Dynamic culture guidance system based on gram-morphology analysis in positive blood culture and body fluid samples.

Abstract

The invention encompasses an integrated microbiological decision support and automation system that automatically classifies Gram staining and morphological analysis results obtained from positive blood culture and direct body fluid samples using artificial intelligence-based image processing algorithms, and utilizes this classification data as closed-loop decision input for dynamic medium selection, robotic inoculation guidance, and laboratory management.

Description

1 TARIFF GRAM MORPHOLOGY IN POSITIVE BLOOD CULTURE AND BODY FLUID SAMPLES DYNAMIC CULTURAL ORIENTATION SYSTEM BASED ON ANALYSIS TECHNICAL AREA 5 The invention relates to Gram staining and positive blood cultures and direct body fluid samples. Morphological analysis results are automatically processed using artificial intelligence-based image processing algorithms. classifying and using this classification data for dynamic medium selection, robotic inoculation guidance. and an integrated microbiological system that uses closed-loop decision input for laboratory management. It includes decision support and automation systems. 10 The invention specifically applies to positive blood culture vials or directly from clinical microbiology laboratories. automated or semi-automated processing of microbiological preparations obtained from body fluid samples It begins with a staining module that enables the Gram staining process. The stained preparations are digitally processed via a high-resolution imaging unit. These images are converted into visual data, and these images are then processed using deep learning and machine learning. It is processed by an AI-based analysis module that includes algorithms. The analysis module includes Gram-positive / Gram-negative differentiation and morphological classification of microorganisms. automatically performs density analysis and potential contamination assessment. It classifies the data obtained. This classification data is used by the system's decision support algorithm. 20 It is evaluated by and includes information on microorganism type, morphological structure and density. Dynamic selection of the most suitable culture medium and culture environment based on the specific sample. The system provides alternatives to the use of fixed and standard culture media for Gram-positive samples. Selective media, different media for Gram-negative samples and for samples containing mixed flora. A medium selection module capable of dynamically selecting from a wide spectrum of culture media. 25 It includes information about the selected culture medium, which is then integrated into the system by automated seeding robots and culture devices. The data is transmitted via the integration interface, thus automatically adding, for example, the data to the relevant culture medium. The planting and cultivation process is guided. All these processes include data collection, A centralized system that manages analysis, decision-making, guidance, recording, and reporting processes end-to-end. It is controlled by a software-based management system, and the system is currently in laboratory 30. It is configured to work integrated with automation systems and blood culture devices. PREVIOUS TECHNIQUE In clinical microbiology laboratories, from positive blood cultures or direct body fluid samples The process of identifying the obtained microorganisms and directing them to the appropriate culture medium, 35 It plays a critical role in the management of sepsis and other invasive infections. Current 2 In practice, this process typically involves the detection of a positive signal using automated blood culture devices. then manual or semi-automatic Gram staining, microscopic images are obtained by expert personnel. Interpretation by [the relevant authority], and based on this interpretation, manual or robotic application to standard culture media. This process includes the stages of inoculation and ultimately incubation in culture devices. In this flow, Gram Interpretation of staining results, morphological classification and density assessment 5 This depends entirely on the operator's visual perception, experience, and subjective interpretation. This leads to high levels of inter-observer variability and potential interpretation errors. This opens up the interpretation process, especially during night shifts or intense work hours. Due to limited access to specialists during peak hours, there are delays, which can affect critically ill patients. In these cases, it leads to a prolongation of the diagnostic period and a delay in empirical treatment. 10 In current systems, Gram staining and microscopic evaluation results are only available for reporting purposes. It is used for this purpose, and these results are automatically transferred to subsequent laboratory processes. Integration as a dynamic decision input is not provided. For example, a Grammatical example... If interpreted as positive cocci or Gram-negative bacilli, the following culture stage The choice of culture medium usually follows a fixed and standard protocol, such as Blood Agar or Chocolate 15. A static panel consisting of classic culture media such as agar or eosin methylene blue agar. This approach is used, for example, based on specific microbiological characteristics and potential contamination. risk or the microorganism density in the sample being very low or containing mixed flora It does not provide optimization according to the situation, therefore the current fixed culture medium selection protocols often result in low efficiency or require additional subculture requirements. This increases both the cost and the workload. However, today's automated blood culture devices, automated planting robots, and imaging equipment in their laboratories. various automation solutions are available, such as systems and laboratory information management systems. However, these systems operate on independent and non-integrated platforms, Data exchange and process flow cannot be established between them. This fragmented structure, pre-analytical, analytical 25 and causes a disconnect between the post-analytic and intermediate stages, with manual interventions in the process. this causes interruptions and creates a significant workload for laboratory personnel. It provides this. In the current technique, it starts with a positive blood culture signal and is based on the Gram staining result. It can dynamically adapt to the situation, select a culture medium specific to the example, and automatically inoculate that selection. integrated 30 capable of communicating to its systems and managing the entire process with a closed-loop decision-making mechanism. There is no such system. In this context, the most fundamental technical problems in current technology are: Long diagnostic time, interpretation errors and variability due to human factors, standard Low culture yield due to the use of culture media, the fragmented and manual progress of the process. Due to the high workload in the laboratory, access to specialists is limited during night shifts. and due to the lack of early referral, treatment is particularly difficult in critical situations such as sepsis. 35 The delays that occur are significant. These problems, particularly in bloodstream infections, lead to increased mortality and... Considering the clinical importance of early and accurate diagnosis, which directly affects morbidity rates 3 These are critical deficiencies that require an urgent solution when they are encountered. This invention, as detailed above, All these technical problems described can only be overcome by AI-based Gram-morphology analysis. Not only for classification or reporting purposes, but also for dynamic culture medium selection and inoculation. Closed-loop decision-making for protocol definition, robotic planting guidance, and laboratory management. 