Method for performing diagnosis, prognosis and / or therapeutic monitoring of diseases impacting the mineral profile of at least one biological fluid, and a device, system and software package used to implement same

Abstract

The aim of the invention is to provide an efficient, reliable and accurate method for performing diagnosis, prognosis and / or therapeutic monitoring of diseases impacting the mineral composition of at least one biological fluid. To this end, the method according to the invention essentially consists in - implementing at least one reference database (BDR) containing data relating, on the one hand, to samples of biological fluid, and, on the other hand, to the healthy or diseased subjects from which the samples of biological fluid were taken, these data comprising for each sample: (i) at least one mineral profile PM of concentrations of mineral elements in the biological fluid and (ii) conventional clinical data on the subjects; -implementing at least one sample Ex of biological fluid of at least one subject Sx to be tested; - analyzing each sample to determine at least one PM of Sx; - defining at least two categories with a view to making at least one prediction of membership of the subject Sx to be tested to at least one category of diagnosis (d), prognosis (p) and / or therapeutic monitoring (t), and, optionally to at least one subdivision of at least one category; - statistically processing the data of the BDR and displaying the processed data graphically, then positioning the data relating to Sx on this graphical display; - and / or carrying out AI-based machine learning using the data of the BDR; in order to make the prediction of membership regarding Sx and assist, by so doing, with the diagnosis, prognosis and / or therapeutic monitoring of diseases impacting the mineral composition of at least one biological fluid, for the subject Sx to be tested. The invention also relates to an electronic device, to a system for implementing the method, and to at least one computer program (and its medium) for executing the method.

Description

DESCRIPTION Title: Method for diagnosis, prognosis and / or therapeutic monitoring of diseases impacting the mineral profile of at least one biological fluid, as well as a device, a system and software useful for its implementation. Field of the invention

[0001] diagnosis, prognosis and / or therapeutic monitoring of diseases impacting the particular mineral profile of cerebrospinal fluid. These diseases include: . in the case of cerebrospinal fluid, diseases affecting the central nervous system, in particular neurological diseases, such as psychiatric diseases, neurodegenerative diseases such as Parkinson's or Alzheimer's. in the case of blood plasma, cancers, neurological diseases, diseases . , diseases linked to surgical complications following complex operations, in particular cardio-. diseases leading to fertility problems and requiring in particular aids for

[0002] a process for diagnosis, prognosis and / or therapeutic monitoring of these diseases.

[0003] of this process. Technological background of the invention

[0004] Neurological diseases [such as neurodegenerative diseases], cancers, fertility problems, heart operations, have a very significant societal impact and represent a growing burden for the community. Identifying these diseases precisely and early enough, establishing a prognosis or therapeutic monitoring forecasts, are crucial challenges for patients and their caregivers and to help the latter propose the most appropriate therapeutic treatments.

[0005] Due to the progressive aging of the population and the lack of curative treatments, the number of people suffering from neurodegenerative diseases has has increased considerably over the past few decades and is expected to grow steadily in the coming years.

[0006] Furthermore, the use of medically assisted procreation will increase steadily in the coming years.

[0007] Cancer and heart disease have also increased significantly in recent years, and the number of patients treated has more than doubled in less than 20 years.

[0008] It is therefore urgent to have better tools for the diagnosis, prognosis and monitoring of the effectiveness of treatments for these diseases.

[0009] Early, effective and accurate diagnosis is a priority objective.

[0010] But a diagnosis is often very complex, given the wide variety of possible pathologies. Moreover, information is difficult to obtain, especially when the main problem is the central nervous system, and more specifically the brain.

[0011] The structural and functional alterations of the brain that cause these diseases modify the biochemical composition of the cerebrospinal fluid. The latter may therefore contain biomarkers of these diseases.

[0012] Many diagnoses are already based in part on biochemical analyses and the detection and quantification of certain organic molecules present in the cerebrospinal fluid.

[0013] Known biomarkers of cerebrospinal fluid are organic molecules and especially proteins. For example, Alzheimer's biomarkers are proteins: amyloid 1-42, total tau, phosphorylated tau isoforms, NF-L, S90B, VILIP-1, YKL-40, PAPP-A, albumin.

[0014] EP1907838B1 describes a method for qualifying the state of Alzheimer's disease in a subject, comprising:(a) measuring the concentration of a VGF-1 peptide consisting of the amino acid sequence of SEQ ID NO: 1 in the cerebrospinal fluid (CSF) of the subject; wherein the VGF-1 peptide is measured by capture on an adsorbent surface of a SELDI probe and detecting the captured peptide by laser desorption / ionization mass spectrometry; and (b) correlating the concentration of the VGF-1 peptide with the state of Alzheimer's disease, or an increased CSF concentration of the VGF-1 peptide is an indicator of Alzheimer's disease. EP1907838B1 also discloses a method for determining the progression of Alzheimer's disease comprising: (a) measuring the LCS concentration of a VGF-1 peptide consisting of the amino acid sequence of SEQ ID N degrees 1 in a sample in a first step; (b) measuring the concentration of LCS of VGF peptide-1 in a sample at a second time; and (c) comparing the first measurement and the second measurement; wherein an increased concentration of LCS of VGF peptide-1 in the second measurement indicates progression of Alzheimer's disease. EP19078388B1 finally relates to software installed on a computer comprising: (a) code which allows access to data assigned to a sample, the data comprising the measurement of a VGF peptide-1 consisting of the amino acid sequence of SEQ ID No. 1; and (b) code which executes a classification algorithm which assigns the Alzheimer's disease status of the sample, based on the extent to which an increase in VGF peptide-1 is an indicator of Alzheimer's disease.

[0015] delicate in many cases. Molecules are often difficult to detect and quantify. They are not very stable, require precautions for sampling and storage. They can , which makes their detection even more complex. Research has been carried out to try to link the presence of certain studies have measured heavy elements and essential metals in the plasma and / or cerebrospinal fluid of patients with dementia. high variability. As a result, metals are still not considered reliable biomarkers. In addition, mineral sampling processes are often complex. Mineral traces are also very indirectly associated with this type of disease, as causes or consequences of these diseases. For all cerebrospinal fluid, in practice, to diagnose a neurological disease.Objectives of the invention

[0016] - provide an efficient method for diagnosis, prognosis and / or therapeutic monitoring of diseases impacting the particular mineral composition of cerebrospinal fluid, - provide a reliable method for diagnosis, prognosis and / or therapeutic monitoring of licular diseases, - provide a precise method for diagnosis, prognosis and / or therapeutic monitoring of , - provide a method for diagnosis, prognosis and / or therapy or therapeutic monitoring, - provide a method for diagnosis, prognosis and / or therapeutic monitoring of diseases, a valuable and effective aid for the doctor, allowing the latter to treat and usefully contribute to the fastest and most complete cure of patients possible. - provide an electronic device and a system including this device, which are efficient, reliable, precise and economical for implementing the above objectives. Description of the invention

[0017] DEFINITIONS

[0018] Throughout this presentation, any singular designates indifferently a singular or a plural.

[0019] The definitions given below as examples may be used to interpret this presentation: - "CNS": Central Nervous System. - "CSF": cerebrospinal fluid. - "CSF": cerebrospinal fluid. - "Differentiating data": discriminating data that provide specific information on a subject S xto be tested which may be a patient. - "AI": Artificial Intelligence. - "ROC": "Receiver Operating Characteristic", also called performance characteristic (of a test) or sensitivity / specificity curve, which is a measure of the performance of a binary classifier, that is to say of a system which aims to categorize elements into two distinct groups on the basis of one or more of the characteristics of each of these elements. - "AUC": "Area Under the Curve", area under the ROC curve. - "PCA": Principal Component Analysis, a statistical technique used to reduce the dimensionality of data while preserving important information. - "t-SNE": "t-distributed Stochastic Neighbor Embedding", a dimensionality reduction method. - “RFE”: “Recursive Feature Elimination”, a feature selection method used in the field of machine learning. - “Cross-validation”: A method of evaluating the performance of a model by testing it on several subsets of the data to obtain a more reliable estimate of its performance. - “FPR”: “False Positive Rate”. A measure used in the evaluation of classification models, representing the rate of values ​​incorrectly predicted as positive compared to the set of actual negative values. - “TPR”: “True Positive Rate”. - “Sensitivity”: Measures the ability to detect true positives, i.e., the proportion of positive cases that are correctly identified by the test. - “Specificity”: Measures the ability to exclude false positives, i.e., the proportion of negative cases that are correctly identified as such by the test. - “approximately”: to within + or - 10%, preferably within 5%.

[0020] PROCESS

[0021] The invention satisfies at least one of the above objectives and relates, according to a first aspect, to a method according to claim 1.

[0022] (i) mineral profile of the biological fluid of the subjects to be tested, based on judicious selection and / or on and, preferably, (ii) conventional clinical data, to feed automatic learning, of the statistical processing type combined with visualization (1 er protocol) and / or type IA (2 ème protocol), to provide medical decision support in terms of . The 1 er and the 2 ème : * ; * ( d,p,t) and / or healthy subjects S m=0 (d,p,t) , with a given reliability (sensitivity / specificity).

[0023] to determine a diagnosis (d), a prognosis (p) and / or therapeutic monitoring (t), one or more samples Ex of biological fluid are taken. An analysis S4 of the mineral profile of this sample of these samples is carried out. The data D Sx obtained are data (i) of concentrations of mineral elements in the biological fluid. data D Sx , which are conventional clinical data (ii), are assigned to the subject Sx. Questions are asked by the physician concerning the subject Sx to be tested. The responses of the subject Sx to be tested, to at least one target, each target comprising at least one diagnostic category (d), prognostic category (p) a category an electronic device described below and shown in the attached figure 24, a reference database BDR comprising data (i) PM and (ii) conventional clinical data on an entire population of sick subjects (d,p,t) and / or healthy subjects S m=0(d,p,t) electronic. A user defines based on the questions asked, at least 2 targets for x to be tested to at least one target. A 1 er statistical protocol and / or a 2 e AI protocol is launched by the electronic device, on expected, with sensitivity and / or therapeutic.

[0024] * the biological fluid is chosen from the group comprising - advantageously consisting of - bile. * and the diseases impacting the mineral profile of the biological fluid, include: . . in the case of cerebrospinal fluid, diseases affecting the central nervous system, in particular neurological diseases, such as psychiatric diseases, neurodegenerative diseases of the Parkinson's or Alzheimer's type, . in the case of blood plasma, cancers, neurological diseases, diseases . in complex operations, in particular cardiovascular, . diseases causing fertility problems and requiring in particular procreation aids.

[0025] Advantageously, the BDR includes data relating to S subjects m=0 (d,p,t) healthy reference subjects and / or subjects (d,p,t)several diseases impacting at least two biological fluids chosen from the group comprising - advantageously consisting of - cerebrospinal fluid, whole blood, lysine, bile.

[0026] one step -S0 ^ - construction of the reference database (RDB) consisting of collect data, including data (i) and / or (ii) and / or data D Sx and store them.

[0027] According to a database includes data relating to at least one control group of healthy subjects S m=0 (d,p,t) reference, of which the total number N tt is greater than or equal to -in ascending order of preference- 100; 500, 1000 and data relating to at least one target group of reference diseased subjects S (d,p,t), whose total number Ntc is greater than or equal to -in ascending order of preference- 50; 500, 1000.

[0028] Advantageously, the mineral elements of the PM mineral profile in the biological fluid of subjects S m=0 (d,p,t) (d,p,t) reference elements comprise at least 3 atomic elements selected from halogens, preferably Cl, Br, I.

[0029] the mineral elements of the PM mineral profile in the biological fluid of S subjects m=0 (d,p,t) (d,p,t) of reference are chosen from a group comprising -advantageously consisting of: at least three alkalis, preferably Li, Na, K, Rb, Cs; at least three alkaline earths, preferably Mg, Ca, Sr, Ba; at least three transition elements, preferably Fe, Mn, Cu, Zn, Cr, Co, Ni and at least 3 halogens, preferably Cl, Br, I and at least two trace elements, advantageously chosen from Al, Ti, Zr, Pb, La, Ce, Gd, U.

[0030] By "trace element" is meant, for example, within the meaning of this presentation, an element present in the biological fluid, in a mass concentration less than or equal to 10 µg / L, and preferably in more than 95, or even 99% of cases.

[0031] conventional clinical data relating to S subjects m=0 (d,p,t) and / or Sm 1(d,p,t) of reference, these data (ii) include and gender.

[0032] Preferably, S4 is semi-quantitative, with inductively coupled plasma mass spectrometry, or ICP-MS (Inductively Coupled Plasma Mass Spectrometry) being preferred.

[0033] Quickly, semi-

[0034] -quantitative according to fluid, preferably biological fluid, consists essentially of: S4.1-: - preferably a biological fluid;- comprising at least one negative control sample - that is, containing one or more positive control elements defined below, in a mass concentration, eg less than or equal to 1 ppb; - comprising at least one positive control sample comprising one or more positive control mineral elements, at least a portion of which have atomic masses and at least a portion of which have atomic masses greater than or equal to; ^ the number N emt ascending order of preference: N emt ..^ 10; N emt ..^ 20 ; N emt ..^ 25 ; S4.2- Measurement, using the device, of the concentrations of mineral elements in the negative standard; S4.3- Measurement, using the device, of the concentrations of positive control mineral elements in the positive standard, S4.4- Calibration, preferably 2-point calibration; production of straight lines, from the concentrations fixed for the N emtpositive control mineral elements; S4.5- Possibly calculation of N concentrations emt positive control mineral elements, from the curves, preferably from; S4.6- Calculation of the concentration of analyte mineral elements in the population of calibrations of the concentrations of N emt - .

