived from at least one determined quality parameter of the biosensor 1.

[0347] The feature vector can be obtained, for example, from data preparation of at least one quality parameter. This data preparation can include, for instance, rescaling the at least one quality parameter. It can also include an error analysis of the at least one quality parameter, culminating in the generation of the feature vector. This data preparation can involve removing unnecessary quality parameters. Furthermore, principal component analysis or another data reduction or selection method can be applied during data preparation to obtain a feature vector that contains as much information as possible about the underlying at least one quality parameter.Data processing can also include storing at least one quality parameter and / or the feature vector in a database or in memory, such as an EEPROM, of biosensor 1. The feature vector can also correspond to at least one quality parameter.

[0348] The exemplary neural network 80 shown in the embodiment in Figure 6A has four layers. A neural network 80 can also have a topology with, for example, more or fewer layers, as well as more or fewer neurons in the respective layers.

[0349] The feature vector based on the quality parameters is passed as input to the trained neural network 80. For example, the feature vector can have n components x = (xi,... ,xn), where the neural network 80 has n neurons 82 in a first layer 84. The neural network 80 in the embodiment shown in Figure 6A has three neurons 82 in its first layer 84. Thus, feature vectors with three components are passed to the neural network 80. A neuron 82 at the i-th position in the second layer 85 of the neural network 80 outputs a value as follows: Oi (w, T + bi), where w = (wi,... ,w n ) is a vector with weights and bi is a bias value. Oi is an activation function. The activation function can be, for example, a ReLU or sigmoid function.

[0350] The neural network 80 outputs performance parameters and their corresponding threshold values ​​as output values ​​at its last layer 86. In the embodiment shown in Figure 6A, the

[0351] Page 33 of 49 Applicant: digid GmbH

[0352] Our reference: DID29174PCT

[0353] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0354] Date: April 5, 2024

[0355] Neural network 80 has five neurons 82 in its last layer 86, where the five neurons 82 output a performance parameter list 70. A performance parameter list can, for example, be a list of tuples, where each tuple from the list of tuples contains a value and a corresponding threshold for each performance parameter.

[0356] The five neurons 82 can, for example, output five values ​​for the performance parameter specificity in the form of a performance parameter list, each with its corresponding threshold.

[0357] For example, the specificity of biosensor 1 is selected from the performance parameter list with the five values ​​for specificity and their associated threshold values.

[0358] The trained neural network 80 in Figure 6A is trained by training data 88 and is suitable to fulfill the function described in S2.

[0359] Figure 6B schematically illustrates how the training data 88 can be obtained.

[0360] A measurement is performed under controlled conditions using a large number of pre-fabricated biosensors 1. In other words, a biosensor 1 is used under normal measurement conditions—that is, the conditions for which the respective biosensor 1 is intended by the user—and under normal environmental influences expected during a measurement, to measure a known analyte concentration of the analyte 90 of interest. Similarly, a large number of measurements can also be performed under controlled disturbances to create a more comprehensive topography. The resulting time-resolved measurement is then stored for each biosensor 1 of the pre-fabricated biosensors, in combination with information about the measured analyte 90—in particular, the known analyte concentration and / or the presence or absence of the analyte 90 in the sample liquid 9 to be analyzed.

[0361] For each individual biosensor 1, a time-resolved measurement curve 800 is obtained from the multitude of already manufactured biosensors, as shown in Figure 6B as an example for one biosensor. The measurement accordingly yields a measurement curve 800, where the measurement curve is based on a voltage signal at the resistance bridge of the biosensor 1 that varies over time.

[0362] It is also known which analyte 90 was measured at which concentration and under which disturbance, so that both the individual measurement curve 800 for each biosensor 1 and the information relating to it are available.

[0363] Page 34 of 49 Applicant: digid GmbH

[0364] Our reference: DID29174PCT

[0365] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0366] Date: April 5, 2024

[0367] The expected outcome is known. In this way, training data can be provided to train a second artificial intelligence, 820.

[0368] In the second artificial intelligence 820, for example a classifier, the individual measurement result 822 is determined from the time-resolved measurement curve 800 of an individual measurement and a validity parameter 824 is output at the same time.

