Method for preparing and applying umbilical cord blood mesenchymal mononuclear cells for treating urinary incontinence

Abstract

This invention relates to the field of medical technology, and more particularly to a method for preparing and applying umbilical cord blood mesenchymal mononuclear cells for the treatment of urinary incontinence. The method includes: constructing an individualized target area weight map by integrating urodynamic data, image annotation data, and symptom scale records; estimating cell demand; preparing high-purity mesenchymal mononuclear cells through umbilical cord blood collection, density gradient centrifugation, and immunomagnetic bead sorting; obtaining cell batches meeting treatment standards through seeding concentration optimization, multi-generation induction culture, and dual-pathway functional marker detection; and dynamically adjusting the dosing regimen based on functional assessment feedback, combined with execution records of multi-point injection into the bladder wall and intravenous infusion, forming a closed-loop treatment system. This invention achieves full controllability and adaptive optimization from patient-specific diagnosis to cell preparation, individualized dosing, and efficacy feedback, significantly improving the targeting and effectiveness of urinary incontinence treatment.

Description

Technical Field

[0001] This invention relates to the field of medical technology, and in particular to a method for preparing and applying umbilical cord blood mesenchymal mononuclear cells for the treatment of urinary incontinence. Background Technology

[0002] In the field of medical technology, existing protocols for the preparation and application of umbilical cord blood mesenchymal mononuclear cells (MCMCs) for the treatment of urinary incontinence typically rely on symptom scale records and raw urodynamic data for preliminary diagnosis and classification, supplemented by bladder and urethral imaging results. Clinicians then choose treatment pathways such as medication, surgery, or single-cell injection based on experience. The cell preparation stage often employs routine umbilical cord blood collection and centrifugation pretreatment combined with density gradient stratification and adherent culture to obtain mesenchymal cells for local injection into the bladder wall. However, this approach suffers from insufficient granularity in case classification and inadequate labeling of bladder and urethral imaging data. The common problems include low correlation between symptom scale records and the inability to support refined treatment planning for specific target areas; lack of unified modeling for process links such as umbilical cord blood collection and centrifugation pretreatment, density gradient stratification and immunomagnetic sorting; resulting in large quality fluctuations of high-purity umbilical cord blood mesenchymal mononuclear cell structure between different batches and difficulty in coordinating with subsequent treatment design; and low coupling between treatment-level cell dose allocation and multi-point injection in the bladder wall and tail vein infusion, incomplete drug administration process records, and difficulty in using them for quantitative analysis. Existing methods largely rely on empirical case classification and manual judgment of total treatment cell requirements. Clinicians, constrained by limited examination results and established operating procedures, roughly plan the dosage and route of each administration. During follow-up, adjustments are made based on subjective impressions and a few indicators. In the context of long-term treatment of urinary incontinence, which requires balancing treatment safety and cell resource utilization efficiency, there is a lack of traceable evidence for dosage and route adjustments, and the source of efficacy differences between different treatment cycles is difficult to analyze. This makes it difficult to achieve stable, refined control of treatment-level cell dosage allocation and application. Regarding the joint processing of data and conditional links related to the target weight map structure and the closed-loop adjustment structure of combined drug administration, existing technologies generally suffer from shortcomings in data fusion, time synchronization, case classification determination, and treatment course-level cell dose allocation and dosing execution records. These shortcomings include reliance on human experience and inconsistent recording. These shortcomings make it difficult to form a consistent process for continuous acquisition, alignment, determination, and recording in the application scenario of umbilical cord blood mesenchymal mononuclear cell preparation and application for the treatment of urinary incontinence. This process involves data acquisition, alignment processing, classification determination, target modeling, estimation of total therapeutic cell requirements, construction of high-purity cell structures, batch structure management of therapeutic cells, and treatment course-level cell dose allocation and treatment course adjustment records. As a result, the data links and decision links are disconnected during the treatment course-level cell dose allocation and application process, making it difficult to iteratively optimize the treatment strategy. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a method for preparing and applying umbilical cord blood mesenchymal mononuclear cells for the treatment of urinary incontinence, comprising:

[0004] We acquire raw urodynamic data, bladder and urethral image annotation data, and symptom scale records. We perform time-aligned coding, parameter field organization, case classification, and target weight modeling to generate a target weight map structure carrying case classification labels and target weight parameters.

[0005] Based on the target area weight map structure, the total cell requirement for treatment is estimated, umbilical cord blood collection and centrifugation preprocessing are performed, and Ficoll Hypaque density gradient stratification combined with CD105 and CD73 immunomagnetic sorting is used to generate high-purity umbilical cord blood mesenchymal mononuclear cell structures with recorded purity, yield and batch number.

[0006] Based on the high-purity umbilical cord blood mesenchymal mononuclear cell structure, the seeding concentration setting, multi-generation induction culture execution, and CK18 and βⅢ tubulin labeling detection combined with cryopreservation steps were processed to generate a therapeutic cell batch structure that records the dual-pathway labeling achievement status and cryopreservation batch number.

[0007] The system acquires the batch structure of therapeutic cells, performs treatment-level cell dose allocation, records of multi-point injection into the bladder wall and tail vein infusion, and drives protocol adjustment through functional assessment. It then generates a closed-loop adjustment structure for combined dosing, recording the dosing execution timeline and the input requirements for the new round of urinary continence function data, as well as the combined dosing adjustment parameters.

[0008] Furthermore, raw urodynamic data, annotated bladder and urethral imaging data, and symptom scale records included:

[0009] The raw urodynamic data include bladder pressure time series, urethral pressure time series, urine flow rate time series, and filling volume time series. The bladder pressure time series includes the bladder pressure curve during filling, the intravesical pressure curve during voiding, and the abdominal pressure compensation curve. The urethral pressure time series includes the resting urethral pressure distribution curve, the urethral pressure change curve under stress, and the urethral pressure response waveform during the cough test. The urine flow rate time series includes the maximum urine flow rate, the average urine flow rate, and characteristic points of the flow time curve. The filling volume series records the changes in bladder volume at different time points.

[0010] The bladder and urethra image annotation data includes dynamic images of the bladder neck position, continuous frames of the urethra course, and images of the sphincter region, acquired by pelvic floor ultrasound or other medical imaging equipment at multiple time phases including resting state, straining state, and before and after urination. The region of interest boundaries of the bladder outlet region, proximal urethra contour, mid-urethral wall structure, distal urethral junction, and surrounding soft tissue support structures are marked on the images in coordinate form by the annotator.

[0011] The symptom scale records include self-reported answers to the standardized urinary incontinence scale collected via electronic terminals.

[0012] Furthermore, the process of performing time-aligned encoding and parameter field organization also includes:

[0013] The time-aligned encoding process includes establishing a timeline for the raw urodynamic data according to the start and end times of the examination, mapping the acquisition time points of each frame of the image annotation data to the urodynamic timeline, encoding the symptom scale records according to the questionnaire completion date and assessment period label, and handling missing timestamps and abnormal sampling intervals.