[0035] This semi-quantitative analytical approach aims to measure indicators in biological fluids, expressed in concentrations, which take into account both the generated by the samples. Matrix effects are linked to the nature of the matrices constituting the samples, eg, viscosity, charges in

[0036] Thanks to good measurement fidelity / repeatability, the traceability of the results, and thus their comparability, are guaranteed in relation to an established reference, which can be, for example, a set of biological fluids (cerebrospinal fluid, plasma, etc.)

[0037] The accuracy of measurements is also improved.

[0038] (measurement of concentration in a biological fluid) 10 to 50, preferably 20 to 40, and, for example, 28 mineral elements, chemical incompatibilities that would require multiplying the standards in a quantitative approach. For example, standards including metal fluorides (Ti, Zr, Hf, Nb, Ta, Mo, Si, Sn, Ge) are incompatible with the presence of earths.

[0039] semi-quantitative can, for example, also be defined as follows: the response factors expressed in equivalent concentrations, in mineral elements are determined on a measurement scale ranging eg from 0.01 µg / L to 10 g / L, using a semi-quantitative method with a standard ranging eg from 20 to 40 elements whose concentration in a biological fluid is eg10 to 30 µg / L, this standard comprising, for example, 28 elements at 20 µg / L and a negative control (a blank) eg consisting of 1% nitric acid. With this semi-quantitative approach, it is easy to determine the response factors expressed in equivalent concentrations in a biological fluid, of several dozen eg different mineral elements. The accuracy and robustness of the method have been validated by multiple tests.

[0040] For the 28 elements present in the standard, the concentration is calculated at the concentration is interpolated from that of the elements present in the calibration standard, by applying response factors, which depend on their mass and their isotopic abundances,

[0041] This provides a rich sampling for the selection of classes or subclasses of mineral elements relevant for diagnosis, prognosis and / or therapeutic monitoring. It should be noted that halogens, preferably Cl, Br, I, soluble anions are very interesting for monitoring the different biological balances and anion pumps.

[0042] Semi-quantitative analysis is particularly suitable for the creation of data (i) of mineral profiles, intended to be stored in the BDR and exploitable by statistical processing and / or by AI (chemometrics).

[0043] Indeed, the concentration of mineral elements in patients' biological fluids is often very variable. And if certain main minerals and trace elements evolve in a well-defined domain and have a homeostasis which limits their deviation from vary over several orders of magnitude, often with concentration differences of more than 1000, on certain elements, between healthy subjects S m=0reference and / or reference sick subjects treated or not (patients) - quantitative allows to measure sufficiently precisely and above all in a sufficiently reproducible way, for patients. By using a simplified sample preparation (for example simple critical for trace elements) and an ICP analyzer - a simple general multi-elemental calibration (without necessary use it is possible to obtain a multimineral profile differentiating minerals, X corresponding to - in ascending order of preference - 3, 5, 10, 15, 15, 20, 25, 30, for example, usable for a database. With a variability in the value obtained which can be less than 5% for the main elements and less than 20% for the trace elements.

[0044] It is nevertheless possible to carry out a more precise quantification of certain elements of interest. To do this, a calibration range specific to the concentrations of the elements is carried out. The semi-quantitative and quantitative results are then obtained in parallel and in a single acquisition for a sample and can be compared.

[0045] a subdivision, according to -S8- x to be tested, and the BDR includes data relating to a number N s of healthy subjects S m=0 (d,p,t) of reference and a number N m of sick subjects (d,p,t) of reference; with [N s / (N s + N m )] *100 greater than or equal to -in ascending order of preference- 20%, 30%, 40%.

[0046] Advantageously, the 1 erprotocol -S9.1- includes the following steps: -S9.1.1- Collection of data in the BDR; -S9.1.2- Normalization of data; -S9.1.3- Selection from data (i), and possibly data (ii), of differentiating data; -S9.1.4- Reduction of the dimensionality of the data in 2D or 3D, preferably by an -SNE; -S9.1.5-Visualization of targets on graphs; -S9.1.6-Positioning on the graphical visualization of the targets, the data D Sx on the Sx subject(s) to be tested, to visualize the positioning of the Sx(s), in relation to the targets.

[0047] The targets are established from the BDR data collected, standardized, selected and statistically processed in -S.9.1.4-.

[0048] Depending on the variants, -S9.1.3- consists of using: * [Variant V1 of 1 er protocol] of the “volcano plot” curves and / or * [Variant V2 of 1 erprotocol] a selection algorithm, preferably a recursive features elimination-RFE selection algorithm, possibly associated with the following steps: -S9.1.3.1- Definition of models; -S9.1.3.2- Data separation; -S9.1.3.3- Learning and evaluation loop; -S9.1.3.4- (AUC score); -S9.1.3.5- choice of the model with the best AUC score.

[0049] Advantageously, the 2 ème protocol in S9.2 includes the following steps: -S9.2.1- Collection of data values ​​in the BDR; -S9.2.2- Normalization of data values; -S9.2.3- by AI; -S9.2.4- Data separation; -S9.2.5- Learning and evaluation loop; -S9.2.6- (AUC score); -S9.2.7- on the best AUC score; preferably using “volcano plot” curves and / or -S9.2.7-. -S9.2.8-; -S9.2.9- selected in -S9.2.7-.

[0050] -S9.2.8- consists of using “volcano plot” curves -S9.2.7-.

[0051] According to remarkable modalities *a prognosis is established consisting of evaluating the risks of clinical complications linked to a renal problem, a pulmonary problem, an infection, and / or the risks of lethal complications, for a subject Sx to be tested, having undergone a surgical operation, in , for example of more than 1 hour and / or blood transfusions and / or extracorporeal blood circulation; this prognosis constituting an aid to the medical decision to maintain or not the subject Sx in intensive care; and / or * in one or more times, to the subject Sx to be tested, for one or more durations after; this therapeutic monitoring constituting an aid to the medical decision to modify or not the dose of medication to be administered to the subject Sx to be tested; and / or * to be tested, the clinical response of the tumors and / or the occurrence of therapeutic constituting an aid to the medical decision to modify or not the doses tion of the subject Sx to be tested; and / or * t / or of .

[0052] DEVICE :

[0053] According to a second of its aspects, the invention relates to an electronic device: * -S9.1-: -M9.1.1- Collection of data in the BDR -M9.1.2- Normalization of the data; -M9.1.3- Selection from the data (i), and possibly the data (ii), of differentiating data; -M9.1.4- Reduction of the dimensionality of the data in 2D or 3D, preferably by an -SNE (t- distributed Stochastic Neighbor Embedding); -M9.1.5- Visualization of the targets on graphs; -M9.1.6- Positioning on the graphic visualization of the targets the data D Sx on the or the Sx subjects to be tested, to visualize the positioning of the Sx(s), in relation to the targets ; * èmeprotocol -S9.2-: -M9.2.1- Collection of data values ​​in the BDR -M9.2.2- Normalization of data values ​​-M9.2.3- by AI -M9.2.4- Data separation -M9.2.5- Learning and evaluation loop -M9.2.6- Plotting the ROC curve -M9.2.7- -M9.2.8- -M9.2.9- -S9.2.7-

[0054] Figure 24 simplified architecture of an electronic device This electronic device comprises an electronic assembly comprising a memory 41, a processing unit 42 equipped for example with a microprocessor, and controlled by at least one computer program 43.

[0055] In one embodiment, the is driven by programs installed in part or in whole on the electronic device.

[0056] In another embodiment, the is driven by a processing unit and installed in part or in whole on the electronic device.

[0057] Furthermore, the device also comprises communication means (CIE) presented for example in the form of network components (Wi-Fi, 3G / 4G / 5G, wired, RFID / NFC, Bluetooth, BLE, LPWan, VLC, etc.) which allow the device to receive data (I) from entities connected to one or more communication networks and to transmit processed data (T) to such entities. These entities may be terminals allowing one or more users to communicate with the device.

[0058] in compiled or interpreted languages. The different components of the device are connected to each other to communicate via protocols such as Bluetooth, Ethernet or WiFi No. The device also includes memory components that will record the data and programs necessary for the device to operate.

[0059] SYSTEM

[0060] According to a third of its aspects, the invention relates to a system (rectangle to the right of the diagram in Figure 1) comprising the electronic device, at least one computer server hosting this electronic device, the BDR and at least one terminal allowing at least one user to communicate with the electronic device, in particular to transmit data D Sx relating to a subject Sx to be tested and of the same type as the data (i) & (ii) defined in S0 in the and to receive in return information, relating to the subject Sx to be tested, diagnosis, prognosis and / or therapeutic monitoring of diseases impacting the composition in

[0061] COMPUTER PROGRAM

[0062] According to a fourth of its aspects, the invention relates to at least one computer program for computer processing.

[0063] DATA SUPPORT

[0064] According to a fifth of its aspects, the invention relates to a data medium on which at least one computer program is recorded according to

[0065] The data carriers may be any entity or device capable of storing the programs. For example, the carriers may comprise a storage means, such as a ROM, for example a CD-ROM or a microelectronic circuit ROM, or one or more often a Flash memory. In addition, the carriers may be transmissible carriers such as an electrical or optical signal, which may be conveyed via an electrical or optical cable, by radio or by other means. The programs according to the invention may in particular be downloaded from a network such as the Internet. Alternatively, the information carrier may be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the method in question. Description of the figures

[0066] The attached figures illustrate non-limiting examples of embodiment, and in which: Fig.1

[0067] [Fig. 1] Figure 1 is a general diagram of the presentation of the process according to Fig.2

[0068] [Fig.2] Figure 2 is a curve showing the distribution of concentrations (µg / L). Fig.3

[0069] [Fig. 3] Figures 3A, 3B, 3C, 3D, 3E, 3F attached represent the volcano plot healthy subjects S m=0 or reference patients. Fig.4

[0070] [Fig.4] Figures 4A,4B attached represent the ROC curves of the AI ​​models used in . Fig.5

[0071] [Fig. 5] les figures 5A,5B,5C annexées illustrent , sous forme de two-dimensional graphs, the distributions between the different targets: C1 including the S category m=0(d5) of healthy subjects free from neurological diseases, compared to the target C2 comprising the category / subcategories (d1;1.1;1.2;1.3) of subjects suffering from neurodegenerative diseases (Figure 5A), C2 compared to the target C3 comprising the category / subcategories (d2;2.1;2.2;2.3) of subjects suffering from inflammatory diseases of the CNS (Figure 5B), and finally a combined comparison C1 / C2 / C3 (Figure 5C). Fig.6

[0072] [Fig.6] Figures 6A,6B,6C attached represent Figures 5A,5B,5C with an Sx to be tested, according to its mineral profile [data D Sx (i)] and conventional clinical data D Sx (i) . Fig.7

[0073] [Fig.7] Figures 7A (C1 / C2), 7B (C2 / C2) attached represent the matrices of . Fig.8

[0074] [Fig.8] Figure 8 is a curve giving the distribution of concentrations (µg / L) in plasma samples Fig.9

[0075] [Fig.9] The attached figure 9 represents the ROC curves of the AI ​​models used. Fig.10

[0076] [Fig.10] The attached Figure 10 represents the confusion matrix for the model. Fig.11

[0077] [Fig. 11] Figure 11 is a set of graphs giving the distribution of concentrations (µg / L) of 37 Fig.12

[0078] [Fig.12] The attached figure 12 represents the volcano plot made in the 1 er case of , showing the impact of mineral concentrations in the urine of sick subjects (p) on prognosis. Fig.13

[0079] [Fig.13] The attached figure 13 represents the volcano plot made in the 2 ème cases of: respiratory failure during the stay (lowest P / F in the 5 days <200), showing the impact of the mineral concentrations in the urine of the sick subjects (p) of reference in prognosis. Fig.14

[0080] [Fig.14] The attached figure 14 represents the volcano plot made in the 3 ème case of: , showing the impact of mineral concentrations in the urine of sick subjects (p) reference in prognosis. Fig.15

[0081] [Fig.15] The attached figure 15 represents the volcano plot made in the 4 ème case of: case study, showing the impact of mineral concentrations in the urine of sick subjects (p) of reference in prognosis.

[0082] [Fig. 16] The attached figure 16 represents in the form of two-dimensional graphs, the distributions between the targets C5** / C6** with a cross [Urine2023_CRB002 of the subject Sx to be tested, according to his mineral profile [data D Sx (i)] and data D sx (ii) comprising . Fig.17

[0083] [Fig.17] The attached figure 17 represents the ROC curves of the AI ​​models used. Fig.18

[0084] [Fig.18] The attached figure 18 represents the confusion matrix for the model. Fig.19

[0085] [Fig.19] The attached figure 19 is a set of graphs giving the sample distribution of . Fig.20

[0086] [Fig.20] The attached figure 20 represents the volcano plot made in the 1 er C1*** case: Category (d1) subcategory (d1.1) endometriosis vs C2***: Category (d1) subcategory (d1.4) reference, showing the impact of mineral concentrations in the follicular fluids of the diseased subjects (p) of reference in prognosis. Fig.21

[0087] [Fig.21] Figure 21 two-dimensional, cross [follicular fluid CHUL4] of the subject Sx to be tested, according to his mineral profile [data D Sx (i)] and data D sx (ii) comprising . Fig.22

[0088] [Fig.22] The attached figure 22 represents the ROC curves of the AI ​​models used 5. Fig.23

[0089] [Fig.23] The attached figure 23 represents the confusion matrix for the model. Fig.24

[0090] [Fig. 24] The attached figure 24 is a schematic representation of the device. EXAMPLES

[0091] Exemple 1 : Diagnostic d sujet Sx statistique (1 er protocol -S9- ) of data (i) & (ii) CSF samples, m=0 or data sick (i) & (ii) S x to test

[0092] -S0-

[0093] -S0.1- conventional clinical data (ii)

[0094] A database is constructed from 1956 CSF samples from over 2000 reference subjects who consulted the Hamburg hospital in Germany, for suspected neurological disease(s).