5. Solving it by offering an integrated microbiological decision support system that uses it as input. This innovative approach aims to create an automated and interconnected link between the analytical and post-analytical processes. By establishing a dynamic bridge, image acquisition is AI-based, in contrast to existing piecemeal systems. An integrated workflow between classification, dynamic culture medium selection, and automated culture guidance. It ensures the workflow; thus shortening the diagnosis time, minimizing human errors, Increasing standardization, optimizing laboratory efficiency, and improving sepsis management. It aims to improve clinical outcomes by providing the opportunity for early intervention. Patent document CN120816543A describes artificial intelligence for a laboratory automation system. The invention refers to a system based on interaction with the user to obtain input information. The intelligent human-machine interaction module processes the received input information and standardizes it into 15. converting into machine instructions and optimizing the user experience plan via a knowledge base. The intelligent analysis module, which provides suggestions, spatially analyzes the biological experimental equipment in the study area. By sensing its positioning, the robotic arm performs grasping, carrying, and experimentation operations. the experimental process through an information sensing module and a robotic arm that enables it to be carried out It includes an experiment execution module that automatically completes the process. The system offers high flexibility and 20 It offers low operational threshold characteristics. In the present invention, however, this system differs from the previous one. AI-based analysis is used not only for automating the experimental process, but also directly. based on Gram staining and morphological classification results of microbiological samples It is used for dynamic culture medium selection and culture guidance, and is a closed-loop decision-making process. It provides integrated microbiological workflow management with a support mechanism. 25 Patent document CN117491231A describes an AI-powered home experience platform. The invention is described as an air purifier designed to meet the air purification needs in a laboratory. By verifying the air velocity, the air purifier's performance in the laboratory after a specific operating period. dust concentration, formaldehyde concentration, bacteria concentration in the experimental environment and 30 By collecting images of the air purifier filter plate, we can assess the cleaning effectiveness of the air purifier. to evaluate its competence and air purification according to the number of personnel in the experimental environment By analyzing the needs, the system optimizes the cleaning air velocity. The system uses laboratory air... It aims to improve staff comfort by enhancing quality. In the present invention, however, this Unlike the previous system, instead of controlling environmental conditions, it uses direct microbiological diagnosis. 35 Automation of the process and AI-based Gram-morphology analysis for dynamic culture 4 Guidance is provided, and the diagnostic period in clinical microbiology laboratories is shortened. The aim is to shorten the process and minimize human error. Patent document US2025093851A1 describes a highly scalable laboratory and process. The automation platform is being discussed. The invention consists of a side shell, bottom cover, top cover, and power 5. a connection point, an instrument containing upstream and downstream chained connection points It includes an initial module with a housing, and this housing contains circuit boards, sensors, Motors, solenoids, solenoid valves, rotary valves, heating systems, piston pumps, fluid control elements, pneumatic control elements, airflow controllers, vacuum / air pressure valves, air / gas pumps, vacuum pumps, linear motion stages, rotary motion 10 stages, sensor modules, computer vision modules, motion controller modules, distribution modules, actuator modules, linear actuators, grippers, light sensors, fluorescence meters, photomultipliers, camera sensors, microscope heads, spectrometer units, mechanical components. actuators, microplate cover removers, thermal incubation modules, magnetic beads Any component from the group consisting of extraction modules and mixing modules 15 It is structured to accommodate. The system also includes a meta-instrument enclosure and a A structure that allows multiple modules to communicate with each other via the background card. It offers. In the present invention, however, unlike this platform, a modular instrument Instead of integration, specifically Gram staining and morphological analysis of microbiological samples. focusing on analysis, these analysis results are used in dynamic medium selection and robotic inoculation 20 a software workflow that uses this as decision input for guidance and processes the entire process in a closed-loop manner. An integrated decision support system is offered that manages this. Patent document CN119380330A describes a laboratory strain of YOLOv8-based artificial intelligence. The method and system for identification are discussed. The invention relates to the laboratory detection of strains using data 25. data collection, cleaning, labeling, truncation, grouping, and pre-processing of the collected dataset. processing, model training and algorithm tuning of processed data based on YOLOv8, The creation of a basic strain identification inference model based on data obtained after training, and inference is made and after the inference is completed, the recognition results of the strain recognition system are displayed. It includes the steps of automatically displaying and transmitting the data to the business system. The system is advanced 30 The visual model structure and the low level of automation in the laboratory strain identification process are inadequate. to solve the problems of recognition accuracy and low classification and counting efficiency It aims to achieve this. In the present invention, however, unlike this system, it only uses strain identification. not limited to but directly integrating the result of the definition into the culture orientation process, Dynamic culture medium selection based on Gram-positive / negative discrimination and morphological classification 35 a closed-loop decision support mechanism that makes this selection and transmits this choice to robotic planting systems It includes. Patent document CN121304085A describes an intelligent system that combines graphical machine learning. The invention refers to a laboratory management system. It provides data related to laboratory management. storing and managing job databases, process databases, device status databases, and infographic databases. a data layer, a machine learning engine, and a reinforcement learning engine included in 5 and an AI middleware that connects to the data layer, providing the services of the AI ​​middleware. An application layer that performs laboratory job functions by making calls, and different roles It includes a user layer that provides an interactive interface for users. The system, artificial intelligence... integrating intellectual ability into the laboratory management process, integrating multiple data sources. and breaking down information islands through mining, optimizing resource allocation through dynamic planning 10 It aims to provide early risk identification and intelligent control through active risk forecasting. In contrast to this system, the present invention focuses specifically on laboratory management instead of general laboratory management. focusing on the microbiological diagnostic process, artificial intelligence based on Gram staining and morphology analysis. Pre- an integrated 15 that combines analytical, post-analytical processes in a single closed-loop flow. It offers a decision support system. Patent document number US20260031232A1 describes automated preparation, analysis, and testing for suspected cancer. The text describes an AI-based method and device for scanning. The invention... by automating traditional laboratory processes and the analysis of urine samples and 20 By making it smarter, fully automated preparation of raw samples, automated data collection, and digital imaging are possible. Automated analysis of data using recognition and artificial intelligence based on this data analysis. It includes a urine diagnostic device or system for automated cancer screening. The system detects abnormalities. When no cells are detected, it