[0014] The parameter field processing includes standardizing the naming and unit normalization of peak bladder pressure, resting urethral closure pressure, maximum urinary flow rate, and residual urine volume according to a predefined urinary continence function parameter dictionary; encoding the field for bladder neck displacement distance, urethral angle change, and pelvic floor descent degree; converting urinary incontinence frequency classification, quality of life score, and nocturia frequency; and setting field version identifiers.

[0015] Furthermore, the process of case subtyping and target weight modeling also includes:

[0016] The case classification process includes extracting sphincter function indicators, bladder neck mobility indicators, and urethral closure pressure indicators from the urinary continence function dataset; performing rule matching and classification label encoding according to the classification rule set; and generating urinary continence function classification results containing main type labels and auxiliary labels.

[0017] The target area weight modeling process includes constructing a basic anatomical model of the bladder outlet and the area around the urethra, identifying the bladder neck contour and the urethral lumen centerline through a boundary fitting algorithm, discretizing the area into target area grids, labeling the grids with functional categories based on the urinary control function classification results, and calculating the functional repair weight value for each grid according to the weight rule set.

[0018] Furthermore, the process of estimating the total therapeutic cell requirements, collecting umbilical cord blood, and pre-centrifuging and pre-processing it also includes:

[0019] The total cell requirement estimation process includes sorting and prioritizing each target area according to the functional repair weight in the target area weight map structure, merging multiple target areas that are too close anatomically or highly related in function into the same injection target group, calculating the theoretical total number of cells required by weighted summation based on the number of basic cells injected per injection and the number of basic cells infused per tail vein infusion in the treatment course design parameters and the number of treatment cycles, and introducing process loss prediction coefficient and quality inspection sampling reserve coefficient.

[0020] The cord blood collection and centrifugation pretreatment process includes calculating the required cord blood volume range based on the estimated total cell demand for treatment, generating a cord blood collection demand list, collecting cord blood under the guidance of the cord blood collection execution terminal and monitoring the collection volume and anticoagulation parameters, and performing centrifugation pretreatment steps, including retrieving the pretreatment centrifugation protocol for centrifugation, removing the red blood cell layer and plasma portion while retaining the white membrane layer, performing cell counting and viability assessment, and generating a mononuclear cell suspension data package.

[0021] Furthermore, the process of combining Ficoll Hypaque density gradient stratification with CD105 and CD73 immunomagnetic sorting also includes:

[0022] The Ficoll Hypaque density gradient stratification combined with CD105 and CD73 immunomagnetic sorting process includes reading the sample volume and total number of cells from a mononuclear cell suspension data package, calculating the Ficoll Hypaque density gradient liquid volume and cell suspension volume ratio, performing density gradient stratification centrifugation, identifying and aspirating the white membrane layer for washing, and then performing CD105 and CD73 antibody incubation and magnetic bead sorting, including adding CD105 and CD73 monoclonal antibodies for incubation, adding magnetic beads coupled with secondary antibodies for magnetic labeling, separating CD105 / CD73 double-positive cells using a magnetic sorting device, and evaluating purity and yield.

[0023] Furthermore, the process of setting the inoculation concentration and executing multiple generations of induction culture also includes:

[0024] The inoculation concentration setting process includes reading the CD105 / CD73 double-positive ratio, total number of cells and number of cells per unit volume from the high-purity umbilical cord blood mesenchymal mononuclear cell structure, calculating the target starting number of cells in combination with the treatment course configuration parameters, dynamically fine-tuning the inoculation density according to the basic inoculation density strategy and calculating the inoculation volume and number of inoculated cells.

[0025] The multi-generation induction culture execution process includes establishing an execution plan based on the induction culture input set, monitoring culture environment parameters, performing medium change and passage operations, observing cell morphology and proliferation status through the cell state assessment subsystem, and generating an induction culture process record set by recording the conditions, environmental parameters and operation events of each generation of culture.

[0026] Furthermore, the registration process for the cryopreservation step, which combines CK18 with βⅢ tubulin labeling detection, also includes:

[0027] The registration and processing of CK18 and βⅢ tubulin labeling detection combined with cryopreservation steps includes: identifying sampling time points from the induction culture process record set; collecting cell samples for immunofluorescence or flow cytometry detection; labeling with CK18 and βⅢ tubulin specific antibodies; analyzing the proportion of CK18-positive cells and βⅢ tubulin-positive cells; determining the labeling achievement status by batch screening according to treatment threshold rules; performing cryopreservation resuspension and programmed cooling operations on the labeled cell batches, including resuspension using cryopreservation solution containing DMSO, human serum albumin, and MEM; cooling according to the programmed cooling protocol; and registering the cryopreservation batch number and cryopreservation location.

[0028] Furthermore, the process of recording the execution of treatment-level cell dose allocation, multi-point injection into the bladder wall, and tail vein infusion also includes:

[0029] The treatment-level cell dose allocation process includes reading the total number of available cells and the recommended thawed cell density from the therapeutic cell batch structure, allocating the total cell dose for the entire treatment course to each treatment cycle in combination with the treatment course configuration parameters, and further splitting the local injection dose and intravenous infusion dose within each cycle according to the basic dose allocation ratio of multi-point injection in the bladder wall and tail vein infusion.

[0030] The bladder wall multi-point injection and tail vein infusion execution record processing includes extracting injection target parameters and intravenous infusion parameters from the initial combined dosing regimen set, performing bladder wall multi-point injection under the guidance of an ultrasound navigation terminal and recording the number of needle pricks, actual injection volume and event timestamp for each target point, performing tail vein infusion under the control of an infusion pump control device and recording the actual infusion volume, duration and rate changes, and generating a periodic dosing execution record set.

[0031] Furthermore, the process of functional evaluation-driven solution adjustment also includes:

[0032] The functional assessment-driven protocol adjustment process includes aligning the timeline of dosing events with follow-up assessment data from the periodic dosing execution record set, summarizing changes in the number of urinary incontinence occurrences, changes in urodynamic parameters and symptom scale scores, generating combination dosing adjustment parameters and new round of urinary continence function data input requirements according to adjustment decision rules, and forming a closed-loop adjustment structure for combination dosing.

[0033] The key innovations of this invention include:

[0034] (1) Based on the acquisition of raw urodynamic data, bladder and urethral image annotation data and symptom scale records, a target area weight map structure is constructed by time alignment coding, parameter field organization and case classification, and target area weight modeling. Furthermore, the total cell demand for treatment is estimated based on the target area weight map structure. The umbilical cord blood collection and centrifugation preprocessing, as well as the Ficoll Hypaque density gradient stratification combined with CD105 and CD73 immunomagnetic sorting are connected into the same data link to generate a high-purity umbilical cord blood mesenchymal mononuclear cell structure with recorded purity, yield and batch number, realizing the correspondence between case classification results, target area weight parameters and high-purity umbilical cord blood mesenchymal mononuclear cell structure.

