Computer-implemented method for determining the states in vivo and in vitro by analyzing the blood parameters measured in a hematological analysis device

Abstract

The invention relates to a computer-implemented method for determining states in vivo and in vitro by analyzing blood parameters, comprising: a) obtaining blood parameters of a blood sample by means of a hematology analyzer, wherein the blood parameters encompass quantitative and qualitative measurement variables, wherein the measurement variables encompass properties of individual cells, and the individual cells encompass blood cells; b) creating at least one scatterplot having at least two axes, wherein each axis of the scatterplot encompasses a different measurement variable; c) determining at least one in-vivo and / or in-vitro and / or post-mortem state by means of at least one deep learning model and / or at least one machine learning model, wherein the input variable for the at least one deep learning model comprises at least one scatterplot, wherein the input variable for the at least one machine learning model comprises at least one 1D vector, and wherein the 1D vector is created by vectorizing the at least one scatterplot; and d) automatically generating a report which comprises at least one result regarding the determination of the at least one state.

Description

[0001]Computer-implemented method for determining in vivo and in vitro conditions by analyzing blood parameters measured in a hematology analyzer. Technical field: The present invention relates to the field of medical diagnostics, in particular to the analysis of blood samples and the determination of in vivo and in vitro conditions. The invention uses advanced artificial intelligence (AI) techniques, such as deep learning and machine learning, for the automated analysis of blood parameters measured in a hematology analyzer to perform precise and efficient diagnoses and condition determinations. Background of the invention: Hematology analyzers are widely used diagnostic instruments for examining blood samples in clinical laboratories. These analyzers measure various parameters of blood cells, such as the number, size, volume, and complexity of the cells. The analysisThese blood parameters allow medical professionals to obtain information about a patient's health status, possible diseases, or other relevant characteristics. However, traditional approaches to analyzing hematology parameters are often time-consuming and require manual interpretation by specialized personnel. State of the art: High-dimensional single-cell information is measured in every complete blood count, but clinical decisions are currently based only on a few derived statistics. The enormous potential of the complete set of blood cell measurements has been well appreciated, and previous efforts attempted to achieve early detection of infections by identifying immature granulocytes or prognosis for some malignancies by counting the number of WBCs with atypical features (Statland BE, Winkel P, Harris SC, Burdsall MJ, Saunders AM. Evaluation of biological sources of variation in leukocyte counts and otherhematologic quantities using very precise automated analyzers. At J Clin Pathol. 1978 Jan;69(1):48-54. doi: 10.1093 / ajcp / 69.1.48. PMID: 563672.). These efforts have had limited impact but indicate the potential for improved clinical decision support (Gijsberts CM, den Ruijter HM, de Kleijn DPV, Huisman A, Ten Berg MJ, van Wijk RHA, Asselbergs FW, Voskuil M, Pasterkamp G, van Solinge WW, Hoefer IE. Hematological Parameters Improve Prediction of Mortality and Secondary Adverse Events in Coronary Angiography Patients: A Longitudinal Cohort Study. Medicine (Baltimore). 2015 Nov;94(45):e1992. doi: 10.1097 / MD.0000000000001992. PMID: 26559287; PMCID: PMC4912281.). In this study [Campuzano-Zuluaga G, Alvarez-Sánchez G, Escobar-Gallo GE, Valencia-Zuluaga LM, Ríos-Orrego AM, Pabón-Vidal A, Miranda-Arboleda AF, Blair-Trujillo S, Campuzano-Maya G. Design of malaria diagnostic criteria for the Sysmex XE-2100 hematology analyzer. At J Trop Med Hyg. 2010Mar;82(3):402-11. doi: 10.4269 / ajtmh.2010.09-0464. PMID: 20207864; PMCID: PMC2829900.] The authors used scattergrams generated by the Sysmex XE-2100 hematology analyzer to differentiate blood samples from patients with malaria from those without. In the study, the authors analyzed the scattergrams and identified specific patterns that occurred in malaria-infected patients. Based on these patterns, they developed diagnostic criteria to detect malaria infections. By analyzing the scattergrams and applying the developed criteria, they were able to achieve high sensitivity and specificity in malaria diagnosis. The study [Chaudhury A, Noiret L, Higgins JM]. White blood cell population dynamics for risk stratification of acute coronary syndrome. Proc Natl Acad Sci US A. 2017 Nov 14;114(46):12344- 12349. doi: 10.1073 / pnas.1709228114. Epub 2017 Oct 27. PMID: 29087321; PMCID: PMC5699055.] examines the dynamicsof white blood cells in relation to acute coronary syndrome (ACS) and identifies specific clusters of lymphocytes, neutrophils, and monocytes using an Abbott Cell-Dyn-Sapphire hematology analyzer. These clusters are analyzed using scattergrams using the Fokker-Plank differential equation to risk stratify healthy patients and patients with ACS. The mathematical model achieves an accuracy of over 70% in identifying patients who initially had negative screening tests but were diagnosed with ACS within 48 hours. In contrast, the Pushkin and Shulkin patent RU 2733077 C1 describes a method for diagnosing ACS based on the measured properties of white blood cells. Scattergrams of cell size and cell complexity are created, which are also measured by an Abbott Cell-Dyn-Sapphire hematology analyzer.Clusters of lymphocytes, monocytes, and neutrophils are manually identified and grouped into 1D vectors. These are then reduced using principal component analysis and used for analysis by multilayer perceptrons (MLPs). The method is evaluated using a small database of 211 measurements, achieving a sensitivity of 0.97 and a specificity of 0.94 (AUC = 0.96). The study [Pushkin AS, Shulkin D, Borisova LV, Akhmedov TA, Rukavishnikova SA. [Algorithm to stratify the risk of myocardial infarction in patients with acute coronary syndrome at primary examination.]]. Klin Lab Diagn. 2020;65(6):111-222. Russian. doi: 10.18821 / 0869-2084-2020-65-6-111-222. PMID: 32459900.)] investigates the application of the method described in the aforementioned patent RU 2733077 C1 for the classification of myocardial infarctions and unstable angina pectoris in patients with acute coronary syndrome (ACS). The authors created aA small database of 307 anonymized measurements taken using an Abbott Cell-Dyn Sapphire hematology analyzer. Of these, 214 measurements were used for training and 93 for method evaluation. The results showed a sensitivity of 0.77 and a specificity of 0.80 (AUC = 0.77) for classifying myocardial infarctions and unstable angina pectoris in patients with ACS. The characterization of acute leukemias by various hematology analyzers has been previously documented. Krause JR et al. evaluated the use of the Technicon H-1 (Technicon Instruments Corporation, Tarrytown, NY, USA) for characterizing acute leukemias. Based on myeloperoxidase activity and nuclear characteristics of the cells, they were able to differentiate between acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL). In the case of AML, they pointed out that AML of the French-American-British (FAB) type - M3,M4 and M5 - exhibit characteristic cytograms. Chronic myeloid leukemia (CML) also exhibited a characteristic pattern [Krause JR, Costello RT, Krause J, Penchansky L. Use of the Technicon H-1 in the characterization of leukemias. Arch Pathol Lab Med 1988;112:889-4.]. Similarly, Kawarabayashi et al. investigated the utility of the Technicon H-1 (Technicon Instruments Corporation, Tarrytown, NY, USA) for detecting blast cells. [Kawarabayashi K, Tsuda I, Tatsumi N, Okuda K. Leukemic blasts detected by the Technicon H-1® blood cell counter. Am J Clin Pathol 1987;88:624-27.]. Hoyer et al. investigated the ability of the Coulter STKS hematology analyzer to differentiate acute leukemias. They concluded that using a set of suspect or definitive flags as screening criteria for microscopic review would be the best approach to correctly identify leukemias. They further concluded thatScatterplot patterns are not useful for classifying acute leukemias. [Hoyer JD, Fisher CP, Soppa VM, Lantis KL, Hanson CA. Detection and classification of acute leukemia by the Coulter STKS Hematology Analyzer. Am J Clin Pathol 1996;106:352-8.] Bruno et al. 1994 and Pettit et al. 1995 also investigated the scatterplot patterns of acute leukemias using the Coulter STKS hematology analyzer [Bruno A, Del Poeta G, Venditti A, Stasi R, Adorno G, Aronica G, et al. Diagnosis of acute myeloid leukemia and system Coulter VCS. Haematologica 1994;79:420-8.], [Pettitt AR, Grace P, Chu P. An assessment of the Coulter VCS automated differential counter scatterplots in the recognition of specific acute leukemia variants. Clin Lab Haematol 1995;17:125-9.]. Virk et al. investigated the utility of cell population data (VCS parameters) from the Coulter LH 780 automated hematology analyzer as a rapid screening tool for AML in resource-limited laboratories.concluded that cell population data, along with scattergrams, can provide a cost-effective and rapid initial diagnosis of acute leukemias. These parameters can then be used to differentiate malignant hematological diseases from non-malignant ones [Virk H, Varma N, Naseem S, Bihana I, Sukhachev D. Utility of cell population data (VCS parameters) as a rapid screening tool for Acute Myeloid Leukemia (AML) in resource-constrained laboratories. J Clin Lab Anal 2019;33:e22679.]. In the study [Aparna N et al, Scattergram patterns of hematological malignancies on sysmex XN-series analyzer, Journal of Applied Hematology, 2021, 12, 2, 83-39 ], the scattergram patterns of various primary hematological malignancies in relation to the use of the Sysmex XN hematology analyzer are examined. The authors conducted a retrospective study in which they analyzed the details of 291 newly diagnosed cases of hematologicalmalignancies and 48 cases of leukemoid reactions. The aim of the study was to determine whether specific hematological malignancies generate specific scattergram patterns. The authors found that various scattergram patterns were observed, and these patterns can be used to differentiate between various reactive and neoplastic conditions. Pattern analysis confirms that all cases examined exhibit individual patterns. These patterns can be used to target cases for further molecular and cytogenetic analyses. The above-mentioned studies and patents mainly focus on manual, statistical and mathematical methods and machine learning, such as principal component analysis and multi-layer perceptrons (MLPs). Deep learning models, especially CNNs, have demonstrated excellent performance in many fields in recent years, including medical imaging and diagnosis.Results achieved [Litjens G, Kooi T, Bejnordi BE, Setio AAA, Ciompi F, Ghafoorian M, van der Laak JAWM, van Ginneken B, Sánchez CI. A survey on deep learning in medical image analysis. Med Image Anal. 2017 Dec;42:60-88. doi: 10.1016 / j.media.2017.07.005. Epub 2017 Jul 26. PMID: 28778026.]. US 2018 / 0247715 A1 describes a method for diagnosing and characterizing cancer using artificial neural networks (ANNs) by analyzing white blood cells using a flow cytometer. The publications discussed above primarily focus on methods for diagnosing diseases in humans. Object of the invention: The object of the invention is to provide a method for determining conditions and / or automatically producing a report on conditions based on the analysis of blood parameters, which is characterized by increased accuracy, wider applicability and less workload compared to the prior art.Description of the invention: The object is achieved by a computer-implemented method for determining conditions in vitro and in vivo by analyzing blood parameters measured in a hematology analyzer, the method comprising: a) Obtaining blood parameters of a blood sample by means of a hematology analyzer, the blood parameters comprising quantitative and qualitative measured variables, the measured variables comprising the properties of individual cells, the individual cells comprising blood cells, the blood cells comprising white blood cells, red blood cells, and platelets, the white blood cells comprising monocytes, lymphocytes, basophils, eosinophils, and neutrophils; b) Creating at least one scattergram with at least two axes, each axis of the scattergram comprising a different measured variable; c) Determining at least one in-vivo and / or in-vitro and / or post-mortem condition by at least one deepLearning model and / or a machine learning model, wherein the input variable for the at least one deep learning model comprises at least one scattergram, wherein the input variable for the at least one machine learning model comprises at least one 1D vector, wherein the creation of the 1D vector is carried out by vectorization of the at least one scattergram; and d) Automatically generating a report comprising at least one result on the determination of the at least one condition. The object is also achieved by a computer-implemented method for automatically generating a report on the determination of conditions in vitro and in vivo by analyzing blood parameters measured in a hematology analyzer, wherein the method comprises: a) Obtaining blood parameters of a blood sample by means of a hematology analyzer, wherein the blood parameters comprise quantitative and qualitative measured variables, wherein the measured variables comprise the propertiesof single cells, wherein the single cells comprise blood cells, wherein the blood cells comprise white blood cells, red blood cells and