automatically generates a negative report or rejects suspicious samples. It refers the urine to pathologist examination. In the present invention, however, unlike this system, the urine 25 Instead of analysis and cancer screening, microorganisms are detected in blood cultures and body fluid samples. It focuses on classification, using artificial intelligence analysis results not only for reporting purposes, also as decision input for dynamic culture medium selection and robotic inoculation guidance. It utilizes and enables the full automation of the microbiological culture process. Patent document CN120295414A describes a laboratory based on an artificial intelligence model. The invention refers to an automation control system. It is a control system that works in a traditional laboratory environment. challenges in independently regulating temperature and humidity in the technology and personnel Multimodal environmental sensing, personnel, to address issues such as neglecting the impact of activities. Behavioral modeling, intelligent decision and control, and accidental touch prevention verification modules 35 It integrates various sensors with a multi-mode environmental sensing module. It collects data and creates a three-dimensional ambient heat map to determine temperature and humidity in real time. 6 monitoring, employee location and activity using the employee behavior modeling module. By defining its density, it measures its effect on temperature and humidity, enabling intelligent decision-making and control. The module implements mode switching strategies based on personnel presence and detects accidental touches. By using LSTM with the prevention verification module, the operational intent can be identified and false. It prevents operations. In the present invention, however, unlike this system, environmental conditions 5 Instead of control, automated staining, imaging, and artificial intelligence-based processing of microbiological samples. It focuses on classification and dynamic culture management processes, and the clinical diagnostic process. The goal is automation and standardization. Patent document CN120565077A describes an AI-based smart health management system. The system is being discussed. The invention consists of a collection layer, a data layer, and an algorithm layer. It includes the application layer and the security layer, and the aggregation layer is the Internet of Things. data collection, laboratory test data collection, and user input collection processes It is carrying out, and laboratory test data collection is an aid in collecting laboratory test data. Using the device, the user collects the samples that need to be tested and completes the testing process in 15 days. The system accomplishes this by using artificial intelligence algorithms and sensor technology. By integrating three-dimensional data, we create a complete health profile and sample family data. automatic and efficient collection and transportation of waste to the laboratory It provides. In contrast to this system, the present invention provides health management and sample transport. instead, it focuses directly on the in-laboratory microbiological diagnostic process, with 20 AI-based systems. Closed system using Gram-morphology analysis for dynamic medium selection and culture guidance The cycle offers a decision support system. Patent document CN121403322A describes a modular intelligent laboratory system. It is mentioned. The invention consists of a test unit with at least one independent function and test unit 25 It includes a suspended central support system mounted above the area, and this system is suspended. Type central handling robotic arm, in-unit sling type robotic arm, with standardized interface and cover. It includes software and control systems. The suspended type central transport robotic arm is based on a rail system. pre-assembled with a quick-change connector and vertical lifting mechanism. It is equipped with an in-unit hanging robotic arm with a modular design, capable of carrying various experimental tools. It integrates a standardized interface and container with automatic identification markings. It includes software and control systems, a planning algorithm, and a laboratory brain that integrates everything. by performing coordinated planning and task sequencing of multiple robotic arms, also includes experimental process monitoring, data analysis, and virtual space real-time mirroring. It completes its processes. In the present invention, however, unlike this system, the general laboratory 35 Instead of automation and robotic handling, specifically Gram staining of microbiological samples and It focuses on dynamic culture medium selection and culture orientation based on morphological analysis, 7 pre-analytical, analytical and post-analytical processes with an AI-based decision support mechanism. It offers a closed-loop system that integrates various components. Patent document CN121504068A describes an artificial intelligence based on an attention mechanism. The text discusses a dark laboratory asynchronous multi-process collaboration method. The invention is related to Strategy 5. Using a snapshot as the sole source of accuracy, model version fingerprinting, attention to detail. The vector records the threshold value, the affected sample set, and the rollback point, pre-writing and It performs atomic operations between the two systems through validation stages, in case of a collision. Performing recovery operations based on the freezing and recovery point, using a designated intersection point. It uses occupancy estimation and congestion signal as attention bias, total variation 10 With the stability regulation, urgent orders are triggered by the attention difference threshold, time Microrefactoring without tracing the critical path backward within the window and local candidate set. doing so, recording the degree of counterfactual replay and remorse in the control log, retesting the near microloop by progressively mapping uncertainty to the penalty term. triggering and prioritizing the idle time window under the token interface, thus reducing congestion 15 It forms a planning-traffic-quality control closed loop that connects the threshold. Existing In contrast to this method, the invention utilizes laboratory process planning and traffic control instead of... Dynamic culture medium selection based on direct Gram-morphology analysis of microbiological samples and focusing on culture orientation, AI-based classification result, robotic planting. and offers an integrated system that uses it as decision input for workflow management. 20 Patent document number US2022051788A1 describes a design for an in vitro fertilization (IVF) laboratory. methods and systems for dynamic automation of quality control and information management It is stated that the invention selects an image of a biological entity associated with an IVF procedure. Image analysis to identify morphological features of biological entities on the selected image 25 Algorithms are applied to identify the morphological features of the biological entity in the reference image. compares morphological features and creates a similarity score based on this comparison. And based on this score, a computing system determines the IVF procedure action for the biological entity. It includes the device. The system also matches user input with the IVF procedure action. It requests input from a client computing device to determine if there is a mismatch. The current 30 In contrast to this system, the invention includes an IVF laboratory and embryo morphology evaluation. Instead, it focuses on Gram staining and morphological classification of microbiological samples, AI analysis results are used for dynamic culture medium selection, robotic inoculation guidance, and laboratory testing. It offers an integrated system that uses closed-loop decision input for workflow management. 35 Patent document CN121842234A describes an interregional protocol based on adaptive analysis. The text describes laboratory management methods and systems. The invention involves terminals distributed to each region. 