[0095] Of these more than 2000 reference subjects, a number N s = 515 are healthy subjects S m=0 (d,p,t) of reference and a number N m= 781 are reference sick subjects (d,p,t), suffering from at least one neurological disease with a specific diagnosis and cannot be considered as reference sick subjects (d,p,t) or healthy subjects S m=0 (d,p,t) of reference.

[0096] The total number N t of S subjects m=0 (d,p,t) and (d,p,t) of reference is therefore 1296. [N s / (N s + N m )] *100 = [515 / 515+781)] *100 = 39.7%.

[0097] Each of the healthy reference subjects S m=0 or patients suffering from at least one neurological disease or several targets or categories, and possibly subcategories or subtargets of diagnosis S m(d) specific(s), by , via a review of conventional clinical data concerning the subjects. These categories / subcategories are "targets" / "sub-targets" or "targets" / "sub-targets", within the framework of the decision process involving statistical representations or artificial intelligence, specific to , in particular neurological, impacting the mineral composition of the cerebrospinal fluid.

[0098] For the 1296 healthy subjects S m=0 (d,p,t) or sick (d,p,t) reference of this example, the diagnostic categories S m (d) specific are five in number, namely: (d1) sick subjects (d,p,t) affected by at least one neurodegenerative disease, (d2) sick subjects (d,p,t) inflammatory disease of the CNS (d3) subjects sick (d,p,t) psychiatric disorder, (d4) sick subjects ( d,p,t) other neurological diseases and (d5) healthy subjects S m=0 (d,p,t)not suffering from specific neurological disease: premium control group.

[0099] Some subcategories are also defined. Table 1 below shows all diagnostic categories and subcategories assigned to the 1296 healthy subjects S m=0 (d,p,t) or patients (d,p,t) of reference of this example, medical of the hospital of Hamburg, as well as the number of corresponding subjects. Among the reference subjects, 367 S m (d1) have neurodegenerative disease, 291 S m (d2) have inflammatory CNS disease, 123 S m (d3) suffer from psychiatric disorders, 317 S m (d4) are diagnosed with another neurological disease generating abnormalities in the CSF. Among the study population, 515 S m=0(d5) Healthy subjects are classified as a premium control group, with no symptoms of neurological disorder and no abnormalities detected in the CSF. In addition, the study population includes 660 subjects with non-specific neurological signs, for whom no precise diagnosis could be established and where no known CSF alterations were observed. This latter group is designated as the GNI (Unidentified Group), and the corresponding subjects are designated as S m (d4.2).

[0100] [Table 1] Distribution of CSF samples from the BDR

[0101] -S0.2- Mineral Profile (MP)

[0102] The database is completed for the 1956 healthy subjects S m=0 (d5) or sick (d1.1; 1.2; 1.3; 1.4; 2.1; 2.2; 2.3; 3; 4.1) of reference or of the Unidentified Group (GNI) S m (d4.2) of this example, by the PM mineral profile of N concentrations em= 31 mineral elements established for the corresponding CSF samples.

[0103] These CSF samples are stored in a freezer, generally at a temperature below -15°C, and thawed just before preparation for analysis.

[0104] Before being analyzed by inductively coupled plasma mass spectrometry, or ICP-MS, CSF samples are diluted 25 times in high-purity 1% nitric acid with indium (In) as an internal standard, according to the following protocol: 100 µL of CSF in 2400 µL of HNO31% with 1.6 ppb of In.

[0105] CSF. It helps dissolve and partially degrade proteins, lipids, and other components of CSF to prepare a homogeneous sample. It stabilizes the sample by preventing precipitation or adsorption on the container walls, thus ensuring a uniform distribution of potential interferences during ICP-MS analysis by keeping the elements in a stable ionic form.

[0106] A dilution factor of twenty-five analyzes major and trace elements in a single analysis.

[0107] Thanks to the high sensitivity of mass spectrometry, the volume of CSF required for analysis is extremely low (less than 100 µL), compared to other analytical methods requiring: Between 0.5 and 1 mL for biochemical analyses (biomarkers), Between 1 and 2 mL for cytological analyses (cells), Between 0.5 and 1 mL for genomic analyses (DNA or RNA).

[0108] ICP-MS analysis

[0109] ICP-MS analyses are performed using an Agilent 7850 Series ICP-MS mass spectrometer equipped with a quadrupole analyzer and an integrated autosampler.

[0110] The analyses are carried out, in semi-quantitative mode, with the following conditions: Standard 28 elements at 20 ppb_Blank: HNO31%_ :50 s_Stabilization time: 30 s : 216 s_Post-acquisition rinsing: 3x 45 s_ : 431 s.

[0111] Semi-quantitative analysis involves estimating the relative concentrations of elements without resorting to the rigorous calibration essential for quantitative analysis. It is particularly useful for quickly obtaining the elemental composition of samples.

[0112] The following example describes this semi-quantitative analysis of serum.

[0113] This example is divided into a first part relating to the method and a second part relating to a validation of this method

[0114] Method: The analyses are carried out on an Agilent 7850 ICP-MS in sample mode 25 times in HNO310%. The method used is as follows for each of the 4 elements a concentration value. Standard 1 negative control: 1 blank composed of HNO31%. Standard 2 positive controls: 1 standard of N emt 28 positive control elements (different from those at 20 ppb in multi-element HNO3 (VWR ®reference 85006.186), certified at 100 mg / L of Al, Ag, As, B, Ba, Be, Bi, Ca, Cd, Co, Cr, Cu, Fe, K, Li, Mg, Mn, Mo, Na, Ni, Pb, Sb, Se, Sr, Ti, Tl, V and Zn. For the 28 elements present in standard 2, the concentrations are calculated from the analyte minerals, the concentrations are interpolated from those of the positive control elements present in the calibration standard 2, by applying response factors, which depend on their mass and

[0115] Intra-laboratory validation Linearity: The linearity of the response is verified using 6 standards whose concentrations are in line with the concentrations found in the samples, i.e. 18 elements distributed over the range of masses and concentrations Al: 5,250ppb; As: 0.1 2.5 ppb; B: 50 2500 ppb; Ca: 120 2500 ppb; Cd: 0.01 2.5 ppb; Co: 0.01 2.5 ppb; Cu: 1250 bp; Fe: 1,250 ppb; K: 125 25000 ppb; Li: 0.2 25 ppb; Mg: 12.5 2500 ppb; Mn: 1,250 ppb; Na: 12.5 2500 ppb; Pb: 0.1 25 ppb; Ti: 0.2 25 ppb; U: 0.01 -2.5 ppb; V: 0.1 25 ppb; Zn: 2,250ppb. Limits of detection and quantification These were calculated on a reference sample from 20 measurements LOD = ^blank + 5^blank LOD = ^blank + 10^blank With µblank corresponding to the blank signal and ^blank, to its standard deviation. Precision: ^ Repeatability at the start of the sequence, then every 20 analyses, in order to control the repeatability e.g. of 5% for the major ones (e.g. ^ 1 mg / L) and 10% on the traces. ^ Intermediate precision: This same reference sample, analyzed every day, -this with the previous reference and this, on eg at least 4 analyses / day for eg10 days. ^ Comparability: precision, so that each sample is always defined by the same response factors expressed in equivalent concentrations. Thus, systematic errors are accepted but not random errors. If the accuracy of the data cannot be established, their traceability can be guaranteed in relation to an established reference. ^ Data analysis: variations in this population are less than eg 3 times this tolerance are not taken into account in data processing. For which the indicators are less than the limits of quantification are not taken into account in data processing.

[0116] It is nevertheless possible to carry out a more precise quantification of certain elements of interest. To do this, a calibration range specific to the concentrations of the elements is carried out. The semi-quantitative and quantitative results are then obtained in parallel and in a single acquisition for a sample and can be compared.

[0117] For is used: Systematic cleaning of the ICP-MS supply lines to avoid any contamination and to ensure a good analysis of trace elements in particular. Systematic verification of the device performance. Verification of the semi-quantitative model by analyzing a reference wine at the beginning, middle and end of each analysis sequence. Monitoring of the analyses by regularly analyzing the 28-element standard throughout the sequence. Monitoring of the concentration of the internal indium standard to prevent drifts related to the equipment.

[0118] Figure 2 shows the distribution of concentrations [µg / L] of 31 mineral elements in CSF samples from the BDR.

[0119] -S1-

[0120] A sample (stored at -18°C -S3-) from a subject S x whose diagnosis is still undetermined within the GNI group according to Table 1 above, is used.

[0121] -S2-

[0122] We assign the data D sx following: (ii) conventional clinics: gender and age of S x corresponding to this sample.

[0123] -S4- The PM of this sample is analyzed as described above for the (NS m=0 (d,p,t) +N (d,p,t)) 1296 healthy subjects Sm=0(d5) or sick (d1.1;1.2;1.3;1.4;2.1;2.2;2.3;3;4.1) of reference of this example.

[0124] The detailed results are presented in Table 2 below, specifying the measured mineral concentrations in ppb, in this sample corresponding to the subject Sx to be tested as well.

[0125] [Table 2] Concentration in ppb of different minerals

[0126] -S8- Define at least two targets each comprising at least one category and / or at least one subdivision (eg subcategory) to do at least

[0127] In this example 1, we define 3 categories, 3 subcategories and 3 subcategories of the other category. These 3 categories and these 6 subcategories are targets for which the method allows us to predict whether S x belongs to one or more

[0128] x to none, one or more targets depending on the data and the categories / subcategories chosen, S xcan belong to 0, 1 or more categories / subcategories. The physician can participate in defining the targets (categories / subcategories) of interest to aim at on the subject S x .

[0129] In this example, we predict x to none, one or more of the following targets: C1: Category S m=0 (d5) healthy subjects free from neurological diseases: premium controls C2: Category_ (d1) subjects suffering from neurodegenerative diseases and under- categories: (d1.1 ) Alzheimer's dementia; (d1.2) non-Alzheimer's dementia; (d1.3) Parkinson's (disease / syndrome); (d1.4) other neurodegenerative diseases C3: Category_ (d2) subjects with inflammatory diseases of the CNS and subcategories (d2.1)pathogen-related inflammatory CNS diseases, including meningitis, encephalitis, radiculitis, due to e.g. VZV; (d2.2) multiple sclerosis -MS- inflammatory CNS diseases, including CIS; (d2.3) non-pathogen-related inflammatory CNS diseases (without MS), including immune neuropathies, autoimmune encephalitis, myasthenia gravis. C4: Subcategory_ (d4.1)

[0130] -S9.1- Statistical processing:

[0131] -S9.1.1- Data collection: *Data (i) concentrations of mineral elements *Data (ii) conventional clinical: age; gender: 0 for male and 1 for female

[0132] -S9.1.2- Normalization of StandardScaler data values

[0133] -S9.1.3- [Variant V1 of 1 er protocol] using “Volcano Plot” curves:

[0134] For the selection of data (i) concentrations of mineral elements and -i.e. the most differentiating, we use in this example, Plot for the PM of the CSF of healthy subjects S m=0 (d5) or sick (d1.1;1.2;1.3;1.4;2.1;2.2;2.3;3;4.1) of reference of this example.

[0135] each target category C1, C2, C3 or C4, for example the categories / subcategories corresponding to the subjects S md1, d2 or d3, or d4.1. So-called volcano plots represent categories. This type of graph allows you to quickly and efficiently visualize significant differences between two categories / subcategories or targets / subtargets. On a volcano plot, the x-axis (horizontal axis) generally represents the degree of change, while the y-axis (vertical axis) indicates the statistical significance of this change, expressed as a P value. The points located at the upper ends of the graphs indicate the elements with the most significant differences. marked and most statistically significant. The curve represents the P value in cypi.stats on python3.

[0136] Thus, the attached figures 3A, 3B, 3C, 3D, 3E, 3F represent volcano plots which compare: - either the concentrations of mineral elements in the CSF of healthy subjects S m=0d5 of reference C1, to the concentrations of the mineral elements of the CSF of reference sick subjects suffering from neurological diseases C2; - or, between them, the concentrations of the mineral elements of the CSF of reference sick subjects suffering from different neurological diseases C2,C3; - Fig.3A the concentrations of the mineral elements of the CSF of healthy subjects S m=0 d5 of reference C1 [C1_category d5], to the concentrations of mineral elements in the CSF of sick subjects d1.2 of reference C2 (non-Alzheimer's dementia) [C2_subcategory d1.2]; - ​​Fig.3B the concentrations of mineral elements in the CSF of healthy subjects S m=0 d5 of reference C1 [C1_category d5], to the concentrations of mineral elements in the CSF of sick subjects d1.3 of reference C2 (Parkinson's disease, Parkinson's syndrome) [C2_subcategory d1.3]; - Fig.3C the concentrations of mineral elements in the CSF of healthy subjects S m=0d5 of reference C1 [C1_category d5], to the concentrations of mineral elements of the CSF of sick subjects d1 of reference C2 (neurodegenerative diseases) [C2_category d1]; - Fig.3D the concentrations of mineral elements of the CSF of healthy subjects S m=0 d5 of reference C1 [C1_category d5], to the concentrations of mineral elements in the CSF of sick subjects d2 reference C3 (inflammatory diseases of the CNS), [C3_category d2];. - Fig.3E the concentrations of mineral elements in the CSF of healthy subjects S m=0 d5 reference C1 [C1_category d5], to the concentrations of mineral elements in the CSF of sick subjects d1.1 reference C2 (Alzheimer's dementia) [C2_subcategory d1.1]; - Fig.3F the concentrations of mineral elements in the CSF of sick subjects d1 reference C2 (neurodegenerative diseases) [C2_category d1], to the concentrations of mineral elements in the CSF of sick subjects S m d2reference C3 (inflammatory diseases of the CNS) [C3_category d2].