platelets, wherein the white blood cells comprise monocytes, lymphocytes, basophils, eosinophils and neutrophils; b) Creating at least one scattergram with at least two axes, wherein each axis of the scattergram comprises a different measurement variable; c) Determining at least one in-vive and / or in-vitro and / or post-mortem state by at least one deep learning model and / or a machine learning model, wherein the input variable for the at least one deep learning model comprises at least one scattergram, wherein the input variable for the at least one machine learning model comprises at least one 1D vector, wherein the creation of the 1D vector is carried out by vectorizing the at least one scattergram; d) Automatically generating a report containing at least one result about the determination of theat least one condition; and e) transmitting the report by means of a data carrier signal f) receiving the transmitted report The present invention relates to a computer-implemented method for determining conditions in vivo and in vitro by analyzing blood parameters measured in a hematology analyzer and for automatically generating a report on the determined conditions. The method uses AI techniques, in particular deep learning and machine learning models, for the automated analysis and interpretation of the measured blood parameters and for determining the conditions. The approaches described in the prior art require numerous manual processing steps, such as cluster analysis to identify the subpopulations of white blood cells in scattergrams. In the invention, cluster analysis is not necessary to determine the conditions, because all components of the blood in the blood sample are analyzed anyway.taken into account, especially red blood cells and platelets. Deep learning models are able to automatically recognize the crucial patterns and structures for condition determination during the training phase. Furthermore, conventional cluster analyses are often not feasible with multidimensional scattergrams. By using deep learning models, manual processing steps can be reduced or even eliminated, enabling more efficient and accurate condition determination. These models are able to automatically capture and learn complex patterns and relationships within the data, accelerating the analysis process and improving the accuracy of the results. The invention enables improved analysis and interpretation of hematology parameters, leading to increased accuracy and efficiency in the determination of conditions in vivo and in vitro. Furthermore, the invention can contribute toto reduce the workload of medical professionals, as automated analysis and reporting minimizes manual interpretation. The application of AI techniques also enables continuous improvement of models by adding new data and experience, thereby optimizing diagnostic performance over time. The method according to the invention relates to the examination of blood samples from humans and animals. The invention can use various deep learning and machine learning models, including convolutional neural networks, recurrent neural networks, support vector machines, decision trees, and random forests, to analyze different aspects of the measured blood parameters and combine the results. These models can be combined in an ensemble approach to improve the predictive accuracy and robustness of the AI ​​by combining different models or model instances.to make a consolidated prediction of the condition. A further advantage of the invention is that it is capable of determining a variety of conditions that can occur in vivo, ex vivo, in vitro, and / or post-mortem. This includes conditions and processes both inside and outside a living organism, including changes in blood cell morphology, cell composition, cell function, or other characteristics of individual cells as a result of storage, handling, and / or analysis. The invention can be used for various applications, such as clinical diagnostics, research, forensic analysis, and veterinary examination. By integrating the invention into existing hematology analyzers or laboratory systems, the diagnostic capabilities of these systems can be expanded and the quality of patient care can be improved. The invention can contribute toto reduce the time required for the analysis of blood samples and the preparation of reports, thereby increasing the efficiency of laboratories and lowering the cost of patient care. Overall, the present invention offers an innovative solution for improving hematology analysis and condition assessment in vivo and in vitro through the use of artificial intelligence. The invention enables automated, precise, and efficient analysis of blood parameters from hematology analyzers and can help improve the quality of patient care and increase the efficiency of laboratories. A blood sample is taken from the subject to be examined. For example, the first venous whole blood is drawn from a cubic vein, e.g. using a 4 ml vacuum system for blood collection, into a e.g. Vacutest tube (KIMA, Italy) and applied to the inner surface of the e.g. 7.2 mg K3EDTA tube walls. Other variants are conceivable.This sample is then used to determine diseases or other conditions by the method according to the invention. Sampling is not part of the method. The tube could be stirred after blood collection by turning it upside down and rotating it horizontally and vertically for 30 seconds. The clinical blood test is then carried out in open mode on an automated hematology analyzer, e.g., CELL-DYN Sapphire (Abbott Laboratories, USA). In this process, the individual cells of the complete blood count are measured in a high-dimensional manner, with the measurements including the properties of the individual cells, as shown by way of example in Fig. 1. The measurements are copied from the analyzer, e.g., as FCS files or in another format, and transferred to an accessible PC or mobile computing device or a cloud for automated processing. These measurements include blood parameters as properties of leukocytes, with leukocytes representing theNeutrophils, eosinophils, basophils, lymphocytes, and monocytes, with characteristics including size, granularity, lobularity, and complexity. a) Complete blood count Blood cells in the circulation of a human or animal continuously circulate through almost all tissues in vivo at a high rate, and their collective level of maturity, activation, proliferation, and senescence reflect the current pathophysiological or health state: healthy quiescence, acute response to pathology, chronic compensation for disease, and ultimately decompensation. Complete blood counts involve the measurement of single-cell characteristics for tens of thousands of blood cells and provide an overview of these conditions. The complete blood count includes the measurement of white blood cells, red blood cells, and platelets. A complete blood count is a common blood test that is often part of a routine physical. A complete blood count can help determine a variety ofto detect conditions such as infections, anemia, immune system disorders, and blood cancers. [MedlinePlus Medical Encyclopedia. (2021). Complete blood count (CBC). US National Library of Medicine. Retrieved from https: / / medlineplus.gov / lab-tests / complete-blood-count-cbc / ]. b) Measuring blood parameters in the hematology analyzer A complete blood count is usually performed using an automated laboratory device called a hematology analyzer. It uses various technologies to measure the various blood cells and blood parameters included in a complete blood count. One of the most common technologies used in hematology analyzers is impedance sensing. This involves passing the blood sample into a tiny chamber filled with a conductive fluid. Electrical pulses are then passed through the fluid, measuring the resistance created by the various blood cells. Based onThese measurements allow the device to determine the number and size of red blood cells, white blood cells, and platelets. Another method used in hematology analyzers is laser scattered light analysis. In this method, the blood sample is passed through a thin beam of laser light, which is reflected or scattered by the various blood cells. The reflected light is then collected by photodetectors that can detect the size, shape, and complexity of the various cells. Most modern hematology analyzers combine these two technologies to achieve greater accuracy and reliability. The blood sample is passed into several channels, each dedicated to a specific analysis. The device can then automatically detect and quantify the various cell types and display the results in a report. The measurement channels of the hematology analyzer only display the raw measurement data.generated. This data contains information such as the size, shape, density, or color of the blood cells, which is captured by the specific measurement methods in each measurement channel. The raw measurement data from the measurement channels is then processed by the analyzer's software and usually converted into a series of results and blood parameters. These results and blood parameters include the total white and red blood cell count, the hematocrit value, the hemoglobin concentration, and other relevant information about the blood cells. The hematology analyzer software usually performs complex algorithms and statistical methods to obtain more accurate results from the raw measurement data. c) Scatterplot Analysis One of the most important tools for displaying the blood cell properties measured by a hematology analyzer is a scatterplot. Scatterplots obtained from hematology analyzers can alsofor conditions other than diseases, such as blood age after collection, as well as for other living beings such as animals. A scattergram is a graphical representation of data points arranged on a two-dimensional plane. Each point in the graph represents a single cell, and the two axes represent different blood parameters measured by the analyzer. Generally, most hematology analyzers will produce two basic scattergrams, one for red blood cells (RBC) and one for white blood cells (WBC). The RBC scattergram shows the size and distribution of red blood cells, while the WBC scattergram shows the size and distribution of the different types of white blood cells. Additional scattergrams may be available depending on the model of hematology analyzer and the required blood parameters. Using these scattergrams, doctors and medical professionals can describe various conditionsmanually identify or suspect that exhibit characteristic patterns. Most modern hematology analyzers have the ability to store the acquired measurement data in a digital form. This data can then be exported and used for further analysis and visualization, including the creation of scattergrams. d) Deep Learning Analysis for Determining Conditions in Vitro and in Vivo This description describes a computer-implemented method for determining conditions in vivo and in vitro by analyzing blood parameters measured with a hematology analyzer, using artificial intelligence and deep learning models. In an exemplary use case, it is demonstrated for the first time that the proposed method is capable of successfully classifying blood age in vitro after collection. Description of Embodiments: Embodiments of theThe methods according to the invention are explained below. For example, the blood age in a blood sample after blood collection is to be determined in vitro. The age of the blood in vitro after blood collection may be of interest, as some studies have shown that it affects the quality and effectiveness of the transfused red blood cells. It is assumed that some properties of red blood cells change over time, including viscosity, the ability to transport oxygen, and the expression of antigens on the cell surface. One possible application of knowledge about the age of blood could be to reduce transfusion-related morbidity and mortality by selecting the most appropriately aged red blood cells. For example, the use of fresh blood could be advantageous in certain patients, such as trauma patients or patients with severe bleeding, toTo maximize the effectiveness of transfusion and reduce complications. However, determining the age of blood after blood collection is not easy, and there is no standardized procedure for doing so. There are various approaches and techniques for estimating blood age, including measuring the expression of certain proteins on the surface of red blood cells and analyzing changes in the red blood cell membrane over time. The blood parameters of a blood sample are obtained in a modern hematology analyzer. There are several manufacturers of hematology analyzers on the market. Some of the best-known and most widely used brands include: - Sysmex: Sysmex is one of the largest suppliers of hematology analyzers worldwide. The company offers a wide range of devices suitable for various applications, from small clinics to large laboratories. - Beckman Coulter: Beckman Coulter is aAnother leading provider of hematology analyzers, it offers a wide range of devices suitable for various applications, from routine analysis to specialized research. - Abbott Laboratories: Abbott Laboratories is a large company that manufactures various medical devices and diagnostics, including hematology analyzers. - Siemens Healthcare Diagnostics: Siemens is a well-known medical device manufacturer and also offers hematology analyzers. - Roche Diagnostics: Roche is a global diagnostics company and also offers a range of hematology analyzers. - Horiba Medical: Horiba is a global medical device manufacturer that also offers a range of hematology analyzers. - Mindray: Mindray is a Chinese company that offers a wide range of medical devices and solutions, including hematology analyzers. -Nihon Kohden: Nihon Kohden is a Japanese manufacturer of medical devices, including hematology analyzers. - Boule Medical: Boule Medical is a Swedish company specializing in the development of hematology