8 Monitoring equipment communication data flow through nodes and based on a deep learning model. using the protocol adaptive analysis module, it automatically identifies the protocol and It analyzes and produces standardized data, uploads this data to a cloud platform, cloud global coordinated control engine based on deep reinforcement learning of the platform Dynamically optimized control across regions by analyzing data through 5 developing and implementing strategies and equipment based on time series forecasting models. It includes a system that performs anomaly detection and predictive maintenance. In the present invention, however, In contrast to this system, instead of cross-regional laboratory management and protocol analysis, a single system is used. Dynamic analysis of microbiological samples in the laboratory based on Gram staining and morphology. Focusing on culture medium selection and culture guidance, AI-based decision support 10 a closed-loop mechanism that integrates pre-analytical, analytical, and post-analytical processes The system offers Patent document CN121219076A describes a complete laboratory automation system. It is mentioned. The invention involves complex robotic 15 advanced analytical devices in the clinical laboratory. systems and daily manual maintenance are required, and each analysis method requires separate analysis equipment. while requiring two or more analysis devices, the fact that they contain expensive redundant components with movement, arranged along the X-axis on the workbench surface and under controller control Y a system containing multiple carriers configured to move in two directions along its axis The system offers one or more stations positioned so that all carriers pass perpendicularly through them x 20. It extends along the axis, and the vehicle at each station is under controller control along the X-axis and potentially moving elements on carriers along the Z-axis It is able to function. However, in the present invention, unlike this system, general laboratory automation is employed. and instead of transport systems, specifically Gram staining and morphology of microbiological samples. It focuses on dynamic culture medium selection and culture guidance based on analysis, artificial intelligence 25 The classification based on the results is used as decision input for robotic planting and workflow management. It offers an integrated decision support system that utilizes Patent document CN102075378A describes a GOOSE communication system based on the IEC61850 standard. The simulation method is mentioned. Invention, SCI, file analysis, communication simulation 30 initial configuration, communication simulation example configuration, communication simulation This includes strategy structuring and online communication simulation, as well as XML language analysis. by creating a GOOSE simulation mechanism for the substation automation system GOOSE performs online simulations of communication behaviors and smart energy grids. It offers an intuitive and reliable simulation method for the simulation laboratory. Currently, 35 In contrast to this method, the invention uses clinical simulation instead of energy network communication simulation. Dynamic culture media based on Gram staining and morphology analysis in microbiology laboratories. 9 focusing on selection and cultural orientation, the result of AI-based classification. an integrated system that uses robotic planting and workflow management as decision input. It offers. Patent document CN117409200A describes a rock mineral component 5 based on clustering analysis. and an automated pore segmentation method is mentioned. The invention relates to laboratory samples. By observing minerals in a photograph, you can determine their types in advance based on their mineral properties. Performing quantitative analysis and calculating the percentage of each mineral and pore, pre-processed. Converting a color image to a grayscale image using the K-means algorithm. Clustering the tonal image into K clusters, setting the lower and upper threshold values ​​of grayscale for each cluster to 10. subtract and divide the original grayscale image into K grayscale images, and each grayscale It includes the steps for calculating the percentage of black pixels in a tonal image. In the present invention... However, unlike this method, instead of mineral segmentation of geological samples, microbiological analysis is used. It focuses on Gram staining and morphological classification of samples, using artificial intelligence. By analyzing images, microorganisms are classified as Gram-positive / negative and morphologically, and these 15 an integrated system that uses the results for dynamic culture medium selection and culture guidance It offers. Patent document CN117173685A describes a liquid iron information system based on visual recognition. The matching method and system are discussed. The invention is a liquid iron casting station (Figure 20). The liquid iron crucible number and sampling time identified by the recognition unit are liquid iron. The information is transmitted to the integration and transmission unit, including temperature and crucible data from the laboratory analysis system. By matching the serial number and component information, the liquid iron information is integrated and transmitted to the unit. loading, converter liquid iron casting station visual recognition unit liquid iron crucible Liquid iron information integration and transmission 25, including number, liquid iron weight, and casting time information. transmitting the information to the unit, and this unit transmitting the integrated information to the converter secondary system. The converter enables its use for secondary model calculation. In the present invention, however, this Unlike the system used for monitoring liquid iron in the metallurgical industry, clinical microbiology is used instead. Dynamic medium selection and culture based on Gram staining and morphology analysis in laboratories. focusing on its guidance, the AI-based classification result, robotic planting and 30 It offers an integrated system that uses workflow management as decision input. Patent document CN120572528A describes a flexible composite for laboratory automation. The robot system is being discussed. The invention involves a flexible composite robot and central planning and It includes a control system and a flexible composite robot with a multi-mode navigation sensor. equipped and capable of simultaneous path planning and obstacle avoidance in a laboratory environment. An autonomous mobile platform that uses a positioning and map-generating algorithm; this platform a high-precision collaborative robotic arm that is assembled and integrates joint torque sensors and Quick-change end effector providing a standardized interface for mechanical, electrical, and pneumatic systems. It includes a central planning and control system that enables human-machine interaction. The intelligent human-machine interaction system breaks down the digital experience plan into executable task sequences. Experiment process separator, multi-intelligent task 5 that distributes tasks and performs mobile path planning. planning and distribution engine and anomalies based on machine vision and force sensor data. an anomaly detection and autonomous recovery module that detects and triggers recovery strategies It includes. In contrast to this system, the present invention involves general laboratory robotics. instead of automation, specifically Gram staining and morphology analysis of microbiological samples. It focuses on dynamic culture medium selection and culture guidance based on artificial intelligence, 10 a closed-loop system integrating robotic planting and workflow management with a decision support mechanism The system offers Patent document number US2023405579A1 describes the use of laboratory material movement. The text discusses laboratory automation. The invention involves multiple laboratory equipment. components and multiple process heads configured to interact with these components It includes a system that facilitates parallel laboratory operations. The system also includes: It is connected to multiple processing heads and moves these processing heads along the first directional axis. The first set of actuators is configured to operate in such a way as to provide at least the laboratory equipment components. The second actuator is