[0035] (2) Based on the high-purity umbilical cord blood mesenchymal mononuclear cell structure, the induction culture process record is obtained by setting the inoculation concentration and performing multiple generations of induction culture. In the same process, the CK18 and βⅢ tubulin labeling detection and cryopreservation steps are combined and the dual-pathway labeling detection results, induction culture condition information and cryopreservation batch number are integrated into the therapeutic cell batch structure, so that the therapeutic cell batch structure carries both the dual-pathway labeling target status and the cryopreservation batch number.

[0036] (3) After obtaining the batch structure of therapeutic cells, the treatment course-level cell dose allocation is used to uniformly model the protocol adjustment process driven by the execution record of bladder wall multi-point injection and tail vein infusion and the functional assessment. This forms a combined dosing closed-loop adjustment structure that records the dosing execution time axis and the new round of urinary control function data input requirements and combined dosing adjustment parameters. The dosing execution process and the functional assessment results corresponding to the urodynamic raw data, bladder and urethral imaging annotation data and symptom scale records are incorporated into the same closed-loop control link.

[0037] The following are its main beneficial effects:

[0038] (1) By constructing a target area weight map structure based on raw urodynamic data, bladder and urethral imaging annotation data and symptom scale records, and estimating the total cell requirement for treatment based on this, the high-purity umbilical cord blood mesenchymal mononuclear cell structure output from umbilical cord blood collection and centrifugation preprocessing and Ficoll Hypaque density gradient stratification combined with CD105 and CD73 immunomagnetic sorting is linked with case typing labels and target area weight parameters. Compared with the existing schemes that plan dosage and cell preparation based solely on experience or single examination results, this approach can link the diagnostic typing, target area modeling and high-purity umbilical cord blood mesenchymal mononuclear cell structure construction process within the same parameter system, reducing the disconnect between clinical planning and preparation process, and facilitating the formation of traceable dosage basis and batch management foundation in the treatment of urinary incontinence.

[0039] (2) By simultaneously recording the induction culture conditions during the setting of the inoculation concentration and the execution of multiple generations of induction culture, and combining the detection results of CK18 and β-III tubulin markers with the registration and processing of cryopreservation steps, the dual-pathway marker detection results and the cryopreservation batch number are written into the therapeutic cell batch structure as an integrated field. Compared with the practice in the background technology of only performing routine phenotypic detection or single cryopreservation management on mesenchymal cells without a function-oriented batch structure, this method can reflect the detection status of urothelial-related markers and neural-related markers and the corresponding induction culture and cryopreservation information in the same structure. This allows the therapeutic cell batch structure to not only characterize the cell source and batch, but also carry marker information closely related to the functional pathways of urinary incontinence. This is beneficial for calling the corresponding batch according to the function orientation when allocating and applying cell doses in subsequent treatment courses, and improves the problem in the prior art that the therapeutic cell batch is difficult to correspond to specific indications and induction conditions.

[0040] (3) By implementing course-level cell dose allocation based on the batch structure of therapeutic cells, the protocol adjustment process driven by the execution records of bladder wall multi-point injection and tail vein infusion and functional assessment is uniformly incorporated into the closed-loop adjustment structure of combined administration. Compared with the existing technology that only records single-dose information or makes empirical adjustments based on scattered indicators during follow-up, it can reflect the correlation between the drug administration time axis, course-level cell dose allocation, execution records of bladder wall multi-point injection and tail vein infusion and the new round of urinary continence function data input requirements and combined administration adjustment parameters in the same closed loop. This transforms the application of cells for the treatment of urinary incontinence from a single-course protocol to a repeatedly adjusted link driven by functional assessment results. This is conducive to gradually optimizing the combined administration strategy in the long course of treatment and reducing the situation of scattered treatment course adjustment basis and difficulty in verification in the existing technology. Attached Figure Description

[0041] Figure 1 This is a flowchart illustrating a method for preparing and applying umbilical cord blood mesenchymal mononuclear cells for treating urinary incontinence, as provided in an embodiment of this application. Detailed Implementation

[0042] Example 1: Refer to Figure 1 This is a flowchart illustrating a method for preparing and applying umbilical cord blood mesenchymal mononuclear cells for treating urinary incontinence, provided by an embodiment of the present invention. The process may include at least steps S100-S400:

[0043] S100: Obtain raw urodynamic data, bladder and urethral image annotation data and symptom scale records, perform time-aligned coding, parameter field organization and case classification, target weight modeling, and generate a target weight map structure carrying case classification labels and target weight parameters.

[0044] S200, based on the target area weight map structure, estimates the total cell requirement for treatment, performs umbilical cord blood collection and centrifugation preprocessing, and combines Ficoll Hypaque density gradient stratification with CD105 and CD73 immunomagnetic sorting to generate high-purity umbilical cord blood mesenchymal mononuclear cell structures with recorded purity, yield and batch number.

[0045] S300, based on the high-purity umbilical cord blood mesenchymal mononuclear cell structure, performs seeding concentration setting, multi-generation induction culture execution, and CK18 and β-III tubulin labeling detection combined with cryopreservation steps registration processing to generate a therapeutic cell batch structure that records the dual-pathway labeling achievement status and cryopreservation batch number;

[0046] S400: Obtain the batch structure of therapeutic cells, perform treatment-level cell dose allocation, record execution of bladder wall multi-point injection and tail vein infusion, and process the protocol adjustment driven by functional assessment to generate a closed-loop adjustment structure for combined dosing, which records the dosing execution time axis and the input requirements for the new round of urinary continence function data and the combined dosing adjustment parameters.

[0047] Step S100 includes at least steps S110-S130:

[0048] S110. Obtain raw urodynamic data, bladder and urethral image annotation data and symptom scale records, perform time-aligned encoding processing and parameter field sorting processing to obtain a urinary continence function dataset containing curve parameters, image region markings and symptom score entries.

[0049] The raw urodynamic data includes, but is not limited to, bladder pressure time series, urethral pressure time series, urine flow rate time series, and filling volume time series continuously recorded by pressure and flow sensors at a preset sampling frequency. The bladder pressure time series includes at least the bladder pressure curve during filling, the intravesical pressure curve during voiding, and the abdominal pressure compensation curve. The urethral pressure time series includes at least the resting urethral pressure distribution curve, the urethral pressure change curve under stress, and the urethral pressure response waveform during the cough test. The urine flow rate time series includes the maximum urine flow rate, the average urine flow rate, and characteristic points of the flow-time curve. The filling volume series records changes in bladder volume at different time points. The bladder and urethral imaging annotation data... The system includes dynamic images of the bladder neck position, continuous frames of the urethra, and images of the sphincter region acquired at multiple time points before and after urination using pelvic floor ultrasound or other medical imaging equipment, in a resting state, a straining state, and before and after urination. The system also includes regions of interest (ROIs) marked on the images in coordinate form by annotators, such as the bladder outlet region, proximal urethral contour, mid-urethral wall structure, distal urethral junction, and surrounding soft tissue support structures. The symptom scale records include self-reported answers to a standardized urinary incontinence scale acquired via an electronic terminal, which at least record the grading score of the frequency of urinary incontinence occurrence, the frequency score of urinary urgency, the number of nocturia episodes, and the quality score of the impact on daily life, along with a questionnaire completion timestamp and assessment period identifier.