analyzers. - Diatron: Diatron is a Hungarian manufacturer of medical devices, including hematology analyzers. - HemoCue: HemoCue is a Swedish company specializing in the development of point-of-care tests and analyzers, including hematology analyzers. - Shenzhen Mindray Bio-Medical Electronics Co.: Shenzhen Mindray is a Chinese company that offers a wide range of medical devices and solutions, including hematology analyzers. - Human Diagnostics: Human Diagnostics is a German company specializing in the development and manufacture of in vitro diagnostics and laboratory equipment, including hematology analyzers. -Erba Mannheim: Erba Mannheim is a global company specializing in the development of diagnostics and medical devices for various applications, including hematology analyzers. - Heska: Heska is a US company specializing in the development of diagnostic and treatment solutions for companion animals, including hematology analyzers. In a blood sample examined using a hematology analyzer, various types of blood cells can be detected and taken into account by the method according to the invention. The blood cells and their functions include: - Erythrocytes (red blood cells): These cells transport oxygen from the lungs to the tissues and carbon dioxide from the tissues to the lungs. The condition of the erythrocytes can be influenced by various factors such as the oxygen content of the blood, the pH value, and the amount of carbon dioxide. - Leukocytes (white blood cells)Blood cells): These cells play an important role in fighting infections and foreign substances in the body. The condition of leukocytes can be influenced by the degree of inflammation, the type of pathogen, and other factors. - Lymphocytes: This type of white blood cell plays an important role in the immune system and can exist in various states depending on the circumstances. For example, they can be activated to fight infections or they can remain inactive when there is no threat to the body. - Monocytes: These cells are also white blood cells and play a role in fighting infections and foreign substances. They can transform into various types of tissue macrophages that phagocytose foreign substances and protect the body from damage. - Basophils: This type of white blood cell is involved in allergic reactions and can exist in variousconditions, depending on the type of allergic reaction and other factors. - Neutrophils: These cells are the most common type of white blood cell and play an important role in fighting infections. They can be in various states, depending on the type of infection and other factors. - Eosinophils: This type of white blood cell is involved in fighting parasites and allergies and can be in various states, depending on the circumstances. - Platelets (blood platelets): These cells play an important role in blood clotting. The condition of platelets can be affected by various factors such as inflammation, infections, and other disorders of the blood system. Other cell types and / or particles may also be present in the blood sample, the properties of which can also be detected by hematology analyzers. These cell types and / or particles can also be detected by theinventive methods. These cells and / or particles may be of clinical interest and include, among others: - Blasts: Immature blood cells that normally occur in the bone marrow and enter the blood in small quantities. An increased number of blasts in the blood may indicate a blood disease such as leukemia. - Cell fragments: Broken cells that arise due to cell damage or degradation. For example, schistocytes (broken erythrocytes) may occur in certain diseases such as hemolytic uremic syndrome or thrombotic thrombocytopenic purpura. - Circulating tumor cells (CTCs): Cancer cells that detach from the primary tumor and enter the bloodstream. CTCs can occur in various cancers and are a potential marker for tumor metastasis. - Endothelial cells: Cells that form the inner lining of blood vessels. An increased number of endothelial cells in the blood may indicateindicate inflammation or injury to the blood vessels. - Microparticles: Small cell fragments secreted by various cell types that circulate in the blood. They can occur in cases of inflammation, clotting disorders, or other diseases. In some cases, hematology analyzers can also detect parasites in the blood. Some blood parasites, such as those that cause malaria (Plasmodium spp.), can be found within red blood cells (erythrocytes). In cases of severe infestation, these infected cells can be detected by the analyzers and possibly identified as abnormal cells. The measured properties or parameters can vary depending on the analyzer type and technology used and include the following: - Number of cells - Size: The size of the cells can help distinguish different cell types, as they typically have different sizes. For example, erythrocytes are smaller thanLeukocytes. - Shape: The shape of the cells can also help with identification. Erythrocytes are usually disc-shaped, while leukocytes and platelets can have different shapes. - Volume: The volume of cells can be related to the diameter of the cell and varies between different cell types. - Granularity or complexity: The internal structure of cells, also known as granularity, can be used to identify cell types, especially leukocytes, which can be divided into different subgroups based on their granularity (e.g., granulocytes and lymphocytes). - Electrical conductivity: The electrical conductivity of cells can also be used to differentiate cell types, as they have different conductivities due to their different membrane structures and cellular contents. - Light scattering: Some hematology analyzers use light scattering.caused by the cells to evaluate size, shape, and granularity. Forward scatter is proportional to the size of the cell, while lateral scatter is related to the granularity of the cell. - Mean Corpuscular Volume (MCV) - Mean Corpuscular Hemoglobin Content (MCH) - Mean Corpuscular Hemoglobin Concentration (MCHC) - Red Blood Cell Distribution Width (RDW) - Mean Platelet Volume (MPV) - Platelet Distribution Width The measurements of the cell properties are transferred from the hematology analyzer to an accessible PC, mobile computer, mobile device, or cloud for machine processing and subsequent AI-based analysis. The measurement results from hematology analyzers can be transferred in various formats, depending on the interfaces supported by the analyzer and the formats accepted by the target information system. Formats for transferring measurement results fromHematology analyzers include: - HL7 (Health Level Seven): This is a standard format for the exchange of clinical and administrative data between different medical information systems. Hematology analyzers can transmit measurement results in HL7 format to an information system via an interface such as a TCP / IP connection. - ASTM (American Society for Testing and Materials): ASTM is another format for the exchange of data between medical devices and information systems. Hematology analyzers can transmit measurement results in ASTM format to an information system via an RS-232 interface. - Text files: The measurement results can be saved and transmitted as text files, which can be in various formats such as CSV, TXT, or XML. These files can then be further processed in other programs such as spreadsheets or databases. - DICOM: DICOM is a standard format forTransfer of medical images and other data between different systems and devices. Some hematology analyzers can create DICOM files, which can then be used in image processing programs or electronic medical records. - FCS: FCS stands for "Flow Cytometry Standard" and is a file format used in flow cytometry to store data from individual cells or particles. The FCS format is an open standard developed by the International Society for the Advancement of Cytometry (ISAC) to ensure compatibility between different flow cytometry devices and software platforms. Machining consists of the automatic creation of at least one scatter plot. Each point in the plot represents a single cell, and the two axes represent different properties measured by the analyzer. In a scatter plot based on data froma hematology analyzer, two properties of the measured cells are often plotted along the x- and y-axes. Typical properties depicted in scattergrams include: - Size (cell diameter or cell volume): The size of the cells can be plotted on the x- or y-axis to distinguish different cell types, as they have different sizes. - Granularity or internal complexity: The granularity of the cells can be plotted on the x- or y-axis and is often used to identify subsets of leukocytes, such as lymphocytes, monocytes, and granulocytes. - Light scatter: Forward scatter and side scatter can be plotted in scattergrams. Forward scatter correlates with the size of the cells, while side scatter provides information about the granularity or internal complexity of the cells. In addition to at least one scattergram in simple form,can also be created and used as at least one scatterplot: - a 3D scatterplot: In a 3D scatterplot, three axes (x, y, and z) are used to represent three properties simultaneously. - a scatterplot matrix: This method displays multiple 2D scatterplots in a matrix, with each plot representing a pair of properties. - parallel coordinates: In parallel coordinates, multiple vertical axes are arranged parallel to each other, with each axis representing a dimension or property. The data points are represented by lines connecting the corresponding values ​​on each axis. - a histogram: A histogram shows the distribution of measured values ​​of a particular property or characteristic. For example, when analyzing blood cells, you could create histograms for size, granularity, or fluorescence intensity. - a density plot: This is equivalent to a scatterplot using an additionalDimension, such as different colors for the points of the scatter plot; this way, even overlapping points can be differentiated and taken into account. At least one scatter plot is then used as an input for at least one deep learning model based on artificial neural networks. This means that it consists of many layers of neurons and can learn a hierarchical representation of data. It is capable of extracting complex features from large amounts of data and making precise predictions of the state based on these features. The at least one deep learning model can include at least one deep learning model from the following group: - Convolutional Neural Networks (CNNs): CNNs are a powerful deep learning model specifically developed for image processing. They can analyze both 2D and 3D images and can be used for a wide variety of applications.such as image recognition, object recognition, face recognition, and medical imaging. - Generative Adversarial Networks (GANs): GANs are a pair of artificial neural networks that work together to generate new images similar to the images in the training dataset. One network, called a generator, creates new images, while another network, called a discriminator, attempts to distinguish between real and generated images. GANs are often used for generating artificial images, image restoration, and reconstruction. - Recurrent Neural Networks (RNNs): RNNs are another deep learning model that can be used for processing multidimensional images. - Long Short-Term Memory Networks - Transformer Networks - 3D Convolutional Neural Networks (3D CNNs): 3D CNNs are an extension of CNNs used for processing 3D images- 4D Convolutional Neural Networks The scattergram can optionally be edited before deep learning analysis (data pre-processing). This editing includes: - Resizing: The images should be resized to a suitable size for the model, such as 224x224 for many CNN models. - Normalization: The pixel values ​​of the images should be normalized by scaling them to a range between 0 and 1 to ensure they are comparable for the model. - Standardization - Noise reduction: It may be necessary to remove noise from the images, especially if the images are of poor quality or heavily compressed. - Test Time Augmentation (TTA) - Clustering - Contrast adjustment - Filtering - Cropping: It may be necessary to remove unwanted areas of the image to ensure that the model only receives relevant features. These editing operations can be performed individually or in batches.combination can be applied. In parallel, the scattergram can be converted into a one-dimensional vector (1D vector) to use it as an input for at least one machine learning model that can only process one-dimensional data structures. The method includes a vectorization step in which the data of the scattergram is converted into a one-dimensional vector. Vectorization is done by summarizing the data in a specific order so that the resulting vector is a unique representation of the scattergram. Vectorization could include the following: - Flattening: This technique refers to the simple conversion of a multi-dimensional matrix into a one-dimensional vector by concatenating the individual rows or columns of the matrix. In image processing, for example, this method can be applied by simplyarranged sequentially in a one-dimensional vector. - Reshaping: Reshaping refers to changing the shape of a multidimensional matrix without changing the individual elements of the matrix. In image processing, for example, reshaping can be used to convert a multidimensional image into a different format, such as a larger or smaller two-dimensional matrix or a tensor. Reshaping can also be used in other areas of data analysis and machine learning to bring multidimensional data into a different structure that is better suited to specific algorithms. - Zigzag scanning: This technique is commonly used in JPEG image compression. It involves scanning the pixel values ​​of a matrix in a "zigzag" pattern, so that the pixel values ​​of the resulting one-dimensional vector are not arranged in the order of the original