configured to move along the second and third directional axes. 20 It includes the first set, the second actuator set, and the second and third sets of laboratory equipment components. one or more magnetic levitation systems that allow movement along directional axes. It may include a system. However, in the present invention, unlike this system, a general laboratory system is used. Instead of material handling and magnetic levitation systems, specifically microbiological Dynamic medium selection and culture based on Gram staining and morphology analysis of samples 25 It focuses on guiding pre-analytical, AI-based decision support mechanisms. an integrated system that combines analytical and post-analytical processes in a single closed-loop flow It offers. Patent document number CN119398461A describes an intelligent laboratory and automated laboratory sample processing system. The distribution system is discussed. The invention involves control of experimental equipment and real-time data. analysis, machine learning optimization, and automated experimental methodology updates by combining them, we perform dynamic optimization of experimental steps and parameters, It automatically obtains and updates experimental method information, deep learning and self-directed learning. Using attention mechanisms, they present suggestions for experimental designs and intelligently utilize experimental equipment. 35 planning and adjusting experimental parameters based on real-time feedback, thus improving the laboratory automation management level and optimizing experimental efficiency. 11 It aims to reduce human intervention and improve efficiency. In the present invention, however, this system... In contrast, instead of general laboratory experiment optimization, it focuses specifically on microbiological aspects. Dynamic medium selection and culture based on Gram staining and morphology analysis of samples. focusing on its guidance, the AI-based classification result, robotic planting and It offers a closed-loop integrated system that uses workflow management as decision input. 5 Patent document number CN218926829U describes mechanical rolling for a conical specimen piece. The device in question is related to the invention, which belongs to the field of surface modification and involves a conical specimen. To overcome the difficulties in processing the parts with existing laboratory equipment, a dual Cylinder hydraulic device, double parallel mechanism, fixed frame, image recognition system and control 10 It presents a device consisting of a system. Image recognition is used while processing the conical sample piece. The system first defines the shape and position of the workpiece and then provides precise positioning. in doing so, then using control system automation technology, the dual parallel mechanism, Static pressure perpendicular to the workpiece surface of the rolling head on the twin-cylinder hydraulic device. to be adjusted to apply and with a double parallel mechanism on the fixed frame and double 15 Cylinder hydraulic device, machine tool guide rail together with machine tool movable plate. It performs hydraulic processing of the conical sample surface by moving along its length. In the present invention, however, unlike this device, instead of mechanical processing and surface modification, Dynamic methods based on Gram staining and morphology analysis in clinical microbiology laboratories. It focuses on culture medium selection and culture guidance, and uses AI-based image analysis to achieve 20 By classifying microorganisms, this result can be used for decision-making in robotic seeding and workflow management. It offers an integrated system that uses [this] as input. The investigations revealed positive blood cultures and direct body image tests in clinical microbiology laboratories. Existing systems for processing fluid samples, automated blood culture devices, 25 Manual or semi-automatic Gram staining stations, microscopic imaging units, various automation solutions such as robotic planting systems and laboratory information management systems Despite including these components, they exist on independent and unintegrated platforms. Because data exchange and process flow could not be established between them, pre-analytic, analytical and A disconnect appears to occur between the post-analytic stages. In current systems, positive 30 The process begins with blood culture signals and continues with the interpretation of Gram staining results and morphological analysis. Classification and density assessment depend entirely on the operator's visual perception, experience, and This depends on subjective interpretation, which leads to a high degree of inter-observer variability and This can lead to potential interpretation errors, especially during night shifts or intense work hours. Limited access to specialists during these hours prolongs the diagnostic process and empirical 35 This leads to delays in treatment. Furthermore, Gram staining and current approaches... The results of the microscopic evaluation are used for reporting purposes only, and these 12 the results as automated and dynamic decision input for subsequent laboratory processes integration is not ensured; instead, the culture medium to be selected for the subsequent culture stage is chosen. It usually follows a fixed and standard protocol and uses blood agar, chocolate agar or eosin methylene blue. This approach uses a static panel consisting of classic culture media such as blue agar. for example, specific microbiological characteristics, potential risk of contamination, or microorganisms in the sample 5 It does not provide an improvement depending on whether the density is very low or whether it contains mixed flora, Therefore, existing fixed culture medium selection methods are often inefficient. this results in or leads to additional subculture requirements, increasing both cost and workload. It is increasing. However, the majority of current laboratory automation solutions, Physical handling of samples, movement of laboratory equipment with robotic arms, temperature 10 and focuses on general laboratory workflow automation such as humidity control, microbiological Decision-making based on Gram staining and morphology analysis, which are key determinants of the diagnostic process. It does not include a mechanism, therefore current systems rely on image analysis and artificial intelligence. Closed loop between classification, dynamic culture medium selection, and automated culture guidance. It is unable to provide an integration. This invention encompasses all of these 15 described in detail above. technical problems, positive blood culture or direct body fluid sample Automatic or semi-automatic Gram staining of microbiological preparations, Digital imaging of stained preparations via a high-resolution imaging unit. converting these images and processing these images using deep learning and machine learning processed by an AI-based analysis module that includes algorithms, 20 Differentiation of microorganisms between Gram-positive and Gram-negative, morphological classification, density. Automatic classification in terms of analysis and possible infection assessment, the obtained This classification data is evaluated by the system's decision support algorithm. Based on information regarding microorganism type, morphological structure, and density, the most suitable sample-specific analysis is performed. Dynamic selection of suitable culture medium and culture environment, system 25 information on selected culture medium. the transmission of information to automated planting robots and cultivation devices via the integration interface and all these processes include data collection, analysis, decision-making, guidance, recording, and reporting. controlled by a centralized, software-based workflow management system that manages end-to-end processes. This solves the problem by reducing diagnostic time and minimizing human errors. Increasing standardization, improving laboratory efficiency, and early sepsis management in the 30s It aims to improve clinical outcomes by offering intervention opportunities. Ultimately, the problems mentioned above, which cannot be solved with the current technology, are related to the technical aspects. This has made it necessary to make an innovation in the field. 