[0050] In practice, raw urodynamic data is collected by the urodynamic testing subsystem during urodynamic examinations, bladder pressure tests, and urethral pressure tests. This subsystem includes a pressure sensor, a flow sensor, and a data acquisition and control unit, capable of continuously recording time-series data such as bladder pressure, urethral pressure, urine flow rate, and filling volume at a preset sampling frequency. Bladder and urethral image annotation data is acquired by the image acquisition and annotation subsystem. This subsystem may include pelvic floor ultrasound equipment or other medical imaging equipment and a corresponding image annotation terminal. It acquires images of the bladder neck position, urethral course, and sphincter region at multiple time points—resting, straining, and before and after urination. Trained annotators then mark the bladder outlet, proximal urethra, mid-urethra, distal urethra, and regions of interest in the surrounding soft tissues on the images. Symptom scale records are obtained by the questionnaire collection subsystem during outpatient visits or follow-ups. The questionnaire collection subsystem presents standardized urinary incontinence-related scale questions to patients through electronic terminals and records patients' self-reported answers to items such as frequency of urinary incontinence, urgency, and nocturia.

[0051] After the three subsystems complete the raw data acquisition, the data management and integration unit integrates the raw urodynamic data, bladder and urethral image annotation data, and symptom scale records into the same data access queue. When all three types of data for the same patient and assessment period are detected as having reached the completion of reception, the time alignment encoding process is automatically triggered. In the time alignment encoding process, firstly, a timeline is established for the raw urodynamic data according to the start and end times of the examination, and each pressure and flow curve entry is appended with an acquisition timestamp and channel number. Then, the acquisition time points of each frame in the image annotation data are mapped to the urodynamic timeline, and the regions of interest annotated on the images are recorded in coordinate form. Interpolation is used to establish a correspondence between discrete image time points and continuous urodynamic time series. Simultaneously, the symptom scale records are encoded according to the questionnaire completion date, assessment period label, and scale entry number, associating each entry score with the corresponding examination period. During the time alignment process, the data management and integration unit handles missing timestamps, abnormal sampling intervals, or duplicate records according to preset rules. Abnormal records are separately registered in the log center and marked with a flag field in the corresponding data entry for subsequent classification rule skipping or correction.

[0052] In the parameter field processing, the parameter field processing submodule, based on a predefined urinary incontinence function parameter dictionary, standardizes the naming and unit normalization of raw measurements in the urodynamic raw data, such as peak bladder pressure, resting urethral closure pressure, maximum urinary flow rate, and residual urine volume, mapping parameters with the same meaning from different devices to unified fields. It also encodes quantitative indicators in the image annotation data, such as bladder neck displacement distance, urethral angle change, and pelvic floor descent, ensuring consistent field names and value ranges across different examination scenarios. Furthermore, it transforms entries in the symptom scale records, such as urinary incontinence frequency classification, quality of life score, and nocturia frequency, converting text-based answers into structured scoring values. During parameter field processing, a field version identifier is set. When the parameter dictionary is updated, the system automatically adds a new version number to the newly processed fields while retaining the mapping relationships of the old versions, facilitating subsequent traceability and auditing. Finally, the data management and integration unit groups the structured data, which has been time-aligned and coded and organized by parameter fields, by patient, by assessment period, and by examination type, forming a comprehensive data entry set containing curve parameters, image region markers, and symptom score entries, and marks the generation time and version number of the data entry set in the metadata.

[0053] In this embodiment, the comprehensive set of data entries compiled above is defined as the urinary control function dataset. The urinary control function dataset serves as the basic input structure for the process of this invention, stored in a record associated with the patient's unique identifier, and registered in the data directory as the standard input for subsequent steps. After the urinary control function dataset is generated, the data management and integration unit sends a notification that new data is available to the subtyping decision subsystem. Upon receiving the notification and verifying the completeness of the urinary control function dataset, the subtyping decision subsystem automatically calls the urinary control function dataset as the input data source for step S120. This allows the urinary control function dataset output from step S110 to be directly used as the input field of S120, while also providing basic functional data for subsequent treatment pathway planning and combined drug administration regimen design.

[0054] S120. Extract sphincter function indicators, bladder neck mobility indicators, and urethral closure pressure indicators from the urinary continence function dataset, perform classification rule matching and classification label encoding processing, and generate urinary continence function classification results with case classification labels.

[0055] In the specific implementation process, the subtyping decision subsystem first extracts parameter fields related to sphincter function from the urinary continence function dataset generated in step S110. These sphincter function indicators include, but are not limited to, morphological parameters of the resting urethral pressure curve, peak urethral pressure during systole, the degree of urethral pressure change during the cough test, and urethral pressure compensation when abdominal pressure increases. The subtyping decision subsystem reads these parameters into the sphincter function analysis module according to predefined field names and value ranges. The bladder neck mobility index consists of quantitative data such as the bladder neck's downward displacement distance, the positional change of the bladder neck between the exertion and resting states, and the change in distance between the bladder neck and the pubic symphysis from the image annotation data. This quantitative data is extracted from the urinary continence function dataset by the image parameter parsing module of the subtyping decision subsystem according to image frame time points and annotation area coordinates, and converted into standardized length and angle parameters. The urethral closure pressure index consists of the resting urethral closure pressure value, the minimum urethral closure pressure value, and the curve of the closure pressure distribution along the urethral length. The classification decision subsystem parses this information from the corresponding curve parameter fields in the urinary control function dataset and records the corresponding measurement time period and acquisition channel information during the parsing process.

[0056] After extracting the three sets of indicators mentioned above, the classification decision subsystem enters the classification rule matching processing stage. This process is collaboratively completed by the rule base management module and the classification engine module. The rule base management module stores classification rule sets for different types of urinary incontinence. These rule sets include logical condition combinations for identifying types primarily characterized by internal sphincter dysfunction, types primarily characterized by external sphincter or neural pathway damage, and mixed types. Each rule carries a rule version number and an description of the applicable population. When the system detects an update in the rule base, the rule base management module automatically freezes the old version rules and activates the new version, while simultaneously recording the version switch time and responsible person information in the classification log. Upon receiving sphincter function indicators, bladder neck mobility indicators, and urethral closure pressure indicators, the classification engine module calculates whether each indicator meets the corresponding conditions based on the current valid rule set. For each patient, it constructs a matching matrix between indicators and rules, marking the degree of satisfaction and priority within the matrix. If multiple rules within a certain rule group are satisfied, the classification engine module selects the rule based on pre-set priorities and conflict resolution strategies. For example, it prioritizes rules with broader indicator coverage and more sample validation times. If there are conflicts between different classification categories, the classification engine module will make a decision based on hybrid priority rules or manual intervention flags and record the decision basis in the classification log.