[0369] For example, the second artificial intelligence (AI) 820 can be a classifier that categorizes the measurement result as positive or negative for each individual. AI 820 can incorporate dynamic time warping (DTW) for this purpose. AI 820 can also include a regression model. This regression model can output an analyte concentration as an individual measurement result.

[0370] The validity parameter 824 is compared to a predefined threshold to indicate the accuracy of the measurement result. The threshold, stored during the biosensor's customization along with the application scenario, may be used as a predefined threshold for evaluating the validity parameter during the measurement assessment. The accuracy can be assessed as either correct or invalid.

[0371] The measurement result may be invalid, for example, if a time-resolved measurement curve 800 ends abruptly due to a premature termination of the associated measurement, resulting in an unknown curve trajectory. In this case, the second artificial intelligence 820 can determine the individual measurement result for the curve, but only with very low validity. Based on a comparison of the corresponding validity parameter with the predefined threshold, the measurement result may then be classified as invalid.

[0372] At the same time, before each measurement of a specific analyte with a biosensor 1, its quality parameters are also determined, for example the self-similarity of the electrical resistances in the resistance bridge.

[0373] In a sorting step 840, the determined data, consisting of the individual measurement result 822, the assigned validity parameter 824, the assigned quality parameter, for example the self-similarity 826, as well as the known result for each biosensor from the multitude of already manufactured biosensors, are statistically evaluated.

[0374] Page 35 of 49 Applicant: digid GmbH

[0375] Our reference: DID29174PCT

[0376] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0377] Date: April 5, 2024

[0378] The statistical analysis can include the following sub-steps. Sorting step 840 can begin with filtering the collected data. During this filtering process, each assigned validity parameter is compared to a defined threshold to determine the accuracy of each individual measurement result. All individual measurement results that are, for example, correct are combined with their assigned validity parameters, their associated quality parameters, and the expected results to form a filtered dataset. The filtered dataset is then divided into groups, with each group containing biosensors with the same quality parameters. A group may also include all related data from the filtered dataset where one or more quality parameters exceed one or more predefined thresholds.For each of these groups, performance parameters such as sensitivity, specificity, and / or yield can now be determined. The filtering and subsequent determination of the performance parameters is repeated for a large number of predefined thresholds to generate the training data.

[0379] Figure 6B shows an exemplary three-dimensional representation of the training data 88, for example, for sensitivity, where the sensitivity values ​​are displayed based on a grayscale coding 72. The three-dimensional representation plots self-similarity as a quality parameter on the y-axis and the threshold value on the x-axis.

[0380] The resulting training data 88, which, when applied to the trained first artificial intelligence, such as the neural network 80 in Figure 6A, are used as training data, lead to the determination of a list of performance parameters 70, such as sensitivity, specificity and / or yield, with the corresponding thresholds for the biosensor for the specific anlyte 90, when a quality parameter for a biosensor 1, for example self-similarity, is entered.

[0381] The training data can also be tagged according to specific application scenarios. For example, it can be noted in the training data that performance parameters within a certain range are suitable for a specific application, such as lactate measurement or the detection of a particular virus. Once the training data is tagged according to specific application scenarios, the corresponding training data can be used to train an initial artificial intelligence that can personalize the training.

[0382] Page 36 of 49 Applicant: digid GmbH

[0383] Our reference: DID29174PCT

[0384] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0385] Date: April 5, 2024

[0386] Biosensors for measuring a specific analyte in step S3, as shown for example in Figure 1, including a specific application scenario for the biosensor 1.

[0387] The second artificial intelligence 820 can also be used to evaluate a measurement at a user (see Figure 8).

[0388] The second artificial intelligence, 820, is trained to output the individual measurement result and validity parameter. For training, stored measurement curves from the numerous previously manufactured biosensors can be used in combination with information about the measured analyte – in particular, the known analyte concentration and / or the presence or...

[0389] Absence of the analyte to be analyzed in the sample liquid. This data is stored in a training dataset.