35 A BRIEF DESCRIPTION OF THE INVENTION 13 The present invention is a positive development designed to eliminate the aforementioned technical shortcomings. Dynamic culture based on gram-morphology analysis in blood culture and body fluid samples. It is related to the routing system. The main purpose of the invention is to improve the performance of positive blood cultures and direct body samples in clinical microbiology laboratories. During the processing of liquid samples, the existing systems are independent and integrated. 5 Due to its non-existent structure, Gram staining results are used solely for reporting purposes. resulting process disruptions, long diagnostic time, human-induced interpretation errors, and standards. The aim is to offer a fundamental solution to the problems of low culture productivity associated with the use of culture media. This Thanks to the automatic or semi-automatic Gram staining module, high-resolution imaging is possible. Unit 10 is an AI-based analysis unit that includes deep learning and machine learning algorithms. The module works in an integrated manner with the decision support algorithm and the dynamic culture medium selection module. Pre-analytical, analytical, and post-analytical processes are combined in a single closed-loop flow for diagnosis. The time required is shortened and the speed of clinical decision-making is increased. Another purpose of the invention is to make Gram staining and morphological analysis results available for reporting purposes only. not only, but also decision input for dynamic culture medium selection and automated culture guidance 15 The aim is to enable its use as either Gram-positive or Gram-negative in current systems. For an interpreted sample, the culture medium to be selected for the subsequent culture stage must be fixed and standardized. Because it depends on the protocol, for example, an optimization based on specific microbiological characteristics. While this was not possible before, this invention makes it possible to obtain information based on microorganism type, morphological structure, and density. It is possible to dynamically select the most suitable culture medium and culture environment for each specific example. 20 This is done in order to increase cultural productivity and to meet the additional subculture requirements. Costs and workload are reduced. Another aim of the invention is to reduce human-induced variability and the risk of error in the microbiological diagnostic process. The goal is to increase standardization by minimizing the use of Gram staining in current systems. Interpretation of results, morphological classification and density assessment are entirely 25 It depends highly on the operator's visual perception, experience, and subjective interpretation. This discovery leads to inter-observer variability and potential interpretation errors. Thanks to the image-based artificial intelligence analysis module, microorganisms can be identified as Gram-positive or Gram-negative discrimination, morphological classification, density analysis, and potential contamination. In terms of evaluation, it is automatically and objectively classified, so the operator 30 Independent, repeatable, and reliable results are obtained. Another purpose of the invention is to enable specialists, especially during night shifts or intense working hours. diagnostic delays resulting from limited access and the resulting empirical issues The aim is to eliminate the problem of delayed treatment. In current practices, during night shifts... 14 Due to the lack of expert personnel, samples that yield a positive blood culture signal are not processed using Gram statutes. While painting and finishing processes were postponed until the morning hours, this invention enables automatic painting. Imaging, AI-based classification, and dynamic culture medium selection processes are performed by personnel. It can operate independently for twenty-four hours a day, thus providing particularly critical conditions such as sepsis. Early diagnosis and treatment in these situations significantly improve clinical outcomes. It is being improved to a certain extent. Another aim of the invention is to improve existing laboratory automation systems and blood culture equipment. By offering a compatible and integrated system, it can easily integrate with the existing infrastructure of laboratories. The aim is to ensure its adaptability. In this way, hospitals and reference microbiology departments will be able to adapt it. 10 by integrating this system without having to replace their existing equipment They can automate their processes, thus avoiding high investment costs. Operational efficiency is being increased. Another aim of the invention is to reduce the high workload in microbiology laboratories by improving the working environment for staff. The goal is to increase efficiency. Current systems involve fragmented and manual processes. While it creates a significant workload for laboratory personnel, this invention facilitates data collection and analysis, 15 The entire decision-making, guidance, recording, and reporting process is managed through a centralized software-based system. The workflow is automatically managed end-to-end by the management system, thus ensuring the safety of personnel. By reducing manual intervention, it allows the system to focus on more valuable analytical tasks. Another aim of the invention is to collect all the data obtained in the clinical microbiology diagnostic process. By recording and monitoring and creating a feedback mechanism, the system enables adaptive learning. 20 The goal is to enable artificial intelligence to improve its performance over time through its capabilities. In this way, artificial intelligence... By continuously training the models with new data, classification accuracy and decision support are improved. The success of algorithms is steadily improving, thus becoming more accurate and efficient in the long run. A diagnostic system is being obtained. Another aim of the invention is to provide closed-loop Gram-morphology analysis, which is not available in current techniques. By offering a combination of dynamic culture guidance based on microbiological diagnostics, The aim is to provide a holistic solution. The Grammatical state and morphological structure obtained through artificial intelligence... Dynamic culture medium selection, inoculation protocol based on density and potential contamination analysis results. workflow with identification, robotic planting guidance, culture apparatus and laboratory automation system. Thanks to its automatic transfer as decision input to management, 30 of the existing fragmented systems In contrast, image analysis, AI-based classification, dynamic culture medium selection, and automated processes... A seamless and integrated workflow is being created between culture referrals, thus preventing sepsis. By providing the opportunity for early intervention in critical situations such as these, patient safety and treatment success are ensured. is being increased. All the purposes mentioned above and those that will emerge from the detailed explanation below. The present invention aims to achieve Gram-positive results in blood cultures and direct body fluid samples. An AI-powered integrated system that provides dynamic culture guidance based on morphology analysis. It is related to a microbiological decision support system. This system analyzes microbiological samples. A Sample Input Unit that enables the sample to be taken into the system, 5 to the aforementioned sample input unit depending on whether the samples undergo Gram staining or equivalent differential staining. An Automatic Painting Unit that carries out the aforementioned automatic painting working in conjunction with the unit and obtaining digital images of the stained samples an Imaging Unit, digital images obtained from the aforementioned imaging unit It operates using deep learning and machine learning algorithms and processes microorganisms. Gram-positive or Gram-negative differentiation, morphological classification, density analysis, and possible An AI analyzer that automatically classifies substances based on contamination assessment. The unit evaluates the