[0057] During the typing label encoding stage, the typing decision subsystem converts the urinary continence typing results obtained through rule matching into structured typing labels. Typing label encoding includes generating a primary type label and several auxiliary labels for each patient in the current assessment period. The primary type label identifies whether the patient's main urinary continence dysfunction is primarily internal sphincter insufficiency, primarily external sphincter or neural pathway damage, or a mixed type. Auxiliary labels record specific abnormal characteristics, such as abnormal bladder neck mobility, generally low urethral closure pressure, or abnormality only in specific positions. During the encoding process, the typing decision subsystem simultaneously adds a timestamp, rule set version number, and a summary of key indicators involved in the calculation to the typing results, forming traceable typing metadata. In an optional implementation, the system can also assign a risk level code to each typing result, which is used to set different initial regional weight values ​​for different typing categories during subsequent individualized target area weight modeling. In another implementation, the system can configure a manual review status for the typing results. When the typing results involve boundary values ​​or have multiple rule conflicts, the results are marked as pending manual review and pushed to specialist doctors for confirmation or revision in a dedicated review interface. The final typing label after manual review is also written into the typing metadata.

[0058] Finally, the classification decision subsystem encapsulates the aforementioned structured classification labels and their metadata to form the urinary continence function classification result. A one-to-one association is established between the urinary continence function classification result and the corresponding urinary continence function dataset, and indexed and managed using unique patient identifiers and assessment cycle labels. At the system data flow level, the urinary continence function classification result serves as the input data source for step S130, read by the target area modeling subsystem during the division of the bladder outlet and periurethral regions, and used as the basic classification information for regional functional repair weight allocation. Simultaneously, the urinary continence function classification result is also stored in a long-term follow-up database, providing historical classification references for the generation and closed-loop application of function feedback-driven individualized combination drug administration regimens in subsequent step S400. This ensures that the urinary continence function classification result output in step S120 supports both current target area modeling and adjustments and version management of cross-cycle treatment plans.

[0059] S130. The bladder outlet and urethral peri-region are divided into regions based on the urinary control function classification results, and the functional repair weight of each region is assigned. A target area weight map structure is generated, which records the spatial coordinates and weight parameters of each region.

[0060] In the specific implementation process, the target area modeling subsystem first reads the main type label and auxiliary label information from the urinary control function classification results generated in step S120, and combines it with the image region labeling data in the urinary control function dataset to construct a basic anatomical model of the bladder outlet and the surrounding urethra. The basic anatomical model uses the patient's own image data as a reference, and through the image coordinate normalization module, maps images obtained from different examination devices and at different examination times to a standardized two-dimensional or three-dimensional anatomical coordinate system. The standardized anatomical coordinate system predefines the reference areas for the bladder neck, proximal urethral segment, mid-urethral segment, distal urethral segment, and related surrounding supporting tissues. Based on these reference areas, the target area modeling subsystem automatically fits the boundary lines of the actual bladder outlet and urethra in the standardized images of each patient. Through a boundary fitting algorithm, it identifies the bladder neck contour and the centerline of the urethral lumen, and discretizes the fitted area into several target area grids that can serve as injection planning units. Each target area grid is assigned a unique spatial coordinate identifier.

[0061] In the segmentation of the bladder outlet and periurethral region, the target area modeling subsystem not only categorizes the target area grids according to anatomical location but also incorporates the primary and secondary type labels from the urinary continence function classification results to denote the functional category to which the grids belong. For types primarily characterized by internal sphincter dysfunction, the target area modeling subsystem prioritizes classifying the sphincter ring region and its adjacent urethral wall region as key repair areas. For types primarily characterized by external sphincter or neural pathway damage, the system focuses more on the periurethral soft tissue regions related to neural pathways and supporting structures related to nerve innervation. In this way, target area segmentation not only incorporates spatial location information but also solidifies the functional attributes related to the classification results. The target area modeling subsystem also considers the influence of bladder neck mobility and urethral closure pressure indices on regional sensitivity during the segmentation process. For example, patients with significant bladder neck displacement will have their periurethral grids classified as unstable areas, allowing for different baseline weight ranges to be obtained during the subsequent weight allocation phase.

[0062] In the functional repair weight allocation process for each region, the target area modeling subsystem calls the weight allocation engine. After reading the functional category, local anatomical location, and associated functional indicators of each target area grid, the weight allocation engine calculates the functional repair weight value for each grid according to a pre-defined weight rule set. The weight rule set is formed by combining clinical experience data and follow-up statistical results, and includes regional priority recommendations for different urinary continence function subtypes, weighting coefficients for specific imaging manifestations, and correction coefficients for special risk factors. The weight allocation engine selects the corresponding basic weight curve according to the main type label, and then corrects the basic weight based on auxiliary labels and local indicators to generate the functional repair weight for each grid. To support dynamic adjustments to subsequent multi-cycle combined dosing regimens, the weight allocation engine adds a weight source description and rule version number to each weight value during calculation, facilitating the tracing of the weight formation basis when adjusting the treatment plan. In an optional implementation, the system can also configure upper and lower limit ranges for the target area weights to limit the magnitude of weight changes during subsequent adjustment phases and avoid excessive fluctuations in target area weights due to single assessment biases.

[0063] After weight calculation, the target area modeling subsystem integrates the spatial coordinates, functional category labels, and functional repair weights of all target area grids to generate a target area weight map structure. As a structured data object, the target area weight map structure contains a list of target area grids. Each grid entry records standardized coordinates, the corresponding original image coordinate index, the name of the associated anatomical region, urinary continence function classification information, and the current functional repair weight value. Simultaneously, the target area weight map structure records the generation time, the classification rule version number used, the weight rule version number, and a summary of key parameters involved in the calculation in its metadata section. This metadata is used for version management and auditing. After generation, the target area weight map structure is stored in the treatment planning database and indexed by patient unique identifiers and assessment cycle tags for subsequent steps. In the process of this invention, the target weight map structure is first read in step S210 and used to calculate the total therapeutic cell requirements and cell allocation targets for each target area during the umbilical cord blood mesenchymal mononuclear cell collection and high-purity preparation stage. Simultaneously, the target weight map structure is also invoked in step S410 to determine the injection target distribution and injection weights for multi-point injection into the bladder wall during the functional feedback-driven generation and closed-loop application stage of the personalized combination drug administration regimen. Through the above data flow relationship, the target weight map structure output in step S130 plays a crucial role in both the cell preparation and combination drug administration main steps, and maintains the stability and auditability of the target weight evolution through a version management strategy during multi-cycle treatment.

[0064] In summary, the technical effects of this step are as follows: By constructing a target area weight map structure that is closely coupled with the urinary control function classification results, the individualized functional defect characteristics of patients are mapped to quantifiable spatial regions and functional repair weights. This provides a unified data foundation for subsequent estimation of the scale of high-purity cell preparation and multi-point injection path planning, making the present invention targeted and adjustable in the generation of treatment plans.