[0137] For each of the categories identified in relation to the controls, we see that they are different, i.e. influenced by the given category. This influence can be in a positive direction (the more there is negative (the more there is strong intensity. This means. These differentiating roles are not the same depending on the categories. The dotted red line in Figures 2A, 2B, 2C, 2D, 2E, 2F represents the limit corresponding to a P Value of 0.05; commonly considered as a significant limit for an individual value.

[0138] Figure 3A: Dementia (excluding Alzheimer's dementia) [C2_subcategory d1.2] versus premium controls [C1_category d5]. Differentiating minerals identified as significantly associated with dementia include Cu, Fe, P, Sr, Zn, Cr, S, Al, Pb, and are selected based on their correlation strength according to Mann-Whitney U statistical analysis, while Gd, Mg, and Tb have a negative effect on the course or prevalence of dementia, as evidenced by the negative correlation values ​​obtained and the corresponding positions on the volcano plot.

[0139] Figure 3B: Parkinson's disease and Parkinson's syndrome [C2_subcategory d1.3] compared to premium controls [C1_category d5]. Differentiating minerals identified as significantly associated with Parkinson's disease include Cu, Fe, P, S, Al, Cr, V, Zn and are selected based on their correlation strength according to Mann-Whitney U statistical analysis, while Mg and Na, have a negative effect on the course or prevalence of Parkinson's disease, as evidenced by the negative correlation values ​​obtained and the corresponding positions on the volcano plot.

[0140] Figure 3C: Neurodegenerative diseases [C2_category d1] versus premium controls [C1_category d5]. Differentiating minerals identified as significantly associated with neurodegenerative diseases include Cu, Fe, P, Sr, Zn, Cr, S, Al, Pb, and are selected based on their correlation strength according to Spearman statistical analysis, while Gd, Mg, Tb, Cl and Na have a negative effect on the evolution or prevalence of neurodegenerative diseases, as evidenced by the negative correlation values ​​obtained and the corresponding positions on the volcano plot.

[0141] Figure 3D: Inflammatory CNS diseases [C3_category d2] compared to premium controls [C1_category d5]. Differentiating minerals identified as significantly associated with this type of diseases include Cu, Zn, S, P, K, Fe, Rb, Cs, Sr, and are selected based on their correlation strength, according to the Mann-Whitney U statistical analysis, while Mg and Cl have a negative effect on the evolution or prevalence of this type of disease, as evidenced by the negative correlation values ​​obtained and the corresponding positions on the volcano plot.

[0142] Figure 3E: Alzheimer's dementia [C2_subcategory d1.1] versus premium controls [C1_category d5]. Differentiating minerals identified as significantly associated with this type of disease include Cu, Fe, Cr, Ni, Zn, Cs, and are selected based on their correlation strength according to Mann-Whitney U statistical analysis, while Mg, Tb, Na, Gd, Cr have a negative effect on the evolution or prevalence of this type of disease, as evidenced by the negative correlation values ​​obtained and the corresponding positions on the volcano plot.

[0143] Figure 3F: Neurodegenerative diseases [C2_category d1] versus inflammatory CNS diseases [C3_category d2]. Differentiating minerals identified as significantly associated with neurodegenerative diseases and negatively correlated with inflammatory diseases include Fe, Sr, Cr, Al, and are selected based on their correlation strength according to Spearman statistical analysis, and Gd, Mg, Tb, Cl, and Na are identified as significantly associated with inflammatory diseases and negatively correlated with neurodegenerative diseases.

[0144] The correlations between the different data (ii) conventional clinical: age; gender: 0 for male and 1 for female; and the different targets are analyzed. , because its correlation with certain pathologies is too important and overwhelms the other information of the volcano plot. We also note that for figures 3A, 3B, 3C and 3D, gender has a significant effect on the evolution or prevalence of certain diseases.

[0145] The differentiating minerals selected from Figures 3A to 3F help to better clarify the relationship between various minerals and associated diseases. They are not necessarily selected for the statistical processing -S9.1- and the IA calculation -S9.2- described above. An RFE model associated with the calculation model used.

[0146] -S9.1.3- [Variant V2 of 1 erprotocol] This variant V2 is tested on three cases -S8-, of these targets, which a category. These 3 cases correspond respectively to: - 1 er scenario: C1 S targets m=0 (d5) healthy subjects free from diseases neurological [C1_category d5]: premium controls & C2(d1;1.1;1.2;1.3)subjects suffering from neurodegenerative diseases[C2_category d1]; - 2 ème scenario: targets C2(d1;1.1;1.2;1.3)subjects suffering from neurodegenerative diseases [C2_category d1] & C3(d2;2.1;2.2;2.3)subjects suffering from inflammatory diseases of the CNS [C3_category d2]. - 3 ème scenario: targets C2 (d1;1.1;1.2;1.3) subjects suffering from neurodegenerative diseases [C2_category d1], C3 (d2;2.1;2.2;2.3) subjects suffering from inflammatory diseases of the CNS [C3_category d2], & C1 S m=0 (d5) healthy subjects free from neurological diseases [C1_category d5]: premium controls.

[0147] -S9.1.3.1- Model Definition A group of 8 machine learning and AI prediction models are defined as follows: Random Forest, Logistic Regression, Gradient Boosting Machine, Decision tree, Conditional Random Forest, Linear SVM, Ridge Classifier and XGBoost. These models are configured with their default parameters

[0148] -S9.1.3.2- Data separation We divide the data set (i) & (ii) for the 1 er and the 2 ème scenario, in training (80%) and test (20%) parts.

[0149] -S9.1.3.3- Training and evaluation loop Application of 5-fold cross-validation with StratifiedKFold for training and evaluating models, calculating for each iteration the false positive rates (FPR) and true positive rates (TPR).

[0150] -S9.1.3.4- Plotting the ROC curve for each model and calculating the average of We plot the ROC curve for each model using the averages of the TPRs and FPRs, and we calculate the average of the AUC over the iterations of the cross-validation.

[0151] The ROC curves of the 2nd case for the 8 tested models are shown in Figure 4A (curves A to I).

[0152] The ROC curves of the 2nd case for the 8 tested models are shown in Figure 4B (curves A to I).

[0153] In the legend of these two figures 4A and 4B, the AUC obtained for each model tested are shown as well as the number n corresponding to the number of categories for each model.

[0154] -S9.1.3.5- C based on best AUC score.

[0155] The best AUC score measuring the ability of each model to differentiate between C2 / C1 targets (Figure 4A) and C2 / C3 targets (Figure 4B), is obtained, respectively, for the Conditional Random Forest model [C2 / C1 targets (Figure 4A)] and for the Gradient Boosting Machine model [C2 / C3 (Figure 4B)], which are therefore chosen.

[0156] x selected automatic and AI prediction models, is used for the selection of differentiating data (i) & (ii).

[0157] Data (i) & (ii) selected according to -S9.1.3- are: - for the 1 er case C1 / C2-: . Data (i) mineral elements: B, Al, Fe, Rb, S . Data (ii) conventional clinical: age - for the 2 ème case C2 / C3-: . Data (i) mineral elements: P, K, Cr, Zn, Rb. . Data (ii) conventional clinical: age - for the 3 èmecase C1 / C2 / C3-: . Data (i) mineral elements: P, S, Cl, Cu, Zn, Rb. . Data (ii) conventional clinical: age These data [Variant V2 of 1 er protocol], rather than by the volcano plot curves [Variant V1 of 1 er protocol], are displayed and used for dimensionality reduction-S9.1.4-.

[0158] -S9.1.4- 2D or 3D stochastic dimension reduction We use -SNE (t-distributed stochastic neighbor embedding) to - reduce the dimensionality of the targeted data by compressing the multi-dimensional data into two or three main components, - and facilitate their visualization in 2D or 3D depending on the selected data and improve the distinction between the targeted categories.

[0159] -S9.1.5-Visualization on graphs of targets (targeted data) (-(i)- concentrations of different minerals more differentiating / relevant for process performance.

[0160] A KDE kernel density plot (Kernel Density Estimate) is produced to visualize the distribution of targets, 1 er case C2 / C1, 2 ème cases C2 / C3, and 3 ème C1 / C2 / C3 cases, as defined previously. To do this, the components extracted following the application of the t-SNE algorithm are plotted respectively on the abscissa axes and the ordinate axis.

[0161] A scatter plot is also overlaid to show individual points.

[0162] The attached Figures 5A, 5B and 5C illustrate, in the form of two-dimensional graphs, the distributions between the different combinations of targets, respectively, C2 / C1; C2 / C3; C1 / C2 / C3 defined above: - Figure 5A: C1 S m=0(d5) [C1_category d5] versus C2 (d1;1.1;1.2;1.3) Figure 5B: C2 (d1;1.1;1.2;1.3) [C2_category d1] versus C3 (d2;2.1;2.2;2.3) [C3_category d2]; Figure 5C: Combined comparison of the three targets C1 S m=0 (d5) [C1_category d5]; C2 (d1;1.1;1.2;1.3) [C2_category d1] and C3 (d2;2.1;2.2;2.3) [C3_category d2].

[0163] Regions where the estimated density is equal to or greater than 50% of the maximum value will be displayed, with different shades or colors representing different density ranges.

[0164] Statistical processing -S9.1- allows for optimal separation of categories, for example by t-SNE. Figures 5A / 5B / 5C represent the different targets on 2-dimensional graphs, in order to best separate the points representing the different subjects S m=0 (d5) [C1_category d5]; (d1;1.1;1.2;1.3) [C2_category d1] (d2;2.1;2.2;2.3) [C3_category d2].

[0165] This visualization helps to distinguish between different targets based on selected parameters, such as concentrations of certain mineral elements.

[0166] -S9.1.6-Positioning

[0167] Data localization D sx (i) formed by the ppb concentrations of the different minerals (Table 2 above) and D data sx (ii) formed the gender and age corresponding to the subject S x to be tested, are integrated into the three previously established t-SNE graphs (Figures 5A / 5B / 5C), with a distinctive red cross to facilitate its identification. These are the attached Figures 6A / 6B / 6C.

[0168] Figure 6A: Healthy Sm=0(d5) subjects [C1_category d5] compared to subjects (d1;1.1;1.2;1.3) with neurodegenerative diseases [C2_category d1]. The red cross indicates the positioning of the mineral profile.

[0169] Figure 6B: Subjects (d1;1.1;1.2;1.3) with neurodegenerative diseases [C2_category d1] compared to subjects (d2;2.1;2.2;2.3) with inflammatory CNS disease [C3_category d2]. selected: [P, K, Cr, Zn, Rb, age]. The cross mineral profile.

[0170] Figure 6C: Healthy Sm=0(d5) subjects [C1_category d5] compared to subjects (d1;1.1;1.2;1.3) with neurodegenerative diseases [C2_category d1] and subjects (d2;2.1;2.2;2.3) with inflammatory CNS disease [C3_category d2. Selected characteristics: [P, S, Cl, Cu, Zn, Rb, age] the red cross indicates the l. On each of the t-SNE maps in Figures 5A / 5B / 5C, the red cross indicating the position of sample ID5116 of subject S xto be tested, is clearly located within the target (region) associated with subjects (d1;1.1;1.2;1.3) suffering from neurodegenerative diseases. This location strongly suggests that the PM mineral profile of CSF sample ID5116 from subject S x to be tested, corresponds to the typical characteristics of neurodegenerative disorders, which could point towards a diagnosis in this category of diseases.

[0171] Exemple 2 : Diagnostic d sujet Sx à tester au moyen de automatic and AI prediction (2 ème protocol -S9- from data (i) & (ii) historical healthy reference subjects S m=0 and / or sick from data (i) & (ii) x to test.

[0172] -S1-

[0173] sample (stored at -18°C -S3-) from subject Sx at GNI according to table 1 above).

[0174] -S2-

[0175] We assign the data D Sx (ii) conventional clinics: gender and age of Sx corresponding to this sample.

[0176] -S4- The PM of this sample is analyzed as described above for the N t 1296 healthy subjects S m=0 (d5) or sick(d1.1;1.2;1.3;1.4;2.1;2.2;2.3;3;4.1)of reference of this example.

[0177] The detailed results are presented in Table 2 presented previously, specifying the measured mineral concentrations in ppb, in this sample corresponding to the subject Sx to be tested.