matrix. - PCA-based techniques: AnotherA method for vectorizing multidimensional data is the use of Principal Component Analysis (PCA) or other linear transformations. This involves transforming the original matrix into a new basis that contains the most important features of the original dataset. The data can then be represented as linear combinations of the new basis vectors, allowing for a more efficient and compact representation. Suitable machine learning models that can process one-dimensional data structures include: - K-Nearest Neighbors (KNN): KNN models are suitable for processing one-dimensional data structures where each point is represented by a set of numerical or categorical features. The model finds the K nearest neighbors to a given point and uses these to make predictions about the class or value of the point. - Support Vector Machines (SVM): SVM models are suitable forsuitable for processing one-dimensional data structures where each feature is numeric. SVMs are particularly useful when the data consists of two classes, as they seek a linear separation between the classes and make a prediction based on this separation. - Decision Trees: Decision trees are also suitable for processing one-dimensional data structures where each feature is numeric or categorical. The model recursively divides the feature space into partitions by making decisions about the values ​​of the features. - Multi-Layer Perceptron (MLP): MLPs are neural network models developed primarily for processing one-dimensional data structures. Each layer of the MLP consists of a series of neurons, each equipped with a set of weights and an activation function to make predictions. MLPs can be used for classification andRegression problems can be used. - Adaboost: Adaboost is a machine learning algorithm suitable not only for one-dimensional data structures, but also for multidimensional data structures. Adaboost is an ensemble model consisting of a series of weak learners trained sequentially. Each weak learner is trained on a portion of the data, and then the data is weighted to focus on the errors of the previous learner. By combining multiple weak learners, Adaboost can achieve high accuracy in predicting classes or values ​​in complex multidimensional data structures. - Gradient Boosting: The gradient boosting method is an ensemble model that uses a series of weak learning algorithms to create a powerful model. By combining multiple weak learning algorithms, the gradient boosting method can generally achieve higher accuracy thanother machine learning models such as decision trees, Naive Bayes, or Support Vector Machines. - Naive Bayes: Naive Bayes is a probabilistic model suitable for processing one-dimensional data structures where each feature is independent of the others. It works by calculating the probability that an instance belongs to a particular class based on the distribution of features in that class. - One-Class Support Vector Machines (OCSVM): One-Class Support Vector Machines (OCSVM) are a special form of Support Vector Machines (SVM) that can be used for anomaly detection in data. Unlike conventional SVMs, which are used for classifying data into two or more classes, OCSVMs are used to detect data in a single class. The goal of an OCSVM model is to define a hyperbolic separation space around the normal data points. EachA point that falls within the separation space is considered normal, while points that lie outside the separation space are considered anomalies. OCSVM models are typically trained with a training dataset that contains only positive examples (normal data points). The model then learns to define a cutoff surface that corresponds to normal data points while separating anomalies from normal data points. The model can then be used to classify new data sets by checking whether they lie within the defined separation space or not. - Isolation Forests: Isolation Forests are a decision tree-based model that can be used for anomaly detection in large and complex data sets. The model separates the normal data points from the anomalies by creating a tree of decision rules. Anomalies can be identified by a shallow depth in the tree, as they are faster from thedeviate from normal data density than normal data points. - Local Outlier Factor (LOF): LOF is a clustering-based model used for detecting anomalies in data-dependent spaces. The model calculates the local densities of the data points and identifies anomalies that lie in areas of low local density. - Support Vector Data Description (SVDD): SVDD is an SVM-based model used for detecting anomalies in data. Unlike OCSVMs, SVDD attempts to define a hyperbolic separation space around the normal data points, rather than separating them from anomalies. Anomalies are then considered data points that lie outside the defined separation space. The one-dimensional vector can optionally be processed before machine learning analysis (data pre-processing), whereby the processing can include at least one of the following operations: - Noise reduction - Feature extraction: Feature extractionFeature selection is the process of extracting relevant features from the 1D vector that are important for the machine learning model. This can be done using statistical methods such as calculating means, variances, or correlations, or more advanced methods such as wavelet transforms or Mel Frequency Cepstral Coefficients (MFCCs) for audio data. - Preprocessing: Preprocessing refers to the process of normalizing, scaling, or otherwise preparing the 1D vector to make it suitable for the machine learning model. This can be done using standardization or min-max scaling to bring the data to a common scale, or using normalization to bring the data to a specific distribution. - Data / Dimensionality Reduction: Data reduction refers to the process of reducing the dimensionality of the 1D vector to reduce processing complexity orTo improve the accuracy of the machine learning model. This can be done by Principal Component Analysis (PCA) or other methods of dimensionality reduction. These processing operations (data pre-processing) can be applied individually or in combination. In an advantageous embodiment, the dimensionality reduction and / or the feature selection of the at least one 1D vector comprises at least one processing method from a group of processing methods, the group of processing methods comprising: principal component analysis, T-distributed stochastic neighbor embedding, linear discriminant analysis, truncated singular value decomposition, uniform manifold approximation and projection, independent component analysis, sparse representation, partial least squares regression, and kernel principal component analysis. For the prediction of blood age, at least one scattergram and / or at least one one-dimensional vector is used. The scattergram isanalyzed by at least one deep learning model, while the one-dimensional vector is processed by at least one machine learning model. This method enables the state to be predicted precisely and effectively in vitro. If at least two deep learning models and / or at least two machine learning models and / or deep learning models are used in combination with a machine learning model, this is referred to as the prediction of the state by an ensemble. The idea behind the ensemble is that different models can have different weaknesses and strengths, and that the combination of several models can compensate for the weaknesses and enhance the strengths. A prediction by an ensemble is made by combining the predictions of several individual models. The ensemble technique can be used to improve the prediction accuracy and robustness of artificial intelligence by using different models orModel instances are combined to make a consolidated prediction of the state. The exact method of prediction depends on the specific ensemble technique used. The techniques could include one or more of the following: - Bagging: In the bagging technique, multiple copies of the same model are created with different training datasets. When the ensemble is to make a prediction, the predictions of each individual model are combined by averaging them. The idea is that combining multiple copies of the same model with different training datasets improves prediction performance by reducing variability. - Boosting: In boosting, multiple weak models are trained sequentially. Each model attempts to correct the errors of the previous model by focusing on the incorrectly predicted examples. When the ensemble is to make a predictionThe predictions of each individual model are combined by weighting them. The idea is that combining several weak models leads to a strong model by correcting the errors and amplifying the strengths. - Stacking: In stacking, several different models are trained and the output of each model is used as input for a higher-level model that makes a final prediction. When the ensemble is to make a prediction, the predictions of each individual model are combined by using them as input for the higher-level model. The idea is that combining several different models that focus on different features leads to better prediction performance. - Hard voting: In this method, the predictions of several individual models are combined and the final prediction is decided by a majority vote. If the ensemble makes a predictionTo determine the class to be predicted, the class predicted by the majority of classifiers is selected. Hard voting is a simple and effective method to improve the prediction performance of classifiers by reducing variance and mitigating overfitting. It can also be used in combination with other ensemble techniques such as bagging or boosting. - Soft voting: In soft voting, the prediction probabilities of the individual models are aggregated and weighted to generate a final prediction probability for each class. The weighting can be adjusted manually or automatically based on the prediction performance of each classifier. The final prediction is then made based on the class with the highest prediction probability. The idea behind soft voting is that models with higher prediction accuracy should be given more weight than those with lower prediction accuracy.Soft voting is particularly useful when the models in an ensemble have different strengths and weaknesses or when the decision boundary between the classes is fuzzy. - Random Subspace - Mixture of Experts - Bayesian Model Averaging The prediction of blood age comprises the class designation and / or the prediction probability. In an advantageous embodiment, in vitro and / or ex vivo conditions include conditions based on processes outside a living organism, including changes in blood cell morphology, cell composition, cell function, or other characteristics of the individual cells as a result of storage, handling, and / or analysis. In an advantageous embodiment, in vivo conditions are based on processes within a living organism, including physiological and pathological conditions such as diseases, biological age, pregnancy, drug exposure, state of health,Nutritional deficiencies, hereditary disorders, dehydration, blood clotting disorders, infections, and / or anemia. In an advantageous embodiment, post-mortem conditions are based on processes of a dead organism, including changes in the morphology, cell composition, cell function, or other characteristics of individual cells as a result of diseases, the presence of medications, state of health before death, drugs, poisons, and / or toxic substances, as well as changes caused by the decay and autolysis of cells and tissues after death. In an advantageous embodiment, the creation of at least one result report is carried out by a computer device, wherein the result report comprises a graphic and / or an information text and / or a probability and / or a score and / or a class for at least one condition, wherein the presentation of the result report includes the presentation on the computer device, for example a PC.or a portable computer and / or a mobile terminal and / or laboratory device. In an advantageous embodiment, all described models are trained and / or validated on the basis of a pre-configured database, wherein the database comprises measured blood parameters and / or scattergrams, wherein the database can be expanded with new measured blood parameters and / or scattergrams in order to continuously improve the performance and accuracy of the models, wherein the database can include information about known conditions, diseases or other relevant information that contributes to a better interpretation and analysis of the measured blood parameters and / or scattergrams. The training methods for the said artificial intelligence models can comprise both supervised and unsupervised learning, wherein supervised learning focuses on the use of annotated data in the database to identify patterns andrelationships, while unsupervised learning enables the recognition of patterns and relationships in the data without prior annotation in order to identify novel findings and possibly previously unknown conditions or diseases. The aforementioned artificial intelligence models can be trained using ensemble learning methods, in which several models or algorithms are combined to increase the performance and accuracy of the predictions and to compensate for the weaknesses of individual models or algorithms. In an advantageous embodiment, the artificial intelligence is able to continuously optimize and adapt the method independently based on new data and findings in order to improve the prediction accuracy and robustness by learning from its own feedback and errors. In an advantageous embodiment, a user interface is additionally provided that enables the input ofParameters and / or scatter plots, as well as the display of results and reports for healthcare professionals and / or patients. In an advantageous embodiment, the method enables the integration and use of external data sources, including clinical data, demographic