[0065] Step S200 includes at least steps S210-S230:

[0066] S210. Obtain the target area weight map structure, perform total treatment cell demand estimation and umbilical cord blood collection batch registration, and obtain an umbilical cord blood collection demand list that records the number of target cells and the collection batch number.

[0067] In practice, this step is collaboratively completed by the cell preparation planning subsystem and the clinical decision support subsystem. The cell preparation planning subsystem acts as the execution entity, while the clinical decision support subsystem provides parameter constraints and version control information. First, the cell preparation planning subsystem retrieves all target area entries associated with the current patient from the target area weight map structure output in step S130. Each target area entry in the target area weight map structure includes at least standardized spatial coordinates, the name of the corresponding anatomical region, information related to the urinary control function classification, and a functional repair weight value. The metadata also includes the generation time, the version number of the classification rule used, and the version number of the weight rule. Upon receiving the target area weight map structure and detecting that the corresponding patient has been marked as entering the cell therapy preparation state, the cell preparation planning subsystem automatically triggers the total treatment cell requirement estimation process and records this status change in the system log, forming a traceable trigger record. The clinical decision support subsystem provides currently applicable treatment design parameters in the background, including the number of treatment cycles, the number of planned bladder wall injections and tail vein infusions per cycle, the minimum number of cells per injection, and the acceptable collection volume range for a single batch of umbilical cord blood. These parameters are considered as the constraint boundaries of the total cell demand estimation process.

[0068] In the estimation of total treatment cell requirements, the cell preparation planning subsystem first sorts each target region according to the functional repair weight in the target region weight map structure and assigns an injection plan priority to each target region. When multiple target regions are anatomically too close or functionally highly related, the system merges them into the same injection target group while retaining their respective weight values ​​to prevent overly dense injection distribution during subsequent multi-point injections into the bladder wall. Subsequently, the cell preparation planning subsystem calculates the theoretically required total number of cells for the entire treatment course by weighted summing based on the preset basic cell count per injection, basic cell count per tail vein infusion, and number of treatment cycles in the treatment course design parameters, combined with the functional repair weight of each target group. On this basis, a process loss prediction coefficient and a quality inspection sampling reservation coefficient are introduced to cover the unavoidable cell loss during subsequent Ficoll Hypaque density gradient stratification and immunomagnetic bead sorting processes, as well as the sample retention required for quality control. The aforementioned basic cell count, loss prediction coefficient, and sampling reserve coefficient constitute the minimum set of core parameters for estimating total treatment cell requirements. Coefficients for fine-tuning cell doses between different treatment cycles and correction factors adjusted according to patient weight or comorbidities can be provided as optional extended parameters by the clinical decision support subsystem through strategy templates, and can be enabled or disabled in specific patient populations.

[0069] Furthermore, after obtaining the estimated total treatment cell requirement, the cell preparation planning subsystem calls the cord blood resource management module to query the current cord blood bank and the source of cord blood to be collected. For cases originating from existing cord blood inventory, the resource management module can directly assess the available cell quantity based on the historical yield records of the inventory units. For cases planned for immediate collection from healthy mothers, this invention focuses more on the latter, i.e., collecting fresh cord blood on demand during childbirth. In the immediate collection scenario, the cell preparation planning subsystem calculates the required cord blood volume range based on the estimated total treatment cell requirement and empirical statistics showing the average yield of mesenchymal mononuclear cells per milliliter of cord blood. It then sets the target volume and minimum volume according to the collection equipment and delivery room operating procedures. The target volume corresponds to the ideal collection volume, and the minimum volume corresponds to the threshold at which the treatment schedule needs to be reassessed if the volume falls below this value. Simultaneously, the system retrieves the currently recommended anticoagulation protocol from the parameter configuration module, including the type of anticoagulant, the target final concentration, and the maximum allowable deviation range, for configuring anticoagulation parameters in subsequent steps.

[0070] In the umbilical cord blood collection batch registration process, the cell preparation planning subsystem generates a unique umbilical cord blood collection batch number for each upcoming collection. This batch number includes the patient identifier, planned treatment start date, collection institution code, and current parameter configuration version number. This code facilitates the association of the source and treatment plan for each batch of cells in subsequent preparation stages. The system summarizes the estimated total treatment cell requirements, target umbilical cord blood volume, minimum umbilical cord blood volume, recommended anticoagulant type and target final concentration, expected collection time window, and umbilical cord blood collection batch number to form a structured data record, which is defined as the umbilical cord blood collection requirement list. This list, as an output field of this step, is available for viewing and printing by obstetric medical staff on the clinical workstation terminal and is simultaneously synchronized to the umbilical cord blood collection execution terminal to guide subsequent collection operations. At the data flow level, this invention registers the umbilical cord blood collection requirement list as the input field of the subsequent step S220, which is automatically read by the collection execution control module when step S220 is started. In addition, the target cell quantity and collection batch number in the umbilical cord blood collection requirement list are also used in step S230 to compare and verify the yield of high-purity umbilical cord blood mesenchymal mononuclear cell structure, thereby forming an auditable link from treatment planning to preparation execution.

[0071] S220. Extract collection volume parameters and anticoagulation parameters from the umbilical cord blood collection requirement list, perform umbilical cord blood collection execution processing and centrifugation pretreatment step control processing, and generate a single-nuclear cell suspension data package that records centrifugation conditions and cell count information.

[0072] Specifically, this step is completed collaboratively by the cord blood collection execution terminal, the centrifugation control subsystem, and the cell counting unit. The cord blood collection execution terminal is typically deployed in the delivery room or operating room to guide medical staff in collecting cord blood at the appropriate time after delivery. Upon receiving the cord blood collection request list from the cell preparation planning subsystem, the cord blood collection execution terminal automatically parses the target cord blood volume, minimum cord blood volume, recommended anticoagulant type, and target final concentration from the list. It also reads the cord blood collection batch number and displays these key parameters in a standard format on the collection process interface. When the monitoring system detects that the corresponding mother's delivery process has entered the appropriate time window for cord blood collection, the system reminds medical staff to perform the collection operation through interface prompts and audio-visual alerts. The triggering conditions and triggering time are recorded in the log as a basis for subsequent auditing.

[0073] During umbilical cord blood collection, medical staff connect a sterile collection bag pre-loaded with anticoagulant to the mother's umbilical vein. The collection terminal calibrates the volume scale on the collection bag according to the target volume recorded in the umbilical cord blood collection requirements list and guides medical staff to check the pre-filled amount of anticoagulant in the collection bag based on the recommended type and target final concentration. These anticoagulant parameters include the specific type of anticoagulant, such as disodium ethylenediaminetetraacetate (EDTA) or citrate anticoagulant, and the target final concentration range of the anticoagulant in the collected sample. The system saves commonly used protocols through pre-configured parameter templates and automatically loads the corresponding template before collection. If the operator adjusts the anticoagulant parameters, the system will require the reason for the adjustment and generate a new parameter version record. During the collection process, the collection terminal monitors the weight change of the collection bag in real time, converts it into the collection volume through a built-in or external electronic scale module, and compares it with the target volume and minimum volume in real time. When the collection volume reaches the target volume or exceeds the safety limit, the system issues a stop collection prompt. When the collected volume falls below the minimum volume threshold and collection cannot continue, the system records this situation as a special event for subsequent adjustments to the treatment plan in the treatment course planning.