[0390] If the second artificial intelligence, 820, is a classifier, for example, then the classifier can be based on a nearest-neighbor classification method. The classifier determines the measurement result, for example, for an unknown time-resolved measurement curve 800 of a biosensor, based on a comparison of the unknown measurement curve with the measurement curves in the training dataset.

[0391] In the case of measurement curves of equal length and without distortion, the comparison can be performed, for example, by calculating a norm between the still unknown measurement curve of the biosensor and all measurement curves in the training dataset. The absolute values ​​are then ordered in ascending order. The measurement result of the unknown measurement curve of the biosensor can be obtained by a calculation formula, such as averaging, of a predefined number of measurement results from the training dataset with the smallest absolute values. The validity parameter is calculated, for example, based on the highest absolute value of the measurement curve in the training dataset whose measurement result is used to determine the measurement result of the still unknown measurement curve.

[0392] The second artificial intelligence, 820, can be adapted to include dynamic time warping (DTW). DTW is based on a method that compares two time series, such as two time-resolved measurement curves (800), where the two time series do not have to match perfectly but can, for example, be shifted or distorted in time. DTW allows, for example, the comparison of two time series of different lengths. Consequently, DTW can also be used to compare two

[0393] Page 37 of 49 Applicant: digid GmbH

[0394] Our reference: DID29174PCT

[0395] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0396] Date: April 5, 2024

[0397] Time-resolved measurement curves 800, which do not run over the same time period, are carried out.

[0398] If the second artificial intelligence, 820, includes a regression model, for example, then the regression model is trained by minimizing a cost function. The validity parameter can be derived from a calculation that incorporates the measurement result and the training dataset.

[0399] Figure 7 shows the training data 88 again in three representations encompassing sensitivity, specificity and yield.

[0400] The values ​​for the performance parameters sensitivity, specificity, and yield are represented by a grayscale bar 72. The self-similarity in the three representations lies between 0 and 1, and the threshold value between 0.5 and 0.9.

[0401] The three diagrams demonstrate a systematic relationship between sensitivity, specificity, and yield, depending on self-similarity and the threshold. Furthermore, a relationship exists between the performance parameters of sensitivity, specificity, and yield themselves for a given level of self-similarity and a specified threshold. The threshold thus depends on the performance parameters and also on the self-similarity.

[0402] The systematic relationship is evident in the fact that, with a certain degree of self-similarity and increasing threshold, greater sensitivity and specificity tend to result, although the yield tends to decrease.

[0403] By including the specified threshold, the sensitivity, specificity, and yield for biosensor 1 can then be determined during measurement based on the respective performance parameter lists 70. For the specific self-similarity of biosensor 1, a corresponding adjustment between the sensitivity, specificity, and yield of biosensor 1 can thus be made by setting the threshold.

[0404] Figure 8 shows, using a flowchart, an exemplary implementation for an evaluation of a measurement of a biosensor 1 based on the individualization of the biosensor 1.

[0405] Page 38 of 49 Applicant: digid GmbH

[0406] Our reference: DID29174PCT

[0407] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0408] Date: April 5, 2024

[0409] In procedure 100, as described above, the individualized biosensor is used to measure the corresponding specific analyte for the application scenario for which it was individualized.

[0410] In step S4, the measurement is then carried out with the appropriately customized biosensor under the specified measurement conditions. The resulting time-resolved measurement curve 800 shows the corresponding time-resolved course of the signal voltage at the resistor bridge for the reference cantilever 3 and the test cantilever 2.

[0411] The measurement of specific analyte 90 can be performed using an apparatus. This apparatus includes, among other things, the sample 9 containing specific analyte 90 to be measured. An apparatus may also comprise several technical components. These technical components could be, for example, one or more mobile devices. The apparatus may also include one or more programs. These programs may be stored on and / or run on the technical components. The programs can be used to evaluate the measurement of the specific analyte, incorporating the individualized biosensor. The evaluation can be triggered, for example, by an electrical signal from biosensor 1 to the apparatus. Alternatively, the measurement data can be stored on a server.The measurement data, as well as the data from the biosensors themselves, such as their at least one quality parameter, application scenarios, performance parameters, and / or threshold, can also be stored and processed in a cloud.