classification results of the aforementioned artificial intelligence analysis unit. Process guidance based on microorganism type, morphological structure, and density information. a Decision Support Unit that generates the decision, 15 depending on the decision generated by the aforementioned decision support unit and instead of using fixed culture media, for example, Gram-negative cultures are used according to their microbiological properties. Selective media for positive samples, different media for Gram-negative samples, and mixed flora. Dynamic is a system that performs dynamic selection from a wide spectrum of culture media for samples containing various ingredients. The Culture Medium Selection Unit selects the culture medium using the aforementioned dynamic culture medium selection unit. by receiving the information and transmitting it to automated planting robots and culture devices, and the existing laboratory 20 A Data Communication Interface that enables data exchange with systems; the aforementioned data communication interface According to the culture medium information transmitted, for example, it can be automatically added to the relevant culture medium. A Robotic Planting Guidance Unit that provides data collection, analysis, and decision-making, managing end-to-end routing, registration, and reporting processes for all the aforementioned units. A Process Management System that ensures coordinated operation, all 25 obtained throughout the process A Data Storage and Monitoring System that stores and tracks data, operator control, and system. A User Interface that enables monitoring processes is connected to the aforementioned artificial intelligence analysis unit. It operates as a system that is continuously trained with new data to improve classification accuracy and decision-making. An Adaptive Learning Model that improves the success of the support unit over time, patient A Clinical Data Integration Unit that incorporates data from 30 different laboratories into the decision-making process. A Flexible Process Management Unit that adapts to various scenarios, and the aforementioned artificial intelligence. Gram status, morphological structure, density, and possible contamination obtained from the analysis unit. the analysis results mentioned dynamic culture medium selection unit, mentioned robotic inoculation routing unit, the aforementioned data communication interface and the aforementioned process management system Dynamic Culture Guidance 35, which enables the transfer of information as automatic decision input. It includes the mechanism. 16 The best way to utilize the advantages of the existing invention, together with its structure and additional elements. For it to be understood, it must be considered together with the figures explained below. BRIEF DESCRIPTION OF THE FIGURES 5 Figure 1: Dynamics based on gram-morphology analysis in positive blood culture and body fluid samples. This is a schematic view of the cultural orientation system. REFERENCE NUMBERS 1. Sample Input Unit 10 2. Automatic Painting Unit 3. Display Unit 4. Artificial Intelligence Analysis Unit 5. Decision Support Unit 6. Dynamic Culture Medium Selection Unit 15 7. Data Communication Interface 8. Robotic Planting Guidance Unit 9. Process Management System 10. Data Storage and Monitoring System 11. User Interface 20 12. Adaptive Learning Model 13. Clinical Data Integration Unit 14. Flexible Process Management Unit 15. Dynamic Culture Guidance Mechanism DETAILED DESCRIPTION OF THE INVENTION This detailed explanation focuses solely on the innovation in the invention to provide a better understanding of the subject matter. This is conveyed without being limited to examples. Accordingly, in the following explanations and figures, Dynamic culture based on gram-morphology analysis in positive blood culture and body fluid samples. The routing system is explained. 30 Figure 1 shows the dynamics based on gram-morphology analysis in positive blood culture and body fluid samples. This is a schematic view of the cultural orientation system. Accordingly, the system: • A Sample Entry Unit (1) that enables the entry of microbiological samples into the system. 17 • Working in conjunction with the mentioned sample input unit (1) and Gram staining or An Automated Painting system that performs an equivalent differential painting process. Unit (2), • Working in conjunction with the aforementioned automatic staining unit (2) and digitally staining samples An Imaging Unit (3) that obtains its images, 5 • Digital images obtained from the mentioned imaging unit (3) can be processed using deep learning and It uses machine learning algorithms to process and identify microorganisms that are Gram-positive or Gram-negative discrimination, morphological classification, density analysis, and potential contamination. An Artificial Intelligence Analysis Unit (4) that automatically classifies in terms of evaluation, • By evaluating the classification results of the mentioned artificial intelligence analysis unit (4), 10 Process guidance based on microorganism type, morphological structure, and density information. A Decision Support Unit (5) that produces the decision, • Working based on the decision produced by the mentioned decision support unit (5) and fixed culture medium instead of using, for example, selective testing for Gram-positive samples based on their microbiological characteristics. The media are varied, with different media for Gram-negative samples and for samples containing mixed flora. 15 A Dynamic Medium Selection Unit (6) that makes dynamic selection among spectral culture media, • By receiving the information about the culture medium selected by the mentioned dynamic culture medium selection unit (6). transmitting data to automated planting robots and culture devices, and to existing laboratory systems. A Data Communication Interface (7) that enables shopping, • According to the culture medium information transmitted via the mentioned data communication interface (7), for example, the relevant 20 A Robotic Planting Guidance Unit (8) that enables automatic planting into the culture medium. • End-to-end data collection, analysis, decision-making, guidance, recording, and reporting processes. A Process Management system that manages and ensures the coordinated operation of all the aforementioned units. System (9), • A Data Storage and Monitoring 25 that stores and tracks all data collected throughout the process. System (10), • A User Interface (11) that enables operator control and system monitoring operations. • Working in conjunction with the mentioned artificial intelligence analysis unit (4) and constantly updated with new data by training the classification accuracy and the success of the decision support unit (5) over time An Adaptive Learning Model that enhances (12), 30 18 • A Clinical Data Integration Unit that includes patient data in the decision-making process (13), • A Flexible Process Management Unit (14) that adapts to different laboratory scenarios and • Gram status, morphological structure, obtained from the mentioned artificial intelligence analysis unit (4), The results of the density and potential contamination analyses are relevant to the dynamic culture medium selection mentioned. to the unit (6), to the aforementioned robotic planting guidance unit (8), to the aforementioned data communication 5 as automatic decision input to the interface (7) and the mentioned process management system (9) A Dynamic Culture Guidance Mechanism that enables its transfer (15) It includes. The system described in the invention involves the sampling of positive blood cultures or direct body fluid samples. It is started by being included in the system via the unit (1) and its working principle is image-based 10 artificial intelligence analysis, dynamic decision-making and automated process guidance mechanisms It is based on an integrated and closed-loop operation. The sample taken into the system is Sample Input. The sample Gram is transmitted from Unit (1) to Automatic Dyeing Unit (2) and in this unit It undergoes painting or an equivalent differential painting process. The painting process The completed sample is sent by the Automatic Painting Unit (2) to the Imaging Unit (3) 15 High-resolution digital images of the painted sample