[0074] After cord blood collection, this step enters the centrifugation pretreatment control phase. The centrifugation control subsystem retrieves the corresponding pretreatment centrifugation protocol from the parameter library based on the cord blood collection requirements list and current process parameter configuration. The centrifugation protocol includes core parameters such as centrifugation speed, centrifugation time, temperature control mode, acceleration curve, and deceleration curve, and specifies the number of centrifugation cycles required for this batch of cord blood. Medical personnel or laboratory technicians aseptically dispense the collection bags containing cord blood into suitable centrifuge tubes or bags, and write the corresponding cord blood collection batch number into a barcode or electronic tag. The centrifugation control subsystem reads the barcode via barcode scanning or radio frequency identification (RFID) to ensure the correspondence between subsequent data records and specific samples. After detecting that all samples have been placed in the centrifuge and the equilibration settings have been completed, the centrifugation control subsystem automatically starts the centrifuge according to the pretreatment centrifugation plan, and records operation data such as centrifugation start time, planned speed, real-time speed, temperature curve and end time. When an abnormality is detected in the centrifugation process, such as speed fluctuations exceeding the set range or temperature control failure, the system will interrupt the process and remind the operator to handle it. The abnormal information is recorded in the centrifugation log and the batch of samples is marked as requiring manual review.

[0075] After centrifugation, technicians remove the red blood cell layer and plasma fraction according to the pretreatment protocol, retaining the white membrane layer rich in mononuclear cells. This white membrane layer is then resuspended and washed using a pre-prepared buffer solution to obtain a mononuclear cell suspension. Subsequently, the cell counting unit uses an automated cell counter or hemocytometer, combined with log information, to perform total cell counts and assess cell viability in the mononuclear cell suspension. The counting results, viability results, and corresponding detection time, detection equipment number, and operator identification are entered into the system. The cell counting unit associates these measurement data with the umbilical cord blood collection batch number, forming a structured data entry containing centrifugation condition parameters and cell count information. The data management module names this entry a mononuclear cell suspension data package. The mononuclear cell suspension data package includes at least the centrifugation protocol number, number of centrifugations, a summary of key centrifugation parameters, sample volume, total number of mononuclear cells, number of cells per unit volume, and viability marker for this batch of samples. The metadata also records the data generation time and the current process version number. The mononuclear cell suspension data package is used as the output field of this step and is designated as the input field of the subsequent step S230 in the data stream. The mononuclear cell suspension data package is directly read by the separation and purification subsystem when performing Ficoll Hypaque density gradient stratification processing. At the same time, the cell number information in the mononuclear cell suspension data package is also used to compare with the target cell number estimated in step S210, which is used by the quality management module to generate the process evaluation report in the future.

[0076] S230. The mononuclear cell suspension data package is subjected to Ficoll Hypaque density gradient stratification and CD105 and CD73 antibody incubation and magnetic bead sorting to generate high-purity umbilical cord blood mesenchymal mononuclear cell structures with recorded purity, yield and batch number.

[0077] This step is achieved collaboratively by the separation and purification subsystem and the phenotypic screening subsystem. The separation and purification subsystem is responsible for the Ficoll Hypaque density gradient stratification process, while the phenotypic screening subsystem is responsible for CD105 and CD73 antibody incubation and magnetic bead sorting. First, the separation and purification subsystem reads the sample volume, total number of mononuclear cells, and number of cells per unit volume from the mononuclear cell suspension data package output in step S220. Based on the standard operating procedure for density gradient stratification in the current process parameter configuration, it calculates the required volume of mononuclear cell suspension and Ficoll Hypaque density gradient solution to be added to each centrifuge tube or separator. The volume ratio and loading order of these two components constitute the core parameter set for this stratification process. The Ficoll Hypaque density gradient solution is a pre-prepared and quality-tested solution whose density is controlled within a specific range to create an environment for stratification based on cell density differences under the influence of the centrifugation field. The separation and purification subsystem instructs operators to first add a set volume of Ficoll Hypaque density gradient solution to the separator, and then slowly spread the mononuclear cell suspension on top of the density gradient solution by controlling the liquid level and operating speed to avoid violent mixing. The operating rhythm during the spreading process can be assisted by an automated liquid handling device or an operating terminal with speed control prompts.

[0078] After sample preparation, the separation and purification subsystem communicates with the centrifugation control subsystem to invoke a centrifugation protocol for density gradient stratification. This protocol differs from the pretreatment centrifugation protocol and typically employs specific rotation speeds, times, and acceleration / deceleration curves to establish a clear cell separation interface within the density gradient environment. The centrifugation control subsystem automatically initiates the centrifugation process according to this protocol and records key operating parameters throughout the centrifugation cycle. Upon completion of centrifugation, the separation and purification subsystem guides the operator, based on a pre-defined interface recognition strategy and operational prompts, to identify the white membrane layer located at the plasma-density gradient solution interface within the separator. In this invention, this white membrane layer is considered the target layer rich in mononuclear cells. Under aseptic conditions, the operator carefully aspirates this white membrane layer using a specified pipette and transfers it to a new centrifuge tube for buffer washing to remove residual Ficoll Hypaque density gradient solution and plasma components. The time, number of aspiration and washing operations, and the volume of buffer used are all recorded by the separation and purification subsystem according to the process version, thereby establishing an updated cell suspension state based on the mononuclear cell suspension data.

[0079] Subsequently, the phenotypic screening subsystem takes over the cell suspension for this batch and begins CD105 and CD73 antibody incubation and magnetic bead sorting. The phenotypic screening subsystem first retrieves currently valid antibody incubation protocols from the parameter library, including parameters such as the types of monoclonal antibodies targeting CD105 and CD73 molecules, recommended dosages (e.g., antibody mass required per million cells), incubation temperature, and incubation time, and records the protocol version number. Based on the number of mononuclear cells, the system calculates the required volume or mass of CD105 and CD73 monoclonal antibodies to be added. Before the incubation step begins, the operating terminal lists the names and quantities of the reagents to be added for technicians to verify. During incubation, the cell suspension is gently mixed in a constant-temperature shaking device to allow the CD105 and CD73 monoclonal antibodies to fully bind to the antigen sites on the surface of the target cells. After incubation, the phenotypic screening subsystem prompts the technician to add magnetic beads conjugated with secondary antibodies that match the types of antibodies mentioned above. The specifications and amount of magnetic beads are also determined by the protocol in the parameter library to achieve magnetic labeling of CD105 and CD73 double-positive cells.