[0412] The resulting measurement curve 800 is fed in step S5 to the second artificial intelligence 820, which outputs an individual measurement result as well as a validity parameter confirming that the measurement result is correct.

[0413] Depending on the validity parameter, the final measurement result is then output to the user in step S6. If the validity parameter falls below a predefined threshold, the measurement result is classified as invalid. However, if the validity parameter is above the predefined threshold, a measurement result can be output that is either positive or negative, with a corresponding probability that is at least equal to the validity parameter. The threshold stored during the customization of biosensor 1, along with the application scenario, may be used as a predefined threshold for the evaluation of the

[0414] Page 39 of 49 Applicant: digid GmbH

[0415] Our reference: DID29174PCT

[0416] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0417] Date: April 5, 2024

[0418] Validity parameters are used in the evaluation of the measurement. The threshold value resulting from the individualization of biosensor 1 preferably also derives from the application scenario of biosensor 1.

[0419] The comparison between the validity parameter and the predefined threshold can also be performed in step S5 by the second artificial intelligence. Step S6 therefore includes an output of the classification of the measurement result, the measurement result itself, and / or the corresponding probability, which is at least equal to the validity parameter.

[0420] This ensures that the measurement result, depending on the given application scenario, can always be absolutely reliable through the individualization of the biosensor on the one hand and the evaluation of the measurement result based on its validity parameter for the corresponding performance parameters on the other. In an extreme case, this can achieve, for example, a specificity and sensitivity of 100% for a virus test.

[0421] Therefore, if, for example, when customizing the biosensor for a specific application scenario, it has been determined that a specificity of 98% and a validity of 59% are required with a threshold of 0.60, then this threshold must be used to evaluate the measurement for this individual biosensor in order to achieve the required performance parameters.

[0422] In other words, if the measurement performed during the evaluation of the measurement curve in artificial intelligence 820 at step S5 yields a validity parameter below 0.95, the corresponding measurement is discarded as invalid. Therefore, it is always ensured that the parameters defined for the specified application scenario are strictly adhered to during the measurement, or else the measurement is discarded as invalid.

[0423] Irregularities in a measurement can lead to an unknown or irregular shape in the measurement curve. However, by training the second artificial intelligence, 820, a precise measurement result can still be determined even with minor irregularities. A minor irregularity occurs, for example, if the measurement is interrupted shortly before completion, or if minor disturbances, such as slight temperature fluctuations, occur during the measurement. Through repeated and adapted training of the second artificial intelligence, 820, such minor irregularities in the measurement can be compensated for.

[0424] Page 40 of 49 Applicant: digid GmbH

[0425] Our reference: DID29174PCT

[0426] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0427] Date: April 5, 2024

[0428] Irregularities can arise, for example, from a significantly premature termination of the measurement, where biosensor 1 is separated from the specific analyte 90 well before the measurement is completed. With such an unknown or irregular measurement curve, the second artificial intelligence 820 will still arrive at a result, but the validity parameter assigned to that result will be very low. In other words, an irregular measurement will always lead to the validity parameter being rejected as invalid when assessed in relation to the threshold in step S6. Consequently, the number of measurements rejected as invalid can be reduced through repeated and adapted training of the second artificial intelligence 820.

[0429] Since the individualization of the respective biosensor is merely advantageous but not essential for the method shown in Figure 8 for determining the measurement result, the exemplary embodiment can also be considered as starting with method 100 for individualizing a biosensor 1. However, the individualization of the biosensor is not absolutely necessary.

[0430] Figure 9 shows an exemplary embodiment of a biosensor 1 undergoing no or at least one further manufacturing step S7 based on a method 100 for individualizing the biosensor, using a flowchart. As already described in Figure 4, successful individualization leads to the determination of an application scenario for the biosensor 1. The application scenario, as well as the performance parameters assigned to the application scenario with their respective threshold values, can be stored in the biosensor 1. Based on the stored data, comprising the application scenario and the assigned performance parameters with their respective threshold values, biosensors, for example, those held in stock, can undergo at least one further manufacturing step S7. Manufacturing steps can, for example, include functionalization.During functionalization, a biosensor is adapted to its specific needs based on the stored data. Functionalization can include, among other things, the application of a receptor layer 24 and / or a reference layer 34. Alternatively, the manufacturing process of a biosensor can be terminated based on its individualization, so that the biosensor does not undergo any further manufacturing steps. These manufacturing steps can occur during production, repair, and / or verification of the biosensor.