[0178] -S8- Define at least two targets of the subject Sx to be tested to at least one target

[0179] Dans cet exemple 2, on reprend les 3 targets C1 including the category [Sm=0(d5)], C2 including the category [(d1] with its 4 subcategories (d1.1), (d1.2), (d1.3) & (d1.4) & C3 including the category [(d2], with its 3 subcategories (d2.1), (d2.2), (d2.3). These 3 categories and these 6 subcategories are targets, with respect to which the method makes it possible to predict whether S x belongs to a

[0180] xto none, one or more targets depending on the data and the categories / subcategories chosen, S x can belong to 0, 1 or more categories / subcategories. The physician can participate in defining the targets (categories / subcategories) of interest to aim at on the subject S x .

[0181] Dans cet exemple 2 x à aucune, une ou plusieurs cibles, est predicted according to the same first 2.

[0182] 2 ème protocol -S9.2- of -S9- .

[0183] -S9.2.1- Collection of data values: *Data (i) concentrations of mineral elements *Data (ii) conventional clinical: age; gender: 0 for male and 1 for female.

[0184] -S9.2.2- Normalization of StandardScaler data values

[0185] -S9.2.3- 7 and AI prediction.

[0186] -S9.1.3.1- .

[0187] -S9.2.4- Data separation:

[0188] -S9.1.3.2-

[0189] -S9.2.5- Learning and evaluation loop:

[0190] -S9.1.3.3-

[0191] -S9.2.6- Plotting the ROC curve:

[0192] -S9.1.3.3-

[0193] s 8 models of Figure 4A (curves A to I) are given in this figure.

[0194] B (curves A to I) is given in this figure.

[0195] In the legend of these two figures 4A and 4B, n corresponds to the number of for each model.

[0196] automatic and prediction by AI, we can with certain reliability, to category C1, C2 and / or C3.

[0197] -S9.2.7- AI model selection (same as -S9.1.3.5-) based on the best AUC score measuring the ability of each model to differentiate categories / targets. The Conditional Random Forest model is selected for targets C2 / C1 (Figure 4A) and the Gradient Boosting Machine model C2 / C3 (Figure 4B) the best AUC score in these 2 cases.

[0198] -S9.2.8- Selection of data (i) & (ii):

[0199] The data to be selected are the categories / subcategories of the reference sick subjects and / or the healthy subjects S m=0 reference, the concentrations of the different minerals [data (i)] and the conventional clinical data (ii) associated with these reference subjects most differentiating / relevant for the performance of the Conditional Random Forest or Gradient Boosting Machine model selected in -S9.2.7-, thanks to the RFE estimator associated with these models.

[0200] by AI Conditional Random Forest or Gradient Boosting Machine chosen, is also used for the selection of differentiating data (i) & (ii). These data are: - for the 1 er case C1 / C2-: . Data (i) mineral elements: B, Al, S, Cl, Ni, Zn, Rb, I, Tb . Data (ii) conventional clinical: age - for the 2 èmecase C2 / C3-: . Data (i) mineral elements: B, Na, Mg, P, Cr, Co, Cu, Rb, Tb . Data (ii) conventional clinical: age are displayed and used for

[0201] Confusion matrices (Figures 7A & 7B attached), showing true positives, false positives, true negatives and false negatives, are generated from the test dataset (representing 20% ​​of the entire dataset).

[0202] The sensitivity and specificity of the model are calculated. The hyperparameters are then adjusted to optimize the model's performance.

[0203] The confusion matrices (Figures 7A & 7B attached) for this example 2 were performed on 20% of the 815 subjects belonging to at least one of the two targets C1 / C2 or C2 / C3, i.e. 163 subjects. The average sensitivity obtained is 0.79. The average specificity obtained is 0.78. In addition, the average AUC score obtained is 0.85. - S9.2.7-

[0205] We put Conditional Random Forest or Gradient Boosting Machine on the selected data of subject Sx and the selected data of the target categories of the reference subjects.

[0206] The prediction is displayed, along with the sensitivity, specificity, and AUC score of the Conditional Random Forest or Gradient Boosting Machine model for the analyzed data.

[0207] As shown in Figures: 7A: Confusion matrix target C2 subjects with neurodegenerative diseases versus target C1 healthy subjects; Sensitivity: 0.7903225806451613; Specificity: 0.7821782178217822; AUC score: 0.8510060683487704 & 7B: Confusion matrix target C2 subjects with neurodegenerative diseases versus target C3 subjects; Sensitivity: 0.7903225806451613; Specificity: 0.7821782178217822; AUC score: 0.8510060683487704

[0208] The Conditional Random Forest and Gradient Boosting Machine models reliably predict C2 neurodegenerative disease targets, AUC- ROC = 85%, based on the selected data for the chosen targets.

[0209] -S9.2.9- from subcategory d.4 (unknown from [Table 1] above) -S9.2.7-

[0210] Conventional clinical data (ii) are supplemented by the sex of the 655 subjects and by quantitative data on plasma components measured in plasma samples from these 655 Sx subjects. [Table 3] below lists these plasma components.

[0211] [Table 3]

[0212] Conventional clinical data (ii) as well as data (i) formed by the most differentiating concentrations of mineral elements, are chosen by, used: Conditional Random Forest or Gradient Boosting Machine for each of the chosen categories / targets, the Conditional Random Forest or Gradient Boosting Machine model of 252 of the 655 Sx subjects with a specificity of more than 95%, as follows: * 93 to the target C2 subjects category (d1) suffering from neurodegenerative diseases, * 29 in the C3 target category subjects (d2) suffering from inflammatory diseases of the CNS, * 40 target C2 subjects category (d1) subcategories: (d1.1 ) Alzheimer's dementia; ( d1.2) non-Alzheimer's dementia, * 31 at target C2 subjects category (d1), subcategory (d1.4) other neurodegenerative diseases * and 59 to target C1 subjects category S m=0 (d5) healthy, free from neurological diseases: premium controls.

[0213] Example 3: Prognosis / Therapeutic follow-up x to be tested through 1 er protocol -S9.1- -S9.2- :) to of data (i) & of data (i) & (ii) x to test

[0214] -S0-

[0215] -S0.1- Conventional clinical data (ii)

[0216] A database is constructed from 379 plasma samples from 379 ERB2-positive female subjects treated with antibodies (Trastuzumab_Herceptin) and (JAMA Oncol.2022 May 1;8(5):698-705. ) Efficacy of HD201 vs Referent Trastuzumab in Patients WithERBB2- Positive Breast Cancer Treated in the Neoadjuvant Setting. A Multicenter Phase 3 Randomized Clinical Trial”.

[0217] Each of the reference subjects suffering from breast cancer (ERB2 positive), through conventional clinical data concerning the subjects: . to a target including category S m (d1) corresponding to a diagnostic criterion S m (d) specific (criterion 1): ECOG oncology group index; and to a subcategory S m (d1.1) ECOG 0 or subcategory S m (d1.2) ECOG 1; . to a target including category S m (d2) corresponding to a specific diagnostic criterion (criterion 2) ) : Largest clinical diameter of the tumor; and to a subcategory S m (d2.1) >30mm, or to a subcategory S m (d2.2) between 10 and 30 mm or to a subcategory S m (d2.3) Not interpretable ; . to a target including category S m (d3)corresponding to a specific diagnostic criterion (criterion 3): number of lymph nodes; and to a subcategory S m (d3.1) >3, or to a subcategory S m (d3.2) between 1 and 3, or to a subcategory S m (d3.3) none, or to a subcategory S m (d3.4) Not interpretable ; . to a target including category S m (p1) corresponding to a specific prognostic criterion (criterion 1): Complete local pathological response (tpCR) and to a subcategory S m (p1.1) Yes p=0, or to a subcategory S m (p1.2) Not p=1, or to a subcategory S m (p1.3) Not interpretable; . to a target including category S m (p2) corresponding to a specific prognostic criterion (criterion 2): Complete pathological response of the breast at the time of surgery (bpCR) and to a subcategory S m (p2.1)Yes p=0, or to a subcategory S m (p2.2) Not p=1, or to a subcategory S m (p2.3) Not interpretable; . to a target including category S m (p3) corresponding to a specific prognostic criterion (criterion 3) a subcategory S m (p3.1) Yes p=0, or to a subcategory S m (p3.2) No p=1; . to a target including category S m (t1) corresponding to a specific therapeutic monitoring, given by the measured residual trastuzumab concentration (Ctrough), and to a subcategory S m (t.1.1) Ctrough on cycle 5, or to a subcategory S m (t.1.2) Ctrough on cycle 8, or to a subcategory S m (t.1.3) Ctrough at the end of processing. These targets taken into account as part of the decision-making process involving statistical representations or artificial intelligence, specific to the process according to

[0218] For the reference sick subjects (d,p,t) of this example 3, the targets or categories and the sub-targets or sub-categories are given in table 4 below.

[0219] [Table 4] BLQ: Below the limit of quantification.

[0220] -S0.2- constituted by the Mineral Profile (MP)

[0221] The database is completed for the N t = 379 subjects (d,p,t) sick (d1.1; d1.2; d1.3; d2.1; d2.2; d2.3; d3.1; d3.2; d3.3; d3.4; p1.1; p1.2; p1.3 p2.1; p2.2; p2.3 p3.1; p3.2; t1.1; t1.2; t1.3; of reference of this example 3, by the mineral profile PM of concentrations of N em = 34 mineral elements, established for the corresponding plasma samples.

[0222] This mineral profile on plasma samples was performed as described in

[0223] Figure 8 shows the distribution of concentrations [µg / L] of 34 plasma mineral elements from the BDR.

[0224] -S1-

[0225] A sample (stored at -18°C -S3-) from a subject S x 112-003-001, the therapeutic response of which is still undetermined, is used.

[0226] -S2-

[0227] We assign the data D sx following -003-001 and conventional clinical data (ii): weight and age of S x corresponding to this sample.

[0228] -S4- The PM of this sample is analyzed as described above for the N t = 379 subjects (d,p,t)) sick (d1.1 ;d1.2 ; d1.3 ;d2.1 ;d2.2 ;d2.3 ;d3.1 ;d3.2 ;d3.3 ;d3.4 ;p1.1 ; p1.2 ;p1.3 p2.1 ; p2.2 ;p2.3 p3.1 ; p3.2 ; t1.1 ;t1.2 ;t1.3 of this example 3.

[0229] The detailed results are presented in Table 5 below, specifying the measured mineral concentrations in ppb, in this sample corresponding to the subject Sx to be tested as well.

[0230] [Table 5] Concentrations in ppb of different minerals / Weight / Age

[0231] -S8- Define at least two targets at least one category and , to make at least a target

[0232] In this example 3, we define 3 categories f ( p1) ; (p2) ; (p3) 3 subcategories of the category (p1) :(p1.1); (p1.2); 3 subcategories of the category (p2) :(p2.1) ; (p2.2) ;2 subcategories of the category (p3) :(p3.1); (p3.2). These 3 categories and 6 subcategories are intended to form targets vis-à-vis

[0233] targets depending on the data and the categories / subcategories chosen, Sx can belong to 0, 1 or several targets. The doctor can participate in defining the targets (categories / subcategories) of interest to aim at on the subject Sx.

[0234] to one or more targets, is predicted in 3 cases.

[0235] 1st scenario: Complete local pathological response (tpCR) Target C1*: Category (p1) subcategory (p1.1) p=0 Yes Target C2*: Category (p1) subcategory (p1.2) p=1 No.

[0236] 2nd scenario: Complete pathological response of the breast at the time of surgery (bpCR) Target C3*: Category (p2) subcategory (p2.1) p=0 Yes Target C4*: Category (p2) subcategory (p2.2) p=1 No.

[0237] 3rd scenario: Recurrence of the disease during the study or similar (Presence Target C5*: Category (p3) subcategory (p3.1) p=0 Yes Target C6*: Category (p3) subcategory (p3.2) p=1 No.

[0238] 2 ème S9.2- protocol of -S9- .

[0239] -S9.2.1- Collection of data values: *Data (i) mineral element concentrations: see figure 8 & table 5. *Data (ii) conventional clinical: age; weight: table 5.

[0240] -S9.2.2- Normalization of data values ​​e StandardScaler

[0241] -S9.2.3- 4 by AI.

[0242] These 4 models are random forest, logistic regression, XGBoost and decision tree, configured with their default parameters.

[0243] -S9.2.4- Data separation: Division of the entire data set into training (80%) and test (20%) parts.

[0244] -S9.2.5- Learning and evaluation loop:

[0245] Application of 5-fold cross-validation with StratifiedKFold for model training and evaluation, calculating for each iteration the false positive rates (FPR) and true positive rates (TPR).

[0246] -S9.2.6- Plotting the ROC curve:

[0247] Plotting the ROC curve for each model using the averages of the TPRs and FPRs, and calculating the average AUC over the cross-validation iterations.

[0248] Figure 9 attached shows the 8 ROC curves (curves A to I) corresponding to the 8 for the resumption of the disease during Target C5*: Category (p3) subcategory (p3.1) p=0 Yes Target C6*: Category (p3) subcategory (p3.2) p=1 No.

[0249] The AUCs for the 8 models in Figure 9 (curves A to I) are given in this figure.

[0250] In the legend of Figure 9, n corresponds to the number of data categories for each model.