information, medical history, and / or genetic data, to provide additional context and improved predictive accuracy in determining conditions. In an advantageous embodiment, the method offers the possibility of incorporating human expertise and feedback into the artificial intelligence training and validation process in order to increase model accuracy and reliability, particularly in cases where the amount of data is limited or incomplete. In an advantageous embodiment, the method enables the creation of personalized reports tailored to individual needs and requirements.by medical professionals and / or patients by adapting the presentation, information content and format of the reports. In an advantageous embodiment, the method comprises supplying real-time blood parameter data from the hematology analyzer in order to carry out continuous monitoring and real-time analysis of conditions. In an advantageous embodiment, the method comprises training the at least one deep learning model and / or the at least one machine learning model, wherein the training comprises supervised and / or unsupervised learning, wherein the supervised learning comprises using annotated data in a database to identify patterns and correlations, while the unsupervised learning enables recognition of patterns and correlations in the data of the database without prior annotation in order to obtain novel findings and possibly previously unknown conditions orto identify diseases; In an advantageous embodiment, the training comprises transfer learning, in which pre-trained models from related domains or applications are used as a starting point for training and adapted to the specific blood parameters and / or scattergrams in order to increase the efficiency and effectiveness of the training and reduce the required amount of training data; In an advantageous embodiment, the training comprises active learning, in which the at least one deep learning model and / or the at least one machine learning model specifically searches for examples in the database that can most improve its performance and accuracy. In an advantageous embodiment, the training comprises incorporating input from a user for annotating and / or confirming the examples in order to optimize the training process. In an advantageous embodiment, the training comprisesat least one ensemble learning method in which several learning methods are combined to increase the performance and accuracy of the predictions and to compensate for the weaknesses of individual models or algorithms. In an advantageous embodiment, the training comprises incremental learning in which the deep learning model and / or the machine learning model are continuously and gradually updated based on newly added blood parameters and / or scattergrams in the database in order to improve the performance and accuracy of the models over time and to be able to react to changes in the underlying data. In an advantageous embodiment, the method performs at least one data augmentation process to increase the size and diversity of the training data in the database in order to reduce the risk of overfitting and to improve the performance and accuracy of said models. In an advantageous embodimentThe at least one data augmentation process comprises a synthetic generation of blood parameters and / or scattergrams based on existing data, wherein stochastic methods, statistical models, or artificial intelligence algorithms are used to generate realistic and representative data for training the models. In an advantageous embodiment, the at least one data augmentation process comprises at least one transformation of existing blood parameters and / or scattergrams, wherein the at least one transformation comprises rotations, scalings, mirrorings, shearings, noise, and / or distortions in order to increase the diversity of the training data and increase the robustness of determining the at least one state. The noise can be understood as adding numerical noise. In an advantageous embodiment, the at least one data augmentation process comprises a combination of blood parametersand / or scatter plots from different sources, such as different devices, techniques, patient populations, or clinical studies. This can improve the representativeness of the training data for a broader application of the artificial intelligence models. In an advantageous embodiment, the at least one deep learning model and / or the at least one machine learning model is designed to adapt the degree of data augmentation based on boundary conditions, such as factors such as the size of the existing database, the number of previous training iterations, and / or the current performance and accuracy of the model in question, in order to improve the efficiency of training and make the training process more efficient. In addition to blood age, other in vitro and in vivo conditions can be determined using the method, which conditions include: - Anemia: A deficiency of red blood cells or hemoglobin in the blood can affect the size,Affect the shape and number of red blood cells. - Infections: Infections can change the number and type of white blood cells in the blood. - Inflammation: Inflammation can also affect the number and type of white blood cells, particularly the number of neutrophils. - Blood clotting disorders: Conditions that impair the function of platelets or clotting factors in the blood can affect blood clotting and the number and function of platelets. - Dehydration: Dehydration can increase the hematocrit level and increase the concentration of red blood cells in the blood. - Bone marrow disorders: Bone marrow disorders such as leukemia and myelodysplastic syndromes can affect the production of red and white blood cells in the bone marrow and produce abnormally shaped blood cells. - Hereditary disorders: Some hereditary disorders such as sickle cell anemia and thalassemia can affect the size and shape of redAffect blood cells. - Cancer: Cancers such as lymphomas and leukemias can affect blood counts by altering the number and function of blood cells in the blood. - Autoimmune diseases: Autoimmune diseases such as systemic lupus erythematosus and rheumatoid arthritis can affect the number and type of white blood cells. - Immune system diseases: Immune system diseases such as HIV / AIDS can affect the number and function of white blood cells and increase the risk of infections. - Kidney diseases: Kidney diseases can affect blood counts by impairing the production of erythropoietin, a hormone responsible for the production of red blood cells in the bone marrow. - Nutritional deficiencies: Nutritional deficiencies, particularly a deficiency in iron, vitamin B12, or folic acid, can affect blood counts and lead to anemia. - Medications: Some medications can affect blood countsby inhibiting the production of blood cells in the bone marrow or shortening the lifespan of blood cells. - Pregnancy: During pregnancy, the number of white blood cells in a woman's blood may be increased, while the number of red blood cells may be decreased due to increased plasma intake in a woman's body. - Genetic disorders: Certain genetic disorders such as sickle cell anemia and thalassemia can affect the size and shape of red blood cells and lead to anemia. - Liver disorders: Liver disorders such as cirrhosis can affect the number and type of blood cells, as the liver plays an important role in blood cell production. - Arteriosclerosis: A disease in which there is a buildup of fat and other substances in the artery walls. This can restrict blood flow and lead to an increased number of white blood cells in the blood. - CoronaryHeart disease: A condition in which plaque builds up in the coronary arteries, which can lead to a reduction in blood flow to the heart. This can result in the heart receiving less oxygen and an increase in the number of red blood cells in the blood. - Heart attack: A heart attack occurs when blood flow to the heart muscle is suddenly blocked, usually by a blood clot. This can cause damage to the heart muscle and affect the number of blood cells in the blood. - Heart failure: A condition in which the heart is no longer able to pump enough blood to supply the body with oxygen and nutrients. This can lead to an increase in the number of white blood cells and impairment of the red blood cells. - Blood age after blood collection: After blood samples are collected, various factors, such as temperature, storage conditions, and type of treatment, can affect theChanges in blood cell properties. Blood cells, especially red blood cells, can age over time and change their functions and properties. The decomposition of blood cells can be caused by bacteria or enzymes that are present in the blood or that can be added during storage. Changes in cell membranes: The cell membranes of blood cells can be damaged by the action of chemical substances or physical forces. - Biological age: With increasing age, the number of white blood cells in the blood can decrease, while the number of red blood cells and platelets can increase. The type of blood cells can also change, e.g., the size and shape of the red blood cells can change. The invention further relates to a system for determining conditions in vivo and in vitro by analyzing blood parameters measured in a hematology analyzer, comprising aA computer device having a computing unit, a storage unit connected thereto, and an input unit, wherein the system is designed to a) acquire blood parameters measured in a hematology analyzer via the input unit, wherein the blood parameters comprise quantitative and qualitative measured variables, wherein the measured variables comprise properties of individual cells, wherein the individual cells comprise white blood cells, red blood cells, and platelets, wherein the white blood cells comprise monocytes, lymphocytes, basophils, eosinophils, and neutrophils; b) create at least one scattergram by the computing unit, wherein each axis of the scattergram comprises a different measured variable; c) determine at least one in-vivo and / or in-vitro and / or post-mortem state by means of at least one deep learning model and / or a machine learning model, wherein the at least one deep learning model comprises at least oneScatter diagram image as input, wherein the at least one machine learning model contains at least one 1D vector as input; and d) automatically generate a report by the computing unit, wherein the report comprises at least one result about the determination of the at least one condition. In an advantageous embodiment, the system has a communication interface for transmitting results and reports to other computer systems, laboratory information systems (LIS), hospital information systems (HIS) and / or electronic patient records (EPA). The system is further preferably designed to carry out at least one or more of the method steps outlined above. Embodiment: Age of the Blood This demonstrates for the first time how, with the aid of deep learning models, the age of in vitro blood samples can be precisely determined after collection, based on the properties of the blood cells, which are determined bya hematology analyzer. A total of 228 venous blood samples were measured on the hematology analyzer. 149 of the 228 blood samples were measured within the first two hours after blood collection, and 79 of the 228 blood samples were measured 24 hours after blood collection. All measurements were performed on an Abbott Cell-Dyn Ruby hematology analyzer. The measurements were transferred from the hematology analyzer to the laboratory information system (LIS) according to ASTM protocol and downloaded and saved as TEXT files (Fig. 7). Following the instructions in the Cell-Dyn Ruby System Host Interface Specification (LIST NO. 09H05-01 Revision C), the inventors of the patent were able to decrypt the encrypted information in the TEXT files using a self-written algorithm in Python and save it as 228 Excel spreadsheets in a folder named database. 