are transferred and obtained in this unit. The obtained digital images are processed by the Imaging Unit (3) using Artificial Intelligence. The data is sent to the Analysis Unit (4) and in this unit deep learning and machine learning algorithms are processed. Images are processed using the Artificial Intelligence Analysis Unit (4), which processes the images. Distinguishing between Gram-positive and Gram-negative microorganisms, morphological classification, 20 automatically performs density analysis and potential contamination assessment. classify and send the results of this classification to the Decision Support Unit (5). It sends. The Decision Support Unit (5) receives from the aforementioned artificial intelligence analysis unit (4). classification results, i.e., information on microorganism type, morphological structure, and density. By evaluating the specific case, it produces the most appropriate course of action decision, and these 25 The decision is communicated to the Dynamic Medium Selection Unit (6). Dynamic Medium Selection Unit (6) operates based on the decision it receives from the mentioned decision support unit (5) and is fixed. Instead of using culture media, selective media are used, for example, in cases where the animal is classified as Gram-positive. When classified as Gram-negative, it is associated with diverse environments and mixed flora. Dynamically selecting the most suitable culture medium from a wide spectrum of culture media. 30 The information about the culture medium selected by the Dynamic Culture Medium Selection Unit (6) is recorded in the Data. It is transmitted to the automated planting robots and culture devices via the Communication Interface (7) and Data Communication Interface (7) also exchanges data with existing laboratory systems By providing this, it enables the system to work integrated with the existing infrastructure. Data Communication According to the culture medium information transmitted via the interface (7), Robotic Planting Guidance Unit 35 19 (8) intervenes and ensures that, for example, it is automatically inoculated into the relevant culture medium. At this point, the Robotic Planting Guidance Unit (8) uses the aforementioned data communication interface (7) Based on the instructions received, the sample is transferred to the selected culture medium and the culture process is initiated. It initiates. Throughout this entire process, data collection, analysis, decision making, guidance, and recording take place. and reporting processes are managed end-to-end by the Process Management System (9) and 5 The coordinated work of all the units mentioned is ensured. All the results obtained throughout the process data is stored and monitored by the Data Storage and Monitoring System (10), This allows the system to conduct retrospective analysis and performance evaluation. The operator can control the system via the User Interface (11) and It can monitor the process in real time. The system's performance is based on Adaptive Learning 10. Thanks to the model (12), it is constantly being improved and the aforementioned adaptive learning The model (12) operates depending on the mentioned artificial intelligence analysis unit (4) and is constantly By training with new data, the classification accuracy and the aforementioned decision support unit (5) It increases its success over time. In addition, the Clinical Data Integration Unit (13) patients By incorporating this data into the decision-making process, more accurate referral decisions can be made in a clinical context. 15 It contributes to its production. The Flexible Process Management Unit (14) is different laboratories. by adapting to various scenarios, the system operates flexibly under different operating conditions. It enables processing. Ultimately, obtained by the Artificial Intelligence Analysis Unit (4) Gram status, morphological structure, density, and possible contamination analysis were performed. The results were obtained through the Dynamic Culture Orientation System (15) using the aforementioned dynamic medium 20 to the selection unit (6), to the aforementioned robotic planting guidance unit (8), to the aforementioned data communication as automatic decision input to the interface (7) and the mentioned process management system (9) This data is transferred, thus enabling image analysis, AI-based classification, and dynamic culture media. a seamless and closed-loop workflow between selection and automated culture guidance This ensures the full automation of the microbiological diagnostic process. 25

Claims

REQUESTS 1. Dynamics based on gram-morphology analysis in positive blood culture and body fluid samples. It is a cultural orientation system, and its characteristic feature is... • A Sample Entry Unit (1) that enables the entry of microbiological samples into the system. • Working in conjunction with the mentioned sample input unit (1) and processing samples in Gram 5 performing dyeing or equivalent differential dyeing process an Automatic Painting Unit (2), • Working in conjunction with the aforementioned automatic staining unit (2) and the stained samples An Imaging Unit (3) that obtains digital images, • The digital images obtained from the mentioned imaging unit (3) are processed using deep learning and 10 It operates using machine learning algorithms and targets Gram-positive microorganisms. or Gram-negative discrimination, morphological classification, density analysis, and possible An AI that automatically classifies based on contamination assessment. Analysis Unit (4), • By evaluating the classification results of the mentioned artificial intelligence analysis unit (4), 15 Processing based on information about microorganism type, morphological structure, and density. A Decision Support Unit (5) that generates a steering decision, • It operates based on the decision produced by the mentioned decision support unit (5) and is fixed. For example, Gram-positive samples can be sampled according to their microbiological characteristics, instead of using culture media. Selective media for Gram-negative samples, different media for Gram-negative samples, and 20 media containing mixed flora. Dynamic is a system that performs dynamic selection from a wide spectrum of culture media for samples. Culture Medium Selection Unit (6), • By receiving the information about the culture medium selected by the mentioned dynamic culture medium selection unit (6). transmitting data to automated planting robots and culture devices, and to existing laboratory systems. A Data Communication Interface (7) that enables data exchange, 25 • According to the culture medium information transmitted via the mentioned data communication interface (7), for example, the relevant A Robotic Planting Guidance Unit that enables automatic planting into the culture medium. (8), • End-to-end data collection, analysis, decision-making, guidance, recording, and reporting processes. A Process Management 30 that manages and ensures the coordinated operation of all the units mentioned. System (9), 21 • A Data Storage system that stores and monitors all data collected throughout the process. Monitoring System (10), • A User Interface (11) that enables operator control and system monitoring operations. • Working in conjunction with the mentioned artificial intelligence analysis unit (4) and constantly developing new By training with data, the accuracy of classification and the success of the decision support unit (5) were improved. An Adaptive Learning Model that increases over time (12), • A Clinical Data Integration Unit that includes patient data in the decision-making process (13), • A Flexible Process Management Unit that adapts to different laboratory scenarios. (14) and • Gram status, morphological structure, 10 obtained from the mentioned artificial intelligence analysis unit (4) The results of the density and potential contamination analyses are relevant to the dynamic culture medium selection mentioned. to the unit (6), to the aforementioned robotic planting guidance unit (8), to the aforementioned data communication as automatic decision input to the interface (7) and the mentioned process management system (9) A Dynamic Culture Guidance Mechanism that enables its transfer (15) It is characterized by its inclusion. 15 25

Images