[0251] By studying the PM mineral profile of N plasmas t= 379 subjects (d,p,t)) sick (d1.1;d1.2;d1.3;d2.1;d2.2;d2.3;d3.1;d3.2;d3.3;d3.4;p1.1;p1.2;p2.1;p2.2;p3.1;p3.2;t1.1;t1.2;t1.3) of reference of this example 3 these 4 by AI to the target C5* to the target C6*. automatic and AI prediction to differentiate categories / targets. The AI ​​Gradient Boosting Machine model is selected for C5* / C6* targets (Figure 9

[0254] The data to be selected are the categories / subcategories C5*: Category (p3) subcategory (p3.1) p=0 Yes & C6*: Category (p3) subcategory (p3.2) p=1 No of the reference sick subjects, the concentrations of the different minerals [data (i)] and the conventional clinical data (ii) associated with these reference subjects which are the most differentiating / relevant for the performance of the AI ​​logistic regression model selected in -S9.2.7-, using the RFE estimator associated with the logistic regression model.

[0255] to the logistic regression model, is used for the selection of differentiating data (i) & (ii). These data are: - for the 3 ème scenario C5* / C6*: Recurrence of the disease during the study or similar. Data (i) mineral elements: Na, K, Fe, Zn, Ge, As, Sn, Gd, Pb. . Data (ii) conventional clinical: age in years; are displayed and used for

[0256] The confusion matrix (Figure 10 attached), showing true positives, false positives, true negatives and false negatives, is generated from the test dataset (representing approximately 20% of the entire dataset).

[0257] The sensitivity and specificity of the model are calculated. The hyperparameters are then adjusted to optimize the model's performance.

[0258] The confusion matrix (Figure 10 attached) for this example 3 is carried out on 20% of the 379 subjects belonging to at least one of the two targets C5* / C6*, i.e. 76 subjects. The average sensitivity obtained is 0.45. The average specificity obtained is 0.93. In addition, the average AUC score obtained is 0.69. From the confusion matrix, we see that the chosen logistic regression AI model has predictive capabilities to know the therapeutic response of patients to the treatment. Thus, with an average of 90% of the Sx subjects to be tested, whose prognosis is a recurrence of their disease during the 24 months with the treatment carried out.

[0260] the AI ​​model of logistic regression prediction chosen on the selected data of the subject Sx to be tested and the selected data of the targets C5* / C6* of the reference subjects.

[0261] The prediction, with the sensitivity, specificity and AUC score of the chosen logistic regression model, is displayed for the analyzed data.

[0262] -003-001 on the resumption of the disease (Logistic Regression) is as follows: no resumption of the disease during -003-001, the model, trained on the corresponding database, gives a probability score. If the score indicating the probability of the model is higher than the threshold of the model's prediction, defined according to the need, in terms of specificity and sensitivity, within the limit of the model, then the model predicts that the subject will belong to the selected category. We can choose this threshold to increase the sensitivity or specificity of the model according to the needs. In this example, we opt for prediction (compromise between sensitivity and specificity). In this case -003-disease with . In addition, we have displayed the sensitivity, specificity, the selected characteristics, the AUC score as well as the total number of patients used to make this prediction and the number of targets. Model Probability: 0.18; Prediction Threshold: 0.20; Model Sensitivity: 0.72, Model Specificity: 0.57. Selected Features: Na, Mg, S, Cl, Ga, Ge, Zr, Cs, Ba, La. Average AUC Score: 0.65. Total N: 378. Target N: 61.

[0263] Example 4 x to be tested through 1 er S9.1 protocol (2 ème protocol S9.2 from data (i) & (ii) from data (i) & (ii) urine from this subject S x to test

[0264] -S0-

[0265] -S0.1- Conventional clinical data (ii)

[0266] A database is constructed from 197 samples from 197 reference patients (Blood Cell Supernatant on Renal Dysfunction and Posttransfusion Immunomodulation - TRANSNEPHRON-). The investigators selected cardiac surgery patients for the transfusion frequency and significant post-surgical renal morbidity.

[0267] Each of the reference subjects sick conventional clinical data concerning the subjects: . to a target including category S m (p1) corresponding to a specific prognostic criterion (criterion 1): Renal failure during the stay and to a subcategory S m (p1.1) Yes p=0, or to a subcategory S m (p1.2) No p=1; . to a target including category S m (p2) corresponding to a specific prognostic criterion (criterion 2): Respiratory failure during the stay (lowest P / F in the 5 days <200) and to a subcategory S m (p2.1) Yes p=0, or to a subcategory S m (p2.2) Not p=1, or to a subcategory S m (p2.3) Not interpretable . . to a target including the category corresponding to a prognostic criterion S m (p3) specific criterion 3: Infection during resuscitation and in subcategory S m (p3.1)Yes p=0, or to a subcategory S m (p3.2) Not p=1, or to a subcategory S m (p3.3) Not interpretable; These categories are "targets" or "targets", within the framework of the decision process involving statistical representations or artificial intelligence, specific to impacting the composition of plasma minerals.

[0268] For sick subjects (d,p,t) reference of this example, the targets or categories and the sub-targets or sub-categories are given in Table 6 below.

[0269] [Table 6] P / F is an index of severity of hypoxia in cases of ventilation / perfusion disturbances at the pulmonary level resulting in a shunt effect. The normal value is 95mmHg / 21% = 452. A value below 300 indicates severe impairment of gas exchange within the lungs.

[0270] -S0.2- constituted by the Mineral profile (PM)

[0271] The database is completed for the N t = 197 subjects (d,p,t) ) sick (p1.1; p1.2; p2.1; p2.2; p2.3; p3.1; p3.2; p3.3, p4.1; p4.2); of reference of this example 4, by the mineral profile PM of concentrations of N em = 37 mineral elements, established for the corresponding samples.

[0272] This mineral profile on samples 1, for CSF samples.

[0273] Figure 11 shows the distribution of concentrations [µg / L] of 37 elements in the database.

[0274] -S1-

[0275] A sample (stored at -18°C -S3-) from a subject S x S2023_CRB002, whose prognosis is still undetermined, is used.

[0276] -S2-

[0277] The subject Sx to be tested corresponding to this sample is assigned the data D sxfollowing S2023_CRB002 and conventional clinical data (ii): weight (kg); age (years), smoking (yes= 1 / no=0), diabetes (yes= 1 / no=0), hemoglobin (in g / dL, preoperatively on the assessment before surgery), platelets (in number of platelets / mm3 preoperatively on the assessment before surgery), lymphocytes (in number of lymphocytes / mm3 preoperatively on the assessment before surgery), preoperative creatinine (in micromol / L preoperatively on the assessment before surgery), extracorporeal circulation duration (in minutes: duration of circulatory assistance during the block), clamp duration (in minutes: duration of interruption of aortic flow in minutes), dobutamine maximum dosage (dose in microgram / kg / min in continuous infusion), high lactate (in mmol / L hypoxia marker), low hemoglobin (same in g / dL: lowest value postoperatively), neutrophil count.

[0278] -S4- The PM of this sample is analyzed as described above for the N t = 197 sick subjects (d,p,t)) (p1.1; p1.2; p1.3, p2.1; p2.2; p2.3; p3.1; p3.2; p3.3; p4.1; p4.2) of this example 4.

[0279] The detailed results are presented in Table 7 below, specifying the measured mineral concentrations in ppb, in this sample corresponding to the subject Sx to be tested as well as the conventional clinical data (ii) for Sx to be tested including the .

[0280] [Table 7] Concentration in ppb of different minerals

[0281] -S8- Define at least two targets of the subject Sx to be tested to at least one target (each target comprising at least one category

[0282] In this example 4, we define 4 categories (p1); (p2); (p3); (p4); 3 subcategories of category (p1): (p1.1); (p1.2); (p1.3); 3 subcategories of category (p2): (p2.1); (p2.2); (p2.3); 2 subcategories of category (p3): (p3.1); (p3.2); 2 subcategories of category (p4): (p4.1); (p4.2).

[0283] These 3 categories and 10 subcategories allow you to form targets vis-à-vis

[0284] targets depending on the data and the categories / subcategories chosen, Sx can belong to 0, 1 or several targets. The doctor can participate in defining the targets (categories / subcategories) of interest to aim at on the subject Sx.

[0285] In this example 4 subcategories are predicted in 4 cases.

[0286] 1st scenario: Target C1**: Category (p1) subcategory (p1.1) p=0 Yes Target C2**: Category (p1) subcategory (p1.2) p=1 No.

[0287] 2nd case: (lowest P / F in the 5 days <200) Target C3**: Category (p2) subcategory (p2.1) p=0 Yes Target C4**: Category (p2) subcategory (p2.2) p=1 No.

[0288] 3rd scenario: Target C5**: Category (p3) subcategory (p3.1) p=0 Yes Target C6**: Category (p3) subcategory (p3.2) p=1 No.

[0289] 4 ème according to the first 3 cases, complication respectively, renal, pulmonary or infection during the stay Target C7**: Category (p4) subcategory (p4.1) p=0 Yes Target C8**: Category (p4) subcategory (p4.2) p=1 No.

[0290] -S9.1- Statistical processing: *Data (i) mineral element concentrations: see figure 11. *Conventional clinical data (ii): weight (kg); age (years), smoking (yes= 1 / no=0), diabetes (yes= 1 / no=0), hemoglobin (in g / dL, preoperatively on the assessment before surgery), platelets (in number of platelets / mm3 preoperatively on the assessment before surgery), lymphocytes (in number of lymphocytes / mm3 preoperatively on the assessment before surgery), preoperative creatinine (in micromol / L preoperatively on the assessment before surgery), extracorporeal circulation duration (minutes: duration of circulatory assistance during the block), clamp duration (in minutes: duration of interruption of aortic flow in minutes), dobutamine maximum dosage (dose in microgram / kg / min in continuous infusion), high lactate (in mmol / L hypoxia marker), low hemoglobin (same in g / dL: lowest value postoperatively), neutrophil count.

[0292] -S9.1.2- Normalization of StandardScaler data values

[0293] -S9.1.3- [Variant V1 of 1 er using “Volcano Plot” curves:

[0294] For the selection of data (i) concentrations of mineral elements and -i.e. the most t = 379 subjects subjects (d,p,t) ) sick(p1.1; p1.2; p1.3; p2.1; p2.2; p2.3; p3.1; p3.2; p4.1; p4.2) of reference of this example 4.

[0295] The volcano plots are plotted for the 4 scenarios in the same way as in

[0296] Thus, figures 12, 13, 14 & 15 attached represent volcano plots which correspond to the 4 cases mentioned above.

[0297] platelet count. But we also see impacts of the concentrations of certain mineral elements, so statistically the presence of cadmium will promote the occurrence of zirconium (Zr) and selenium (Se) are linked to the presence of fewer complications, while high concentrations of potassium (K) and lead (Pb) are linked to the occurrence of fewer complications.

[0298] -S9.1.4- 2D or 3D stochastic dimension reduction -SNE (t-distributed stochastic neighbor embedding) to - reduce the dimensionality of the targeted data via the compression of multi-dimensional data into two or three principal components, - and facilitate their visualization in 2D or 3D depending on the selected data and improve the distinction between the targeted categories.

[0299] -S9.1.5-Visualization on graphs of the targets (targeted data) (-(i)- concentrations of the different minerals and -(ii)- conventional clinical data) most differentiating / relevant for the performance of the process.

[0300] A KDE kernel density plot (Kernel Density Estimate) is produced to visualize the distribution of targets 3 ème C5** / C6** cases, as defined previously. To do this, the components extracted following the application of the t-SNE algorithm are plotted respectively on the abscissa axes and the ordinate axis.

[0301] A scatter plot is also overlaid to show individual points.

[0302] The attached Figure 16 illustrates, in the form of two-dimensional graphs, the distributions between the C5** / C6** targets defined above:

[0303] Regions where the estimated density is equal to or greater than 50% of the maximum value will be displayed, with different shades or colors representing different density ranges.

[0304] Statistical processing -S9.1- allows for optimal separation of categories, for example by t-SNE. Figure 16 represents on 2-dimensional graphs, the targets C5** / C6** subjects S m (p3.1) Infection during resuscitation: yes p=0 / target C5**; and S m (p3.2) Infection during resuscitation: no p=1 / target C6**.

[0305] This visualization helps distinguish C5** / C6** targets based on selected data, such as concentrations of certain mineral elements.

[0306] -S9.1.6-Positioning

[0307] Data localization D sx (i) formed by the ppb concentrations of the different minerals (Table 7 above) and D data sx(ii) including (Table 7 above) in particular the weight x to be tested [Urine2023_CRB002], are integrated into the t-SNE graph of figure 16, with a distinctive red cross mark to facilitate its identification.

[0308] This location strongly suggests that the PM mineral profile of the Urine2023_CRB002 sample from subject S x to be tested, corresponds to the typical characteristics of the prognostic criterion S m (p3.1) specific criterion 3: Infection during resuscitation yes p=0 / target C5**, which could point towards this prognosis. .

[0310] -S9.2.1- Collection of data values: *Data (i) mineral element concentrations mineral element concentrations: see figure 11 *Data (ii) conventional clinical: see -S9.1.1- above.

[0311] -S9.2.2- Normalization of StandardScaler data values.

[0312] -S9.2.3- by AI.

[0313] These 4 models are random forest, logistic regression, XGBoost and decision tree, configured with their default parameters.

[0314] -S9.2.4- Data separation:

[0315] Division of the entire data set into training (80%) and testing (20%) parts.

[0316] -S9.2.5- Learning and evaluation loop:

[0317] Application of 5-fold cross-validation with StratifiedKFold for model training and evaluation, calculating for each iteration the false positive rates (FPR) and true positive rates (TPR).

[0318] -S9.2.6- Plotting the ROC curve:

[0319] Plotting the ROC curve for each model using the averages of the TPRs and FPRs, and calculating the average AUC over the cross-validation iterations.