[0431] Page 41 of 49 Applicant: digid GmbH

[0432] Our reference: DID29174PCT

[0433] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0434] Date: April 5, 2024

[0435] This allows biosensors to be flexibly adapted to their specific application scenario, even after extended storage periods. Furthermore, biosensors can potentially undergo a customization process at any time, after which they can be reworked or their production halted.

[0436] Where applicable, all individual features shown in the exemplary embodiments can be combined and / or exchanged without leaving the scope of the invention.

[0437] Page 42 of 49 Applicant: digid GmbH

[0438] Our reference: DID29174PCT

[0439] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0440] Date: April 5, 2024

[0441] List of reference signs

[0442] 100 procedures

[0443] 51 Determine at least one quality parameter

[0444] 52 Determine at least one performance parameter and corresponding threshold value 53 Customize a biosensor

[0445] 54 Performing a measurement

[0446] 55 Classifying the measurement

[0447] 56 Evaluation of a measurement

[0448] 57. Go through no or at least one further manufacturing step

[0449] 1 Biosensor

[0450] 2 Test cantilevers

[0451] 3 Reference cantilevers

[0452] 4 Electronics

[0453] 9 Sample

[0454] 90 Analyt

[0455] 200 passive test transducers

[0456] 220 active test transducers

[0457] 300 passive reference transducers

[0458] 320 active reference transducer

[0459] 400, 401, 402, 403 electrodes

[0460] 20 base

[0461] 22 Deformable part

[0462] 24 receptor layer

[0463] 30 base

[0464] 32 Deformable part

[0465] 34 Reference layer

[0466] 44 Bridge cross-voltage detector

[0467] 440 AD converter

[0468] 50 Multimeter

[0469] 52 measuring probes

[0470] 60-camera system

[0471] 62 computers

[0472] Page 43 of 49 Applicant: digid GmbH

[0473] Our reference: DID29174PCT

[0474] Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence

[0475] Date: April 5, 2024

[0476] 70 Performance parameter list

[0477] 72 grayscale

[0478] 80 Neural Network

[0479] 82 neurons

[0480] 84 First shift

[0481] 85 Second shift

[0482] 86 Last shift

[0483] 88 training data

[0484] 800 Measurement curve

[0485] 820 Second artificial intelligence

[0486] 822 Measurement result

[0487] 824 Validity parameters

[0488] 826 quality parameters

[0489] V0 Excitation voltage

[0490] VB Bridge cross-tension

[0491] Page 44 of 49

Claims

Applicant: digid GmbH Our reference: DID29174PCT Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence Date: April 5, 2024 Claims 1. Method (100) for individualizing a biosensor (1) for measuring a specific analyte, comprising: Determine at least one quality parameter of the biosensor (1), Determine at least one performance parameter and an associated threshold of the biosensor (1) based on the at least one quality parameter using a trained first artificial intelligence (S2), and Individualizing the biosensor for measuring a specific analyte by evaluating at least one of the performance parameters and the associated threshold for an individual application scenario (S3).

2. Method according to claim 1, characterized in that the individualization of the biosensor (1) for measuring a specific analyte takes place using a trained first artificial intelligence and including a specific application scenario.

3. Method according to claim 1 or 2, characterized in that the quality parameter is determined on the basis of a property of the biosensor (1), wherein at least one electrical resistance is determined as the property, preferably all electrical resistances of a resistance bridge of the biosensor.

4. Method according to one of the preceding claims, characterized in that the biosensor (1) comprises at least one resistance bridge and at least one of the quality parameters is a self-similarity of the electrical resistances of the resistance bridge.

5. Method according to one of the preceding claims, characterized in that at least one of the performance parameters of the biosensor comprises a sensitivity and / or a specificity and / or a yield.