149 tables with the measurement data of the 2-hour blood sample were stored in a folderand 79 tables with the measurement data of the 24-hour blood sample were saved in the other folder (Fig. 8). Each table consisted of 7 columns and 2000 rows (Fig. 9), where: - The 1st column contained the measurements from the ALL channel (Axial Light Loss), representing cell size, or light scatter signals below 0°; - The 2nd column contained the measurements from the IAS channel (Intermediate Angle Scatter), representing cellular complexity, or light scatter signals below 10°; - The 3rd column contained the measurements from the PSS channel (Polarized Side Scatter), representing the nuclear lobularity of the cells, or light scatter signals below 90°; - The 4th column contained the measurements from the DSS channel (Depolarized Side Scatter), representing the granularity of the eosinophils; - Columns 5 to 7 contained the measurements of the properties of the red blood cells and platelets. This database was randomly divided into three parts (Fig. 10):Training, validation, and test datasets. The training dataset was used to train a deep learning model. The model was trained on the training dataset to learn how to transform specific inputs into specific outputs. During training, the model iteratively traversed the training data, adjusting its internal parameters to make the outputs as close as possible to the expected outputs (<2h=0, 24h=1). The validation dataset was used to monitor the performance of the deep learning model during the training process after each training epoch and to prevent model overfitting. Overfitting occurs when the model fits the training data too well but performs poorly when predicting new data. The validation dataset consists of data that is not used for training the model but solely for monitoring the model during training. After trainingThe model was tested on an independent test data set to evaluate its performance. If the model performs well on the test data, it can be used to make predictions on new, previously unknown data. A convolutional neural network (CNN) in Keras was chosen as the deep learning model (Fig. 11). Keras is an open-source library for deep learning in Python. It is designed to facilitate the development of deep learning models through a simple, intuitive, and modular interface. Keras offers a variety of pre-built layers and models that allow developers to quickly and easily build and train complex models and can be used for both scientific and commercial purposes. Keras is designed to be simple and user-friendly and offers a fast and effective way to build and train deep learning models without having to worry about the details of theimplementation. It supports both CPU and GPU computation and can run on various platforms such as Windows, Linux, and macOS. Keras is licensed under the MIT License. The MIT License is a permissive open-source software license that allows users to use, copy, modify, and distribute the code for various purposes, as long as the license text and the copyright notice are maintained. The MIT License contains few restrictions and is therefore business-friendly, as it allows the use and adaptation of the software in commercial and proprietary projects. The deep learning model that was trained consisted of several layers, each of which performed specific operations to transform the input image (scatterplot) into an output. Here is a brief explanation of each layer: - Conv2D: A convolutional layer that applies a fixed number of filters to theinput image and extracts the image's features. - BatchNormalization: A normalization layer that normalizes the activations in the previous layer to reduce gradient exploding and vanishing and accelerate convergence. - MaxPooling2D: A pooling layer that reduces the size of the output image by extracting the maximum value in each pooling window. - Flatten: A layer that converts the 2D output image into a 1D vector to pass it to a fully connected layer. - Dropout: A layer that randomly sets some of the previous layer's output elements to zero to reduce overfitting. - Dense: A fully connected layer that transforms the previous layer's activations into a new space. - Activation: A layer that activates the previous layer's output using an activation function such as ReLU or Sigmoid. The last dense layer had only oneOutput nodes with the sigmoid activation function, as this is a binary classification problem (<2h=0, 24h=1). The model was trained using a binary cross-entropy loss and the Adam optimizer. During training, validation, and testing, 2D scatterplots of size 256x256 were created from the ALL-channel measurements (1st column in the Excel spreadsheets) and IAS-channel measurements (2nd column in the Excel spreadsheets) as input for the deep learning model (Fig. 12). After training, the deep learning model was tested on an independent test dataset to evaluate its performance. AUROC was chosen as the performance metric. AUROC stands for "Area Under the Receiver Operating Characteristic Curve" and is a metric for evaluating the performance of a binary classifier, e.g., a deep learning model that distinguishes between two classes. The Receiver Operating Characteristic (ROC) curve is a graphical representation that shows how well aThe classifier's ability to distinguish between classes is measured by plotting the true positive rate (TPR) against the false positive rate (FPR). The area under the ROC curve (AUROC) indicates how well the classifier performs overall and is a measure of the model's performance. A perfect prediction would have an AUROC of 1, while a random prediction would have an AUROC of 0.5. A higher AUROC means the classifier achieves better separation between classes and thus has higher performance. AUROC is an important metric for evaluating the performance of deep learning models, especially in medical image classification, diagnosis, or disease prediction. The following figure shows the ROC curve and AUC achieved on the test dataset. The AUROC is 0.96 (Fig. 13). An AUROC of 0.96 indicates that the model has a very high probability of correctly predictingTo make predictions in the future. With a threshold of 0.788, the following results were achieved on the test dataset (Fig. 14): The confusion matrix shows that all 46 blood samples less than 2 hours old were correctly identified. Thus, the specificity is 100.00%. In this example, the specificity indicates how well the model identifies the blood samples that are 2 hours old. Of 21 blood samples taken after 24 hours, 20 blood samples were correctly identified. Thus, the sensitivity is 95.24%. The sensitivity indicates how well the model actually identifies blood samples that are 24 hours old. Sensitivity and specificity are important metrics for evaluating the performance of a classification model. However, since a high AUROC was achieved on a test dataset with only 67 data points, the results were interpreted with caution, as a small test dataset may not be representative of the entire dataset.In this case, a high AUROC could also be the result of randomness. It was decided to perform a 4-fold cross-validation. The use of cross-validation methods such as the 4-fold cross-validation method serves to evaluate the performance of a deep learning model and ensure that the model can generalize well not only to the specific data, but also to other data. In the 4-fold cross-validation method, the dataset is divided into four roughly equal subsets, with each subset being used once as a test dataset and three times as a training dataset. This division allows the model to be trained and tested on different combinations of training and test datasets, thereby improving the robustness of the results. In a 4-fold cross-validation method, the AUROC is calculated for each test dataset, and the average AUROC across all test datasets is used as a measure of performance.of the model. By using a 4-fold cross-validation procedure, the robustness of the results can be improved, as they are based on a larger set of test datasets. The following figure shows AUROCs for all 4 test folds, with the average AUROC being 93.13% (Fig. 15). A mean AUROC of 0.93 after the 4-fold validation means that the deep learning model still shows very good to excellent performance in binary classification. An AUROC of 0.93 is close to the ideal performance of a binary classifier (AUROC=1), indicating that the model is able to distinguish the two classes (2-hour blood vs. 24-hour blood) very well. Such a high level of accuracy is particularly important in medical applications where it is important to determine a clinical condition with high certainty or to make a prediction with high confidence. The solution described in the example forPredicting the age of in vitro blood samples using deep learning models has the potential to be used in various fields. Some possible applications are: - Medical research: Predicting blood age can be important in the investigation of blood disorders, blood cell aging, and the effects of medications or treatments on the blood. - Transfusion medicine: In blood transfusion, it is important to ensure the freshness of blood units. This solution could help monitor the quality of blood units and ensure that they are used within the appropriate timeframe. - Forensics: In forensic science, determining the age of blood samples could help more accurately determine the time of crimes or accidents. - Biotechnology and pharmaceutical industry: The technology can be used in the development of new drugs or therapeutic approaches by helpingContributes to a deeper understanding of the effects of substances on blood and blood cells. It is important to note that the performance and practical applicability of this solution depend on the quality and robustness of the trained model. To ensure the model's effectiveness and reliability, it can be tested on larger and more diverse datasets in clinical trials and adapted as needed. Example: Suspected acute coronary syndrome. A 36-year-old patient entered the admissions department with a preliminary diagnosis of "arteriosclerotic heart disease, acute coronary syndrome without ST elevation, acute heart failure, Killip class I," two hours after experiencing a typical pain syndrome. An electrocardiographic examination was performed in the admissions department, which also showed no ST-segment elevation on the electrocardiogram. Venous blood samples were then taken for laboratory testing.Blood samples collected outside of the procedure were provided for examination, which also included the highly sensitive cardiac troponin I method and clinical blood analysis. The results of laboratory studies were: BUN 4.2 mmol / L (3.0-9.2); ALT 16 units / L (0-55); AST 12 units / L (5-34); total protein 70 g / L (64-83); creatinine 74 µmol / L (64-111); total bilirubin 6.2 µmol / L (3.4-20.5); glucose 7.5 mmol / L (3.9-5.5); potassium 3.7 mmol / L (3.5-5.1); sodium 137 mmol / L (135-145); Calcium ionized 1.23 mmol / l (1.13 - 1.32); APTV 78.7 s (25.1 - 36.5); MNO 0.97 (0.2 - 0.5); 90 - 1.20); Prothrombin 118.0% (70.0 - 140.0); Prothrombin time 11.0 s (9.4 - 12.5); Leukocytes 12.4 10E9 / l (4.0 - 9.0); (NEUT) Neutrophils 10.0*109 / l (2.0 - 5.5); (NEUT%) Neutrophils 80.0% (48.0 - 78.0); (LYM) Lymphocytes 1.79*109 / l (1.20 - 3.00); (LYM%)-lymphocytes 14.3% (19.0-37.0); (MON)-monocytes 0.57 109 / l (0.09-0.60); (MON%) monocytes 4.6%(3.0 - 11.0); (EOS) Eosinophils 0.07*109 / l (0.00 - 0.30); (EOS%) Eosinophils 0.52% (1.00 - 5.00); (BAS) Basophils 0.07*109 / l (0.00 - 0.06); (BAS%) Basophils 0.52% (0.00 - 1.00); (HGB) Hemoglobin 134 g / l (130 - 160); (HCT) Hematocrit 40.7% (40.0 - 48.0); (RBC) Erythrocytes 4.45 10*12 / l (4.00 - 5.60); (MCH) mean red blood cell hemoglobin concentration 30.1 pg (24.0-34.0); (MCHC) mean red blood cell hemoglobin concentration 32.9 g / dL (30.0-38.0); (MCV) mean red blood cell volume 91.4 fL (75.0-95.0); (RDW-CV) red blood cell distribution 11.0% (11.5-16.0%); (PLT) platelets 262*109 / L (180-400); (MPV) mean platelet volume 11.5 fL (7.4-10.4); cardiac troponin I (high-sensitivity method) 37.9 ng / ml (upper reference limit 26.2 ng / ml). After the blood was measured, the raw measurement data was transferred to the PC. Two scatter plots were then generated for the trained deep learning models andinto a global scatter plot. In parallel, the global vector with 4216 elements was derived: ^^^^ where 4216 elements represent the measured properties of the leukocyte subpopulations. The global vector is standardized by deleting the mean and scaling to a single dispersion based on a predefined database: V ^ ^ ^ ^ ! ^ ^" =#V^^^^$ − #u$"'^^^('^)#s$"'^^^('^)where the vector {u} represents the mean values ​​of the feature elements of the global vectors in the pre-trained database and the vector {s} represents the standard deviations of the elements of the global vectors in the pre-trained database. After standardization, PCA was applied as an example to reduce the dimensionality of features from 4216 elements to 4 elements, which are called principal components, while retaining the greatest possible variability (information) of the features. After applying principal component analysis, all 4216 elements of the standardized global vector were reduced to the vector in the 4-dimensional subspace: − 40,48 58,66 − 5,4 ^ 0,75At first glance, it seems difficult to extract information for patient diagnosis from the values ​​of this reduced vector. For this purpose, an ensemble of trained machine learning models is used. The ensemble consists of an ensemble of artificial neural networks, an ensemble of K-Nearest Neighbors models, an ensemble of Random Forest models, an ensemble of AdaBoost models, an ensemble of Gradient Tree Boosting models, and an ensemble of Support Vector Machines models, with each ensemble trained on the pre-trained database. The standardized and reduced global vector is used as the input vector for all ensembles, while the global scatterplot is used for a deep learning model. For the above-mentioned case, the votes for AKS are counted from individual ensembles (hard voting): The final result for AKS after the hard voting procedure was positive. The decision was made to perform percutaneous coronary intervention. The patient underwent coronary angiography, followed by transluminal dilation and stenting of the infarct-dependent coronary artery. CORONARYGRAPHY No. 7175 dated June 7, 2018: Left circulation type. Left coronary artery: Barrel - no stenosis. Anterior ventricular branch - 5-50% stenosis of the mouth, subocclusion in the middle third. Intermediate artery - 90% stenosis in the proximal third. Diagonal branches - no stenosis. Bend branch (BB) - no stenosis. Blunt-edged branches - no stenosis. Right coronary artery: hypoplasia. Acute edge branch: no stenosis. Posterior branch (ROB) - no stenosis. Posterior interventricular branch - without stenosis. ORONAROPLASTY AND PMV STENTING No. 7176 dated June 7, 2018: PMV stenosis zone (average third) BC 2.0*20.0 mm, p=18 atm.A drug-coated stent measuring 2.75 x 33.0 mm was implanted in the middle third of the 2.0 x 20.0 mm BC, p = 16 atm. Control: TIMI grade III blood flow. No infiltration shadows were detected on chest X-rays taken on two projections from June 8, 2018. The roots are structural, not enlarged, and the left one is partially blocked. The lung pattern was unchanged. The diaphragm is contoured. The cardiac shadow is without facial features. The sinus is clear. The following treatment was administered: beta-blockers, anticoagulants, double disaggregation therapy, statins (the dose of Crestor was reduced by 20->10 mg / day due to an increase in transaminase levels), and gastroprotectors. The patient refused rehabilitation treatment in a sanatorium. In the postoperative period, the maximum concentration of cardiac troponin I in dynamic observation reached 7522.5 ng / ml. The hospital stay lasted 12 days.The final diagnosis was "arteriosclerotic heart disease." Acute myocardial infarction of the anterior parietal