[0320] Figure 17 attached shows the 4 ROC curves corresponding to the 4 models for the 4 ème scenario: occurrence of complications respectively, renal, pulmonary or infection during the stay. Target C7**: Category (p4) subcategory (p4.1) p=0 Yes Target C8**: Category (p4) subcategory (p4.2) p=1 No.

[0321] In Figure 17 XGBoost and decision tree, is 0.636, 0.754, 0.556 and 0.560 respectively. By studying the PM mineral profile of plasmas of N t = 182 subjects (d,p,t)) sick (p1.1; p1.2;p1.3; p2.1; p2.2;p2.3; p3.1; p3.2; p4.1; p4.2) of reference of this example 4, we can determine: Target C7**: Category (p4) subcategory (p4.1) p=0 Yes Target C8**: Category (p4) subcategory (p4.2) p=1 No.

[0322] -S9.2.7- automatic and AI prediction to differentiate categories / targets. The logistic regression AI model is selected for C5* / C6* targets (Figure 17

[0323] -S9.2.8-: The data to be selected using the RFE estimator associated with the regression model are the categories / subcategories of target C7**: Category (p4) subcategory ( p4.1) p=0 Yes & C8**: Category (p4) subcategory (p4.2) p=1 No, reference diseased subjects, concentrations of different minerals [data (i)] and conventional clinical data (ii) associated with these reference subjects the automatic and prediction by AI logistic regression selected in -S9.2.7-.

[0324] used for the selection of differentiating data (i) & (ii). These data are: - for the 4 èmeaccording to the first 3 cases, complication respectively, renal, pulmonary or infection during the stay. Target C7**: Category (p4) subcategory (p4.1) p=0 Yes Target C8**: Category (p4) subcategory (p4.2) p=1 No. * Data (i) mineral elements: Li, S, Cl, K, Cu, Se, Rb, Zr, Gd. * Data (ii) conventional clinical: weight, smoking, platelets, neutrophil count, extracorporeal circulation duration, noradrenaline_maximum_dosage.

[0325] are displayed and used for

[0326] The confusion matrix (Figure 21 attached), showing true positives, false positives, true negatives and false negatives, is generated from the test dataset (representing approximately 20% of the entire dataset).

[0327] The sensitivity and specificity of the model are calculated. The hyperparameters are then adjusted to optimize the model's performance.

[0328] The confusion matrix (Figure 18 attached) for this example 4 was carried out on 25% of the 182 subjects belonging to at least one of the two targets C7** / C8**, i.e. 45 subjects. The average sensitivity obtained is 0.78. The average specificity obtained is 0.71. In addition, the average AUC score obtained is 0.78. From the confusion matrix, we see that the model has predictive capabilities to know the prognosis after, renal, hepatic failure or infection. Thus, with an average AUC of 78%, we can make a good prognosis on almost 4 out of 5 patients. S2023_CRB002 -S9.2.7-

[0330] selected data from subject Sx to be tested and selected data from target categories C7** / C8** of reference subjects.

[0331] The prediction, with the sensitivity, specificity and AUC score of the chosen logistic regression model, is displayed for the analyzed data.

[0332] S2023_CRB002 minus one complication such as renal failure, liver failure or infection, to “Logistic Regression”)

[0333] model, trained on the corresponding database gives a probability score. If the score indicating the probability of the model, is higher than the prediction threshold of the model, defined according to the need in terms of specificity and sensitivity, within the limit of the model, then the model predicts that the subject will belong to the selected target. This threshold can be chosen to increase the sensitivity or specificity of the model according to the needs.

[0334] Here we opted for high sensitivity. In this case, we can say that minus 88%. In addition, we displayed the sensitivity, specificity, selected characteristics, AUC score as well as the total number of patients used to make this prediction and the number of targets. Prediction Urine sample S2023_CRB002: PF ratio > 200, no renal failure or infection: - in Target C8**, Category (p4) subcategory (p4.2) p=1 No; - outside Target C7**, Category (p4) subcategory (p4.1) p=0 Yes

[0335] Model probability: 0.26; prediction threshold: 0.36; Model sensitivity: 0.88, Model specificity: 0.42

[0336] Selected characteristics: ['S', 'Cl', 'K', 'Cu', 'Se', 'Rb', 'Zr', 'Gd', 'WEIGHT', 'PLATELETS', 'NEUTROPHIL COUNT', 'NORADRENALINE_POSO_MAX PO]

[0337] Average AUC Score: 0.77

[0338] N Total: 181

[0339] N Target: 95

[0340] Example 5 1st protocol -S9.1-for implementation -S9.2- follicular fluid test

[0341] -S0-

[0342] -S0.1- Conventional clinical data (ii)

[0343] A database is constructed from 64 (individually collected at the level of all samples taken from the same sick reference subject) follicular fluid samples from 64 + 22 female reference subjects undergoing in vitro fertilization (IVF).

[0344] Each of the reference subjects sick conventional clinical data concerning the subjects: . to a target including the category corresponding to a diagnostic criterion S m (d) specific criterion 1 S m (d1): pathologies; to a subcategory S m(d1.1) endometriosis or subcategory S m (d1.2) tubal; to a subcategory S m (d1.3) DOR or to a subcategory S m (d1.4) reference; or to a subcategory S m (d1.5) various or unknown; . to a target or category corresponding to a diagnostic criterion S m (d) specific criterion 2 S m (d2): Oocyte status (individual fluid); and to a subcategory S m (d2.1) M2, or to a subcategory S m (d2.2) empty; . to a target including the category corresponding to a diagnostic criterion S m (d) specific criterion 3 S m (d3): fertilization; and to a subcategory S m (d3.1) yes, or to a subcategory S m (d3.2) no, or to a subcategory S m (d3.3) lysed, or to a subcategory S m (d3.4) 3PN; . to a target including the category corresponding to a diagnostic criterion S m (d) specific criterion 4 S m(d4): transfer; and to a subcategory S m (d4.1) yes, or to a subcategory S m (d4.2) no, . to a target including the category corresponding to a prognostic criterion S m (p1) specific criterion 1: result of IVF and to a subcategory S m (p1.1) pregnancy yes p=0, or to a subcategory S m (p1.2) pregnancy not p=1, or to a subcategory S m (p1.3) white puncture; or to a subcategory S m (p1.3) IVF failure. These targets are taken into account in the decision-making process involving statistical representations or artificial intelligence, specific to the process according to, in this example 5.

[0345] For sick subjects (d,p,t) For reference purposes of this example, the categories and subcategories are given in Table 8 below.

[0346] [Table 8]

[0347] -S0.2- constituted by the Mineral profile (PM)

[0348] The database is completed for Nt= 64+22 subjects (d,p,t) reference patients(d,p) (d1.1;d1.2;d1.3;d1.4;d1.5;d2.1;d2.2;d3.1;d3.2;d3.3;d3.4;d4.1;d4.2;p1.1;p1.2;p1.3 p1.4); of this example 5, by the mineral profile PM of N concentrations em = 31 mineral elements, established for the corresponding follicular fluid samples.

[0349] This mineral profile on follicular fluid samples was performed as described

[0350] Figure 19 shows the distribution of concentrations [µg / L] of 31 follicular fluid elements from the database.

[0351] -S1-

[0352] A sample (stored at -18°C -S3-) from a subject Sx CHUL4 whose t is still undetermined, is used.

[0353] -S2-

[0354] We assign the data D SxCHUL4 and conventional clinical data (ii): weight and age of Sx corresponding to this sample.

[0355] -S4- The PM of this sample is analyzed as described above for the Nt= 64+22 subjects (d,p,t) sick reference (d,p) (d1.1;d1.2;d1.3;d1.4;d1.5;d2.1;d2.2;d3.1;d3.2;d3.3;d3.4;d4.1;d4.2;p1.1;p1.2;p1.3 p1.4) of this example 5.

[0356] The detailed results are presented in Table 9 below, specifying the measured mineral concentrations in ppb, in this sample corresponding to the Sx subject to be tested as well as the conventional clinical data (ii): age of the Sx subject to be tested CHUL4.

[0357] [Table 9]: Concentration in ppb of different minerals LOD: Limit Of Detection

[0358] -S8- Define at least two targets of the subject Sx to be tested to at least one target.

[0359] In this example 5, we define 5 categories (d1) (d2) (d3) (d4)(p1) 5 subcategories of the category (d1) :(d1.1); (d1.2); (d1.3); (d1.4); (d1.5); 2 subcategories of the category (d2) :(d2.1) ; (d2.2) ) ;4 subcategories of the category (d3) :(d3.1); (d3.2);(d3.3) (d3.4)2 subcategories of the category (d4) :(d4.1) ; (d4.2); 4 subcategories of the category (p1) :(p1.1); (p1.2); (p1.3); (p1.4).

[0360] These 5 categories and 17 subcategories allow you to form targets vis-à-vis

[0361] targets depending on the data and the categories / subcategories chosen, Sx can belong to 0, 1 or several targets. The doctor can participate in defining the targets (categories / subcategories) of interest to aim at on the subject Sx.

[0362] subcategories, is predicted in 4 cases.

[0363] 1st scenario: Target C1***: Category (d1) subcategory (d1.1) endometriosis Target C2***: Category (d1) subcategory (d1.4) reference: control (linked to a), the known liquid.

[0364] 2nd scenario: Target C3***: Category (p1) resulting from IVF subcategory (p1.1) pregnancy yes p=0 Target C4***: Category (p1) subcategory (p1.2) reference: pregnancy no p=1.

[0365] -S9.1- Statistical processing:

[0366] -S9.1.1- : . Data (i) concentrations of mineral elements: see figure 21. . Data (ii) conventional clinical: age.

[0367] -S9.1.2- StandardScaler

[0368] -S9.1.3- use of “Volcano Plot” curves:

[0369] For the selection of data (i) concentrations of mineral elements and - i.e. the most differentiating, we use, in this example, volcano Plot for the Nt= 64+22 subjects (d,p,t) sick of reference (d,p) (d1.1; d1.2; d1.3; d1.4; d1.5; d2.1; d2.2; d3.1; d3.2; d3.3; d3.4; d4.1; d4.2; p1.1; p1.2; p1.3 p1.4); of this example 5.

[0370] The volcano plot is plotted for this example 5, from 1.

[0371] Thus, the attached figure 20 represents a volcano plot which corresponds to 1 er case of this example 5: Target C1***: Category (d1) subcategory ( d1.1) Endometriosis versus Target C2***: Category (d1) subcategory ( d1.4) reference: follicular.

[0372] The more cobalt there is in the follicular fluid, the more the fluid has a probability of pro (d1) subcategory (d1.1) endometriosis: Target C1***.

[0373] S9.1.3- [Variant V2 of 1 erprotocol]: 1 above

[0374] -S9.1.3.1- Definition of models

[0375] -S9.1.3.2- Data Separation

[0378] -S9.1.3.5- based on best AUC score.

[0379] -S9.1.4- 2D or 3D stochastic dimension reduction

[0380] -S9.1.5-Visualization on graphs of the data used (concentration of different minerals and conventional clinical data) most differentiating / relevant for the performance of the process.

[0381] A KDE kernel density plot (Kernel Density Estimate) is produced to visualize the distribution of C3*** / C4*** targets as defined previously -SNE

[0382] A scatter plot is also overlaid to show individual points.

[0383] Figure 21 attached illustrates, in the form of two-dimensional graphs, the distributions between the categories C3*** / C4*** Target C3***: Category (p1) resulting from IVF subcategory (p1.1) pregnancy yes p=0 Target C4***: Category (p1) subcategory (p1.2) reference: pregnancy no p=1.

[0384] Regions where the estimated density is equal to or greater than 50% of the maximum value will be displayed, with different shades or colors representing different density ranges.

[0385] The statistical processing -S9.1- allows an optimal separation of the categories, for example by t-SNE. Figure 21 represents on 2-dimensional graphs, the 2 targets C3*** / C4***, in order to best separate the points representing the different subjects target C3***: Category (p1) IVF outcome subcategory (p1.1) pregnancy yes p=0 and target C4***: Category (p1) subcategory (p1.2) reference: pregnancy not p=1.

[0386] This visualization helps to distinguish between different targets based on selected parameters, such as concentrations of certain mineral elements.

[0387] -S9.1.6-Positioning of the Sx CHUL4 subject to be tested

[0388] Data localization D sx (i) formed by the ppb concentrations of the different minerals (Table 9 above) and D data sx (ii) including (Table 9 above) x to be tested [CHUL4 follicular fluid], are integrated into the t-SNE graph of Figure 21, with a distinctive red cross mark to facilitate its location. This location strongly suggests that the PM mineral profile of the test sample [CHUL4 follicular fluid] from subject S x to be tested, corresponds to the typical characteristics of the C3*** target: Category (p1) resulting from IVF subcategory (p1.1) pregnancy yes p=0, which could point towards this prognosis.

[0389] 2nd protocol -S9.2- -S9- .

[0390] -S9.2.1- Data (i) mineral element concentrations. Table 9 above. Data (ii) conventional clinical: age. Table 9 above.

[0391] -S9.2.2- StandardScaler

[0392] -S9.2.3- by AI.

[0393] These 4 models are random forest, logistic regression, XGBoost and decision tree, configured with their default parameters.

[0395] Division of the entire data set into training (80%) and testing (20%) parts.

[0397] Application of 5-fold cross-validation with StratifiedKFold for model training and evaluation, calculating for each iteration the false positive rates (FPR) and true positive rates (TPR).