6. Method according to one of the preceding claims, characterized in that the trained first artificial intelligence has at least one of the performance parameters and a Page 45 of 49 Applicant: digid GmbH Our reference: DID29174PCT Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence Date: April 5, 2024 Threshold determined on the basis of a feature vector, whereby the feature vector is obtained from a data preparation of at least one quality parameter.

7. Method according to one of the preceding claims, characterized in that the trained first artificial intelligence is trained on the basis of training data obtained by the following steps: Determining a time-resolved measurement curve (800) of at least one specific analyte, as well as at least one sample without the specific analyte, each with a plurality of biosensors; Determining an individual measurement result (822) and a validity parameter (824) for each biosensor using a second artificial intelligence (820) from the time-resolved measurement curve (800) of the respective biosensor; Determine at least one quality parameter (826) for each biosensor; Statistical evaluation of the individual measurement results (822) based on the assigned validity parameters (824) using threshold values, the assigned quality parameters (826) and the previously known result for each biosensor; Determining performance parameters from statistical analysis; Using quality parameters, performance parameters, and thresholds as training data.

8. Method according to claim 7, characterized in that the time-resolved measurement curve (800) is recorded at at least one defined concentration of the analyte and / or that the time-resolved measurement curve (800) is additionally recorded at a defined disturbance and / or that the time-resolved measurement curve (800) is recorded at at least one defined measurement routine.

9. Method according to one of the preceding claims, characterized in that the trained first artificial intelligence is trained on the basis of training data which includes as quality parameters the self-similarity of the electrical resistances of a resistance bridge of each biosensor. Page 46 of 49 Applicant: digid GmbH Our reference: DID29174PCT Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence Date: April 5, 2024 10. Method according to one of the preceding claims, characterized in that the trained first artificial intelligence is trained on the basis of training data which includes as performance parameters the sensitivity and / or the specificity and / or the yield of each biosensor.

11. Method according to one of the preceding claims, characterized in that the biosensor, based on its individualization, undergoes no or at least one further manufacturing step (S7), wherein the no or at least one further manufacturing step preferably comprises storing quality parameters and / or performance parameters and / or threshold values ​​on the biosensor or a database and / or the at least one further manufacturing step comprises functionalizing the biosensor, wherein the functionalizing of the biosensor preferably comprises applying a receptor layer and / or reference layer.

12. Method for generating training data for training a trained first artificial intelligence for use in the individualization of biosensors, comprising the following steps: Determining a time-resolved measurement curve (800) of at least one specific analyte, as well as at least one sample without the specific analyte, each with a plurality of biosensors; Determining an individual measurement result (822) and a validity parameter (824) for each biosensor using a second artificial intelligence (820) from the time-resolved measurement curve (800) of the respective biosensor; Determine at least one quality parameter (826) for each biosensor; Statistical evaluation of the individual measurement results (822) based on the assigned validity parameters (824) using threshold values, the assigned quality parameters (826) and the previously known result for each biosensor; Determining performance parameters from statistical analysis; Using quality parameters, performance parameters, and thresholds as training data. Page 47 of 49 Applicant: digid GmbH Our reference: DID29174PCT Title: Method for the individualization of a biosensor for the measurement of an analyte with high performance parameters using artificial intelligence Date: April 5, 2024 13. Method according to claim 12, characterized in that the time-resolved measurement curve (800) is recorded at at least one defined concentration of the analyte and / or that the time-resolved measurement curve (800) is additionally recorded at a defined disturbance and / or that the time-resolved measurement curve (800) is recorded at at least one defined measurement routine.

14. Method according to claim 12 or 13, characterized in that the quality parameter determined for each biosensor comprises the self-similarity of the resistances in a resistance bridge of each biosensor.

15. Biosensor for measuring a specific analyte, comprising at least one resistance bridge, wherein the biosensor is individualized for a specific application scenario for measuring the specific analyte based on a quality parameter determined from the resistance bridge.

16. Biosensor according to claim 15, characterized in that one of the quality parameters is the self-similarity of the resistances of the resistance bridge. Page 48 of 49