region, high lateral portions of the left ventricle without ST-segment elevation on June 7, 2018. Coronary artery stenting and stenting were performed on June 7, 2018. The patient was discharged to his home on June 19, 2018, for further outpatient observation. Brief description of the figures: The figures show: Fig. 1 Measurement technology of a hematology analyzer. Fig. 2 Measurement technology of another hematology analyzer. Fig. 3 Measurement technology of another hematology analyzer. Fig. 4 Prediction of a condition using an ensemble approach. Fig. 5 Procedure in a laboratory setting. Fig. 6 Steps of the computer-implemented procedure. Fig. 7 Measurement data. Fig. 8 Creation of a database from measurement data. Fig. 9 Measurement data. Fig. 10 Division of measurement data into training, validation, and test datasets.11 Architecture of a deep learning model for blood age determination Fig. 12 Scatter plot Fig. 13 AUROC curve Fig. 14 Metrics for test data set for blood age determination Fig. 15 Four AUROC curves Fig. 1 provides an example of the measurement technology used in Sysmex hematology analyzers (https: / / www.sysmex.com). Sysmex has developed an innovative fluorescence flow cytometry technology that provides detailed information about cell size, cell structure, and cell contents. In flow cytometry, cells and particles are analyzed by passing them through a very narrow flow cell. The process begins with taking a blood sample, which is then dosed and diluted to a specified ratio. The sample is then labeled with a special fluorescent marker developed by Sysmex that specifically binds to nucleic acids.In the next step, the labeled sample is transported into the flow cell. During analysis, the sample is illuminated with a semiconductor laser beam, which allows cells to be differentiated based on three different signals: 1) Forward scattered light (FSC): This signal is proportional to the size of the cell and allows conclusions to be drawn about the cell size. 2) Side scattered light (SSC): This signal provides information about the internal cell structure and complexity, as it is related to the granularity and structural properties of the cell. 3) Side fluorescence light (SFL): This signal detects the RNA / DNA content of the cell by measuring the fluorescence intensity generated by the fluorescent label bound to nucleic acids. By combining these three signals, Sysmex's FFC technology enables precise analysis and differentiation of cells in blood samples. Fig.Figure 2 illustrates Abbott's measurement technology (https: / / www.abbott.com / ) as an example. Abbott has developed the innovative and patented MAPSS™ (Multi-Angle Polarized Scatter Separation) technology, which delivers precise WBC and differential results in the first run through laser measurement of up to 20,000 cells and four angles of optical analysis from a single dilution. MAPSS™ technology uses four light scattering detectors to determine different cellular properties. A unique feature of this method is the use of a depolarized light detector, which enables the specific identification of eosinophil granulocytes. The four detectors generate the following signals: 0° or Axial Light Loss (ALL): This signal is related to the size of the cell and allows conclusions to be drawn about cell size.0° to 10° Intermediate Angle Scatter (IAS): This signal depends on cellular complexity and provides information about the internal cell structure. 90° Polarized Side Scatter (PSS): This signal relates to the nuclear lobularity or segmentation of the cell and allows the identification of cellular structural features. 90° Depolarized Side Scatter (DSS): This signal relates to eosinophil granules and enables the specific detection of eosinophil granulocytes in the sample. Figure 3 illustrates the measurement technology from Beckman Coulter (https: / / www.beckmancoulter.com / ). For example, the DxH 800 / DxH 600 instruments feature the Multi-Transducer Module (MTM), which measures multiple angles of light scattering.The MTM flow cell measures volume, conductivity, multiple angles of light scattering, and axial light loss: Bottom electrode (DC and RF > conductivity): This electrode measures cell conductivity using DC and AC current. Top electrode (DC and RF > conductivity): Similar to the bottom electrode, this electrode measures cell conductivity using DC and AC current. Axial light loss (ALL) 0°: This signal is related to cell size and allows conclusions to be drawn about the cell size. Low-angle light scattering (LALS) 5.1°: This angle of light scattering provides information about cellular granularity and structure. Low-median angle light scattering (LMALS) 10°-20°: This range of light scattering allows a more precise determination of cell structure and complexity.Upper median angle light scattering (UMALS) 20°–42°: This light scattering range allows for even more precise analysis of cell structure and complexity. The fifth light scattering channel is the sum of the UMALS and LMALS regions (called MALS): Combining the information from both median angle ranges enables a more comprehensive analysis of cell properties. Figure 4 shows an example of the prediction of a condition using an ensemble approach. A scattergram is created from the blood parameters obtained from a hematology analyzer. This plot serves as input for deep learning models such as convolutional neural networks (CNNs). At the same time, a one-dimensional (1D) vector is generated from the scattergram using a flattening method. This 1D vector is used as input for machine learning models that can process one-dimensional structural data.A prediction for the patient's condition is determined from all models. The final determination of the condition is based on a soft-voting or hard-voting procedure, such as the weighted averaging of the probabilities for the presence of a particular condition. By combining different models and approaches in an ensemble, the strengths of each model can be utilized and weaknesses compensated. This leads to improved prediction accuracy and contributes to the effective diagnosis and monitoring of hematological diseases. Fig. 5 shows an example of the procedure in a laboratory environment. The measurement data are taken from the Abbott Cell-Dyn Ruby hematology analyzer 1 and transmitted to the laboratory information system (LIS) 3 in ASCII format according to the ASTM protocol 2. The blood parameters for the creation of scattergrams are provided in binary format 4.The blood parameters are then decoded 5 according to the Cell-Dyn Ruby System Host Interface Specification (LIST NO. 09H05-01 Revision C) and saved in an Excel spreadsheet 6. A scatter plot 7 is created from the columns of this Excel spreadsheet and serves as input for the deep learning model 8. Based on the scatter plot, the deep learning model makes a prediction for the patient's condition. By integrating modern analytical devices, standardized protocols, and AI-based models in a laboratory environment, efficient and reliable diagnosis and monitoring of hematological diseases can be achieved. The exemplary representation in Fig. 6 summarizes all steps of the computer-implemented method. A blood sample taken prior to the method is provided for examination in the method according to the invention, which is referred to here as step (1).In step (2), the quantitative and qualitative properties of the individual cells in the blood sample are measured in an analyzer. In step (3), the measurement data are transferred to the computer. In step (4), the scattergram is created. In step (5), the scattergram is used as input for the deep learning model. In parallel, a 1D vector can be derived from the scattergram and used as input for the machine learning model. In step (6), a condition is determined (e.g., blood age in vitro or a disease in vivo). In step (7), the diagnosis is displayed on a graphical interface or in the form of an automatically generated report. In step (9), the predicted diagnosis can be saved in the pre-built database together with corresponding measurement data from (3) in order to expand the database with a new case and train new models using a semi-supervised approach.The pre-built database can also be expanded with measurement data and the corresponding true diagnoses in step (8). The purpose of the database is to train and evaluate the models. The models in step (5) can be replaced at any time by the newly trained models in step (10), as long as the newly trained models achieve better results on the test data set. Fig. 7 shows an example of the measurement data that was transferred from the hematology analyzer to the laboratory information system (LIS) according to the ASTM protocol. This data was downloaded and saved as text files. The blood parameters for creating the scattergrams are included in binary format. The use of standardized protocols such as the ASTM protocol enables reliable and consistent transfer of measurement data between different systems, such as the hematology analyzer and the LIS.By storing measurement data in text format and providing scattergram blood parameters in binary format, both readability for human users and efficient processing and analysis by computer models can be ensured. Figure 8 shows an example of the creation of a database from measurement data representing blood age. Using the instructions in the Cell-Dyn Ruby System Host Interface Specification (LIST NO. 09H05-01 Revision C), the patent inventors were able to decrypt the encrypted information in the files and save them as 228 Excel spreadsheets in a folder named "database." The database consists of two separate folders: one folder with 149 spreadsheets containing measurement data from 2-hour blood samples and another folder with 79 spreadsheets containing measurement data from 24-hour blood samples.This organization facilitates data management and analysis by allowing the separation of samples of different ages. Fig. 9 shows an example of an excerpt from an Excel spreadsheet representing measurement data acquired in the various channels of the Abbott Cell-Dyn Ruby hematology analyzer. This table was created by transforming the LIS files transferred from the hematology analyzer to the laboratory's LIS. This Excel spreadsheet lists the various measurement values ​​for the respective cells and particles acquired during the analysis. Organizing the data in such a tabular format enables a simple and clear presentation of the information and facilitates further analysis and processing of the measurement data. Fig.Figure 10 shows an example of the random division of the measurement data into training, validation, and test datasets for determining blood age (<2h=0, 24h=1). This random division of the data ensures that the models are not biased towards specific patterns within the datasets and that an appropriate representation of the different blood ages is guaranteed within the training, validation, and test datasets. By using separate datasets for training, validation, and testing, the models can be continuously evaluated and optimized during the training process, and their performance on unknown data in the test dataset can be assessed. Figure 11 shows an example of the architecture of the deep learning model in Keras that was trained for determining blood age.In this illustration, you can see the different layers and components of the deep learning model, such as input layers, convolutional layers, activation functions, pooling layers, dropout layers, and fully connected (dense) layers, which are connected to each other to build the neural network. The Keras library enables simple and user-friendly implementation and customization of the model architecture. By training the model with the prepared training data, the deep learning model can recognize patterns and relationships in the data and thus predict the blood age of unknown samples. Fig. 12 shows an example scatter plot. Fig. 13 shows the area under the receiver operating characteristic (AUROC) curve for the test dataset when determining blood age (<2h=0, 24h=1).The AUROC curve is a graphical representation of the performance of a classification model at different classification thresholds. It is often used to evaluate the diagnostic ability of a model, especially in medical applications. The AUROC curve shows the sensitivity (true positive rate) on the vertical axis and the specificity (1 – false positive rate) on the horizontal axis. A larger area under the curve (AUC) indicates better prediction accuracy of the model. In this example plot, the AUROC curve shows the ability of the trained deep learning model to correctly determine blood age (<2h=0, 24h=1) on the test dataset. A high AUC means that the model is able to distinguish between the different blood ages and make accurate predictions. Fig. 14 shows the metrics for the test dataset in determining blood age (<2h=0, 24h=1). Fig.Figure 15 shows the four AUROC curves for the four test folds generated during the cross-validation procedure for determining blood age (<2h=0, 24h=1). Cross-validation is a common method for evaluating the performance of a model, in which the data is divided into several subsets (in this case, four folds). During the cross-validation procedure, the model is trained on three of the folds and tested on the remaining fold. This process is repeated for all four folds, so that each fold is used as a test dataset exactly once. The figure shows the AUROC curves for the four test folds generated during the cross-validation procedure. Each curve represents the model's performance in determining blood age on one of the test folds. The area under the receiver operating characteristic (AUROC) curve indicates the model's diagnostic ability at different classification thresholds.A high AUC indicates that the model has a good ability to distinguish between different blood ages and make accurate predictions. The robustness and reliability of the model can be assessed by analyzing the AUROC curves for the four test folds.