[0399] Plotting the ROC curve for each model using the averages of the TPRs and FPRs, and calculating the average AUC over the cross-validation iterations.

[0400] Figure 22 attached shows the 8 ROC curves corresponding to the 8 models: pregnancy or non-pregnancy: Target C3***: Category (p1) IVF outcome subcategory (p1.1) pregnancy yes p=0 Target C4***: Category (p1) subcategory (p1.2) reference: pregnancy not p=1.

[0401] The AUCs for the 8 ROC curves in Figure 22 (curves A to I) are given in this figure.

[0402] In the legend of Figure 22, n corresponds to the number of data categories for each model.

[0403] By studying the PM mineral profile of the follicular fluids of Nt = 64+22 (d,p,t) reference patients (d,p) (d1.1;d1.2;d1.3;d1.4;d1.5;d2.1;d2.2;d3.1;d3.2;d3.3;d3.4;d4.1;d4.2;p1.1;p1.2;p1.3 p1.4); of this example 5 CHUL4, with certain reliability, at C3*** or C4***. automatic and AI prediction to differentiate categories / targets. The logistic regression model is selected for targets C3*** / C4*** (Figure 22

[0405] -S9.2.8-: The data to be selected are those of the categories / subcategories Category (p1) resulting from IVF subcategory (p1.1) pregnancy yes p=0 Category (p1) subcategory (p1.2) reference: pregnancy no p=1, sick or healthy subjects S m=0reference, the concentrations of the different minerals [data (i)] and the conventional clinical data (ii) associated with these reference subjects most prediction by AI logistic regression selected in -S9.2.7-, thanks to the RFE estimator associated with the logistic regression model. is used for the selection of the differentiating data (i) & (ii).

[0406] These data are: - for the 2nd scenario C3*** / C4***: pregnancy or not pregnancy. . Data (i) mineral elements: 1 er test Example 5: Mg, P, Cr, Mn, Zn, Rb, Zr, Cd, I, Tb. 2nd test Example 5: Mg, P, Cr, Mn, Zn, Rb, Y, Zr, Cd, I, Ce, Tb. . Conventional clinical data (ii): age Table 9 above.

[0407] are displayed and used for

[0408] The confusion matrix (Figure 23 attached) from IVF pregnancy vs. no pregnancy, showing true positives, false positives, true negatives and false negatives, is generated from the test dataset (representing approximately 20% of the entire dataset).

[0409] The sensitivity and specificity of the model are calculated. The hyperparameters are then adjusted to optimize the model's performance.

[0410] The confusion matrix (Figure 23 attached) for this example 5 was constructed on 25% of the 58 subjects belonging to at least one of the two targets C3*** / C4***, i.e. 14 subjects. The average sensitivity obtained is 0.84. The average specificity obtained is 0.80. In addition, the average AUC score obtained is 0.82.

[0411] From the confusion matrix, we see that the chosen logistic regression model has predictive capabilities to know at this stage, with 80% reliability, at . CHUL4 -S9.2.7-

[0413] selected data from subject Sx to be tested and selected data from target categories C3*** / C4*** of reference subjects.

[0414] The prediction, with the sensitivity, specificity and AUC score of the chosen logistic regression model, is displayed for the analyzed data.

[0415] trained on the BDR gives a probability score. If the score indicating the probability of the model is higher than the model's prediction threshold, defined according to the need in terms of specificity and sensitivity, within the limit of the model, then the model predicts for the subject pregnancy after in vitro fertilization. This threshold can be chosen to increase the sensitivity or specificity of the model according to the needs.

[0416] (compromise between sensitivity and specificity). In this case, we can say that the

[0417] 1 erTest Example 5: Prediction sample CHUL4: no pregnancy

[0418] Selected data (i): Mg, P, Cr, Mn, Zn, Rb, Zr, Cd, I, Tb

[0419] Model Probability: 0.209; Prediction Threshold: 0.421; Model Sensitivity: 0.78; Model Specificity: 0.85; Average AUC Score: 0.76; N t Total subjects: 57;

[0420] 2nd test Example 5: Prediction sample CHUL4 high sensitivity: no pregnancy

[0421] Selected data (i): Mg, P, Cr, Mn, Zn, Rb, Y, Zr, Cd, I, Ce, Tb

[0422] Model Probability: 0.19; Prediction Threshold: 0.20; Model Sensitivity: 0.89; Model Specificity: 0.49; Average AUC Score: 0.80; N t Total subjects: 57;

[0423] The threshold was reduced to increase the sensitivity of the model, resulting in a reliability close to 89%.

Claims

Claims

1. Method for diagnosis, prognosis and / or therapeutic monitoring of: -S0- subjects from which the biological fluid samples originate; - each sample corresponding to a healthy subject S m=0 reference or to a sick subject S m ^ 1 of reference treated or not, m being a positive natural integer; - each subject S m (d,p,t) of reference corresponding to a diagnostic category S m (d) and / or to a prognosis category S m (p) and / or to a therapeutic response category S m (t); each category may optionally be subdivided into one or more levels; - the database comprising data relating to a total number N t of S subjects m=0 (d,p,t) and (d,p,t) of reference; with N t greater than or equal to - in ascending order of preference - 500, 1000; - these data including for each sample: (i) biological of subjects Sm=0 (d,p,t) and / or (d,p,t) of reference; e m mineral elements; * N em being - in increasing order of preference - greater than or equal to 20, 21, 22, 23, 24, 25, 30; ^ these mineral elements being chosen from the group comprising - advantageously consisting of: Li, B, Na, Mg, Al, Si, P, S, Cl, K, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Ge, As, Se, Br, Rb, Sr, Y, Zr, Nb, Mo, Ru, Rh, Pd, Ag, Cd, In, Sn, Sb, I, Cs, Ba, La, Ce, Pr, Nd, Sm, Eu, Tb, Ta, W, Os, Pt, Au, Tl, Pb, Bi and U; and, preferably in the subgroup comprising - advantageously consisting of: Na, Mg, Al, Si, P, S, Cl, K, Ca, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Br, Rb, Sr, Zr, Sn, I, Cs, Ba, Pb, Bi and U; (ii) conventional clinical data relating to subjects S m=0 (d,p,t) and / or (d,p,t) of reference, these data (ii) being chosen from the group comprising - advantageously constituted by - subjects S m=0 (d,p,t) and / or (d,p,t)reference, data on the therapeutic treatment(s) of subjects S m=0 (d,p,t) and / or (d,p,t) exist(s), lifestyle, concentrations of components of the biological fluid other than mineral elements, in particular concentrations of biomarkers, in proteins, lipids, lipoproteins, red blood cells, white blood cells, platelets; -S1- x S x to be tested; -S2- Possibly assign to each sample E x -S1- data D Sx on the subject S x to test, these data D Sx meeting the same definition as the data (ii) referred to in S0; -S3- Possibly keep at least part of the samples E x determined conditions; -S4- defined in S0(i); -S5- at least once, for all or part of the samples; -S6- -S7- S4,S5,S6; -S8- subject Sx to be tested, to at least one target, each target comprising at least one category -S9- -S9.1-: 1 erprotocol consisting of - of these data; *among which are the PM mineral profiles (ii) comprising at least X differentiating mineral elements, X corresponding to -in ascending order of preference- 3, 5, 10, 15, 15, 20, 25, 30; *and relating to the S m=0 (d,p,t) and / or S subjects m (d,p,t) ; to assign the S subjects m (d,p,t) to at least one target defined in S8; 3D and graphical visualization of the targets; - and Position on the graphical visualization of the targets, the data D Sx on the subject(s) S x to be tested, to visualize the positioning of the Sx(s), in relation to the targets; to assist in the diagnosis, prognosis and / or therapeutic monitoring of diseases, in particular neurological diseases, for the Sx subject(s); and / or, -S9.2- ème protocol consisting of: - IA; - *among which are the PM mineral profiles (ii) comprising at least X differentiating mineral elements, X corresponding to - in increasing order of preference - 3, 5, 10, 15, 15, 20, 25, 30, these X differentiating mineral elements being included among the most efficient for the prediction defined in S8; - Possibly optimization of the parameters of the chosen model(s); - a target defined in S8 and help in the diagnosis, prognosis and / or therapeutic monitoring of that, for the subject S xto be tested.

2. Method according to claim 1, characterized . in that the biological fluid is chosen from the group comprising - advantageously consisting of - follicular, bile, . and in that the diseases impacting the mineral profile of the biological fluid, include: . in the case of cerebrospinal fluid, diseases affecting the central nervous system, in particular neurological diseases, such as psychiatric diseases, neurodegenerative diseases of the Parkinson's or Alzheimer's type, . in the case of blood plasma, cancers, neurological diseases, diseases . complex operations, in particular cardiovascular, . or diseases causing fertility problems and requiring in particular procreation aids.

3. Method according to claim 1, characterized in that the BDR includes data relating to subjects S m=0 (d,p,t)healthy reference subjects or subjects (d,p,t) minus two biological fluids chosen from the group comprising -advantageously consisting of- follicular, bile.

4. a step -S0 ^ - construction of the reference database consisting of collecting data, in particular data (i) and / or (ii) and / or data D Sx and to store them.

5. -quantitative, inductively coupled plasma mass spectrometry, or ICP-MS (Inductively Coupled Plasma Mass Spectrometry) being preferred.

6. according to -S8- minus a biological fluid, for the subject Sx to be tested, and in that the BDR comprises data relating to a number N s of healthy subjects S m=0 (d,p,t) of reference and a number N m of sick subjects (d,p,t) of reference; with [N s / (N s + N m)] *100 greater than or equal to - in ascending order of preference - 20%, 30%, 40%.

7. characterized in that the 1 er protocol -S9.1- includes the following steps: -S9.1.1- Collection of data in the BDR; -S9.1.2- Normalization of data; -S9.1.3- Selection among data (i), and possibly data (ii), of differentiating data; -S9.1.4- Reduction of the dimensionality of the data in 2D or 3D, preferably by a -SNE (t- distributed Stochastic Neighbor Embedding); -S9.1.5- Visualization of targets on graphs; -S9.1.6- Positioning on the graphic visualization of the targets, the data D Sx on the subject(s) S x to be tested, to visualize the positioning of the Sx(s), relative to the targets.

8. - S9.1.3- consists of using: * [Variant V1 of 1 er protocol] of the “volcano plot” curves and / or * [Variant V2 of 1 erprotocol] a selection algorithm, preferably a recursive selection algorithm “recursive features elimination-RFE”, possibly associated with the following sub-steps: -S9.1.3.1- Definition of models; -S9.1.3.2- Data separation; -S9.1.3.3- Learning and evaluation loop; -S9.1.3.4- (AUC score); -S9.1.3.5- best AUC score.

9. what the 2 èmein S9.2 includes the following steps: -S9.2.1- Collection in the BDR of the data values; -S9.2.2- Normalization of the data values; -S9.2.3- by AI; -S9.2.4- Separation of the data; -S9.2.5- Learning and evaluation loop; -S9.2.6- Plotting the ROC curve for each model and (AUC score); -S9.2.7- on the best AUC score; -S9.2.8-; -S9.2.9- selected in -S9.2.7-.

10. - S9.2.8- consists of using “volcano plot” curves -S9.2.7-.

11. characterized in that: *a prognosis is established consisting of evaluating the risks of clinical complications linked to a renal problem, a pulmonary problem, an infection, and / or the risks of lethal complications, for a subject Sx to be tested, having undergone a surgical operation, blood transfusions and / or extracorporeal blood circulation; this prognosis constituting an aid to the medical decision of whether or not to maintain the subject Sx in intensive care; and / or * in one or more times, to the subject Sx to be tested, for one or more durations after; this therapeutic monitoring constituting an aid to the medical decision to modify or not the dose of medication to be administered to the subject Sx to be tested; and / or * clinical complications, e therapeutic constituting an aid to the medical decision to modify or not the doses * esse; this prognosis constituting an aid to the decision of

12. at least of the preceding claims, the device comprising * -S9.1-: -M9.1.1- Collection of data in the BDR -M9.1.2- Normalization of the data; -M9.1.3- Selection from the data (i), and possibly the data (ii), of differentiating data; -M9.1.4- Reduction of the dimensionality of the data in 2D or 3D, preferably by an A and / or by a t-SNE (t- distributed Stochastic Neighbor Embedding) method; -M9.1.5-Visualization of the targets on graphs; -M9.1.6-Positioning on the graphic visualization of the targets the data D Sxon the subject(s) S x to test, to visualize the positioning of the Sx(s), in relation to the targets ; * ème protocol -S9.2-: -M9.2.1- Collection of data values ​​in the BDR -M9.2.2- Normalization of data values ​​-M9.2.3- by AI -M9.2.4- Data separation -M9.2.5- Learning and evaluation loop -M9.2.6- Plotting the ROC curve -M9.2.7- -M9.2.8- -M9.2.9- -S9.2.7-.

13. System comprising the electronic device according to claim 12, at least one computer server hosting this electronic device, the BDR and at least one terminal allowing at least one user to communicate with the electronic device, in particular to transmit data D Sxrelating to a subject Sx to be tested and of the same type as the data (i) & (ii) defined in S0 in claims 1 to 11, and to receive in return information, relating to the subject Sx to be tested, diagnosis, prognosis and / or therapeutic monitoring of diseases impacting

14. Computer program comprising instructions for putting

15. Data medium on which a computer program according to claim 13 is recorded.

Images