[0080] After antibody and magnetic bead incubation, the phenotypic screening subsystem controls the magnetic sorting device to begin operation. The magnetic sorting device typically includes a magnetic field generating unit and a sorting column assembly. Technicians slowly pass the incubated cell suspension into the pre-equilibrated sorting column. Under the influence of the magnetic field, CD105 / CD73 double-positive cells carrying magnetic beads are retained within the column, while unlabeled or only single-positive cells are eluted with the effluent. The system controls the execution of the washing steps according to the preset number of washes and washing buffer volume, and records the time point of each wash and the batch information of the buffer used. After washing, the magnetic sorting device switches to elution mode. By removing the magnetic field or changing its direction, the CD105 / CD73 double-positive cells retained in the sorting column are eluted into a new collection tube, forming a high-purity umbilical cord blood mesenchymal mononuclear cell suspension sample. The phenotypic screening subsystem then organizes the purity and yield assessment process for cell samples, including detecting the CD105 / CD73 double-positive ratio using flow cytometry and recounting the total number of cells using a cell counting unit to assess the yield of this sorting compared to the total number of mononuclear cells. The purity and yield determination results, along with information such as the corresponding umbilical cord blood collection batch number, density gradient stratification protocol number, and antibody incubation protocol version number, are written into the data management module.

[0081] The data management module generates a structured record based on the above information, naming it "High-purity Umbilical Cord Blood Mesenchymal Mononuclear Cell Structure." This structure includes at least the proportion of CD105 / CD73 double-positive cells, the total number of cells, the number of cells per unit volume, a comparison with the estimated total treatment cell requirement in step S210, the corresponding batch number, and a summary of all process parameters. These fields constitute the minimum set of parameters reflecting the core improvement of this invention, directly affecting the setting of the inoculation concentration in the subsequent induction culture system and the allocation of cell dosage in each cycle of the treatment. For optional extended functions, this structure may also include supplementary fields, such as cell viability markers, detection results of other surface markers, or additional quality control indicators for specific patient groups. These fields can be enabled or disabled by the quality management module according to different project needs. After generation, the high-purity umbilical cord blood mesenchymal mononuclear cell structure is written into the cell preparation database by the separation and purification subsystem and indexed by patient identification and umbilical cord blood collection batch number. In the overall process of this invention, the high-purity umbilical cord blood mesenchymal mononuclear cell structure is used as the output field of this step and is read by the subsequent step S310 for inoculation concentration setting and induction culture system assembly, becoming an important part of the induction culture input set. At the same time, the purity and yield data of the high-purity umbilical cord blood mesenchymal mononuclear cell structure are also sent back to the total therapeutic cell demand estimation module for updating the statistical model and process evaluation, thereby forming a continuously evolving and auditable process optimization closed loop in the multi-batch preparation process.

[0082] In summary, the technical effects of this step are as follows: By implementing FicollHypaque density gradient stratification and CD105 and CD73 antibody incubation and magnetic bead sorting in series on the basis of a single nucleated cell suspension data package, this invention generates high-purity umbilical cord blood mesenchymal mononuclear cell structures with recorded purity, yield and batch number. It organically connects the upstream collection and pretreatment links with the downstream induction culture and clinical dosage planning, and builds a stable, traceable basic cell resource at the cell source level that is adapted to the needs of urinary incontinence treatment.

[0083] Step S300 includes at least steps S310-S330:

[0084] S310. Obtain high-purity umbilical cord blood mesenchymal mononuclear cell structures, perform inoculation concentration setting treatment and induction culture system assembly treatment, and obtain an induction culture input set that records the culture medium components and inoculation density;

[0085] Specifically, this step is executed collaboratively by the cell preparation management subsystem and the induction culture configuration subsystem. The cell preparation management subsystem first reads key fields of the target cell batch from the high-purity umbilical cord blood mesenchymal mononuclear cell structure generated in step S230. This high-purity umbilical cord blood mesenchymal mononuclear cell structure records at least the CD105 / CD73 double-positive ratio, total cell count, number of cells per unit volume, cell viability markers, and the corresponding umbilical cord blood collection batch number and process version number. When the cell preparation management subsystem detects that a batch of cells has been marked as ready for induction culture, it automatically triggers the inoculation concentration setting process and binds the batch number to the current patient treatment configuration. The treatment configuration is provided by the clinical decision support subsystem and includes parameters such as the planned number of treatment cycles, the cell dosage range for multi-point injection into the bladder wall per cycle, the cell dosage range for tail vein infusion, and the maximum passage size allowed for induction culture. These parameters constitute the boundary constraints for setting the inoculation concentration in this step. After combining the above treatment configuration and the total number of cells recorded in the high-purity umbilical cord blood mesenchymal mononuclear cell structure, the cell preparation management subsystem calculates the target starting cell quantity required to construct a complete therapeutic cell batch, and determines the range of actual cell quantity to be inoculated, taking into account the subsequent multi-generation culture expansion rate and the amount of samples retained for quality testing.

[0086] In the inoculation concentration setting process, the induction culture configuration subsystem retrieves the standard inoculation density range suitable for umbilical cord blood mesenchymal mononuclear cells from the parameter library according to the preset basic inoculation density strategy of the current project, and dynamically fine-tunes the inoculation density based on the CD105 / CD73 double-positive ratio and survival status markers in the high-purity umbilical cord blood mesenchymal mononuclear cell structure. The basic inoculation density strategy clearly defines the recommended cell number density range and the corresponding induction culture cycle length for each type of culture container (e.g., standard culture flask, multi-well plate, or carrier in a bioreactor). This invention considers these recommended inoculation densities, container types, and induction cycles together as a set of core parameters essential for achieving dual-pathway induced differentiation, while feed modes such as slow gradient increases and optimization schemes for different container combinations can be considered as preferred or optional extended parameters, which are only enabled in specific scenarios. The induction culture configuration subsystem calculates the inoculation volume and number of cells for a single container based on the number of cells per unit volume recorded in the high-purity umbilical cord blood mesenchymal mononuclear cell structure and the effective culture area of ​​the candidate container. The system then compares the calculation results with the allowable range in the parameter library. If the result exceeds the allowable range, the system prompts the operator to select another container configuration or adjust the target amplification fold. When the inoculation quantity meets the preset conditions, the system marks the inoculation parameters as a valid configuration and writes it into the configuration draft for this induction culture.

[0087] In the assembly of the induction culture system, the induction culture configuration subsystem retrieves the basal culture medium formulation and inducing factor parameters from the process parameter library for constructing a dual-pathway induction environment for urothelial and neural pathways. The basal culture medium typically uses a DMEM / F12 formulation (a mixture of Dulbecco-modified Eagle medium and Ham F12 medium), with a preset concentration of fetal bovine serum, human serum, or serum-free alternatives added, selected based on clinical safety and project requirements. The fetal bovine serum ratio and whether a serum-free system is used are considered optional extended parameters. Regarding the core improvement of this