Claims

Patent claims 1. Computer-implemented method for determining conditions in vivo, in vitro and / or post-mortem by analyzing blood parameters measured in a hematology analyzer, the method comprising: a) obtaining blood parameters of a blood sample by means of a hematology analyzer, the blood parameters comprising quantitative and qualitative measurements, the measurements comprising properties of individual cells, the individual cells comprising blood cells, the blood cells comprising white blood cells, red blood cells and platelets, the white blood cells comprising monocytes, lymphocytes, basophils, eosinophils and neutrophils; b) creating at least one scattergram with at least two axes, each axis of the scattergram comprising a different measurement from step a; c) determining at least one in vivo and / or in vitro and / or post-mortem condition by i.at least one deep learning model, wherein the input variable for the at least one deep learning model comprises at least one scattergram from step b and / or ii. at least one machine learning model, wherein the input variable for the at least one machine learning model comprises at least one 1D vector, wherein the 1D vector is created by vectorizing the at least one scattergram from step b; and d) automatically generating a report comprising at least one result about the determination of the at least one state.

2. A computer-implemented method for automatically generating a report comprising at least one result on the determination of conditions in vivo, in vitro and / or post-mortem through an analysis of blood parameters measured in a hematology analyzer, the method comprising: a) obtaining blood parameters of a blood sample using a hematology analyzer, wherein the blood parameters comprise quantitative and qualitative measurements, wherein the measurements comprise properties of individual cells, wherein the individual cells comprise blood cells, wherein the blood cells comprise white blood cells, red blood cells and platelets, wherein the white blood cells comprise monocytes, lymphocytes, basophils, eosinophils and neutrophils; b) creating at least one scattergram with at least two axes, wherein each axis of the scattergram comprises a different measurement from step a;c) determining at least one in-vivo and / or in-vitro and / or post-mortem condition by i. at least one deep learning model, wherein the input variable for the at least one deep learning model comprises at least one scattergram from step b, and / or ii. at least one machine learning model, wherein the input variable for the at least one machine learning model comprises at least one 1D vector, wherein the 1D vector is created by vectorizing the at least one scattergram from step b; d) automatically generating a report comprising at least one result about the determination of the at least one condition; and e) transmitting the report from step d by means of a data signal f) receiving the transmitted report.

3. The computer-implemented method according to claim 1 or 2, wherein the measured variables of the individual cells include the number of cells, size, shape, volume, complexity, granularity, electrical conductivity, light scattering at different angles, mean corpuscular volume (MCV), mean corpuscular hemoglobin content (MCH), mean corpuscular hemoglobin concentration (MCHC), red blood cell distribution width (RDW), mean platelet volume (MPV), and / or platelet distribution width.

4. The computer-implemented method according to at least one of the preceding claims, wherein scattergrams are subjected to processing before being analyzed by the at least one deep learning model, wherein the processing comprises resizing, normalization, standardization, noise reduction, test time augmentation (TTA), clustering, contrast adjustment, and / or filtering, individually or in combination. 5.Computer-implemented method according to at least one of the preceding claims, wherein the at least one deep learning model comprises convolutional neural networks, generative adversarial networks, recurrent neural networks, long short-term memory networks, transformer networks, 3D convolutional neural networks, and / or 4D convolutional neural networks.

6. Computer-implemented method according to claim 1 or 2, wherein the at least one 1D vector is subjected to processing before analysis by the at least one machine learning model, wherein the processing comprises normalization, standardization, scaling, dimensionality reduction, etc. Noise reduction and feature selection individually or in combination.

7. The computer-implemented method according to claim 6, wherein the dimensionality reduction and / or the feature selection of the a 1D vector at least one processing method from a of processing methods, the group of processing methods comprising principal component analysis, T-distributed stochastic neighbor embedding, linear discriminant analysis, truncated singular value decomposition, uniform manifold approximation and projection, independent component analysis, sparse representation, partial least squares regression, and kernel principal component analysis.

8. Computer-implemented method according to at least one of the preceding claims, wherein the at least one machine learning model comprises K-nearest neighbors, support vector machines, decision trees, random forests, multi-layer perceptrons, Adaboost models, gradient-boosting models, Naive Bayes, one-class support vector machines, isolation forests, local outlier factors, and / or support vector data descriptions. 9.The computer-implemented method according to at least one of the preceding claims, wherein more than two deep learning models and / or more than two machine learning models and / or a combination of at least one deep learning model and at least one machine learning model comprise an ensemble.

10. The computer-implemented method according to claim 9, wherein the at least one state is determined using an ensemble technique, wherein the ensemble technique comprises bagging. Boosting, stacking, hard voting, soft voting, random subspace, mixture of experts, and / or Bayesian model averaging.

11. The computer-implemented method according to at least one of the preceding claims, wherein in vitro conditions are based on processes outside a living organism, including changes in blood cell morphology, cell composition, cell function, or other characteristics of the individual cells as a result of storage, handling, and / or analysis.

12. The computer-implemented method according to at least one of the preceding claims, wherein in vivo conditions are based on processes within a living organism, including physiological and pathological conditions such as disease, biological age, pregnancy, drug exposure, health status, nutritional deficiencies, hereditary disorders, dehydration, blood clotting disorders, infections, and / or anemia. 13.A computer-implemented method according to at least one of the preceding claims, wherein post-mortem states are based on processes of a dead organism, including changes in the morphology, cell composition, cell function, or other characteristics of individual cells due to diseases, the presence of medications, pre-mortem health status, drugs, poisons, and / or toxic substances, as well as changes caused by the decay and autolysis of cells and tissues after death.

1. A computer-implemented method according to at least one of the preceding claims, wherein the at least one deep learning model and / or the at least one machine learning model is based on a pre-built database. trained and / or validated, wherein the database comprises measured blood parameters and / or scattergrams, wherein the database is expandable with new measured blood parameters and / or scattergrams to improve performance and accuracy, wherein the database comprises information about known conditions, diseases, or other relevant information that contributes to the interpretation and analysis of the measured blood parameters and / or scattergrams; 1. A computer-implemented method according to at least one of the preceding claims, wherein it enables the integration and use of external data sources, including clinical data, demographic information, medical history, and / or genetic data, to provide additional context and improved predictive accuracy in determining conditions. 16.Computer-implemented method according to at least one of the preceding claims, wherein the method comprises supplying real-time blood parameter data from the hematology analyzer in order to perform continuous monitoring and real-time analysis of conditions.

1. Computer-implemented method according to at least one of the preceding claims, wherein the method comprises training the at least one deep learning model and / or the at least one machine learning model, wherein the training comprises supervised and / or unsupervised learning, wherein the supervised learning comprises using annotated data in a database to identify patterns and relationships, while the unsupervised learning enables recognition of patterns and relationships in the data of the database without prior annotation in order to. to identify novel findings and possibly previously unknown conditions or diseases; 18. The computer-implemented method according to claim 17, wherein the training comprises transfer learning, in which pre-trained models from related domains or applications are used as a starting point for training and adapted to the specific blood parameters and / or scattergrams in order to increase the efficiency and effectiveness of the training and reduce the required amount of training data; 19. The computer-implemented method according to claim 17 or 18, wherein the training comprises active learning, in which the at least one deep learning model and / or the at least one machine learning model specifically searches for examples in the database that can most improve its performance and accuracy.The computer-implemented method of claim 19, wherein training comprises incorporating user input for annotating and / or confirming the examples in order to optimize the training process.

21. The computer-implemented method of at least one of claims 17 to 20, wherein training comprises at least one ensemble learning method in which multiple learning methods are combined.

22. The computer-implemented method of at least one of claims 17 to 20, wherein training comprises incremental learning in which the deep learning model and / or the machine learning model is continuously and step by step based on. of newly added blood parameters and / or scattergrams in the database.

23. The computer-implemented method according to at least one of claims 17 to 21, wherein the method performs at least one data augmentation process to increase the size and diversity of the training data in the database to reduce the risk of overfitting.

24. The computer-implemented method according to claim 23, wherein the at least one data augmentation process comprises synthetically generating blood parameters and / or scattergrams based on existing data, using stochastic methods, statistical models, or artificial intelligence algorithms to generate realistic and representative data for training the models. 25.The computer-implemented method of claim 23 or 24, wherein the at least one data augmentation process comprises at least one transformation of existing blood parameters and / or scattergrams, the at least one transformation comprising rotations, scalings, mirrorings, shearings, noise, and / or distortions to increase the diversity of the training data and increase the robustness of determining the at least one state.

26. The computer-implemented method of at least one of claims 23 to 25, wherein the at least one data augmentation process comprises a combination of blood parameters and / or scattergrams from different sources.

2. The computer-implemented method of at least one of claims 23 to 26, wherein the at least one deep learning. Model and / or the at least one machine learning model is designed to adapt the degree of data augmentation based on boundary conditions, such as the size of the existing database, the number of previous training iterations and / or the current performance and accuracy of the model in question, in order to improve the efficiency of training.

28. System for determining conditions in vivo, in vitro and / or post-mortem by analyzing blood parameters measured in a hematology analyzer, comprising a computer device with a computing unit, a memory unit connected thereto and an input unit, wherein the system is designed to a) acquire blood parameters measured in a hematology analyzer via the input unit, wherein the blood parameters comprise quantitative and qualitative measured variables, wherein the measured variables comprise properties of individual cells, wherein the individual cells are white blood cells,red blood cells and platelets, wherein the white blood cells comprise monocytes, lymphocytes, basophils, eosinophils and neutrophils; b) to create at least one scattergram by the computing unit, wherein each axis of the scattergram comprises a different measured variable from step a; c) to determine at least one in-vivo and / or in-vitro and / or post-mortem state by means of at least one deep learning model and / or a machine learning model by means of the computing unit, wherein the at least one deep learning model receives at least one scattergram image from step b as input variable, wherein the at least one machine learning model contains at least one 1D vector as input variable, wherein the creation of the 1D vector is carried out by, Vectorization of the at least one scattergram from step b is carried out; d) Automatic generation of a report by the computing unit, wherein the report comprises at least one result regarding the determination of the at least one condition.

29. The system according to claim 28, wherein the system has a communication interface for transmitting results and reports to other computer systems, laboratory information systems (LIS), hospital information systems (HIS), and / or electronic patient records (EPA).

30. A data signal that transmits the report automatically generated in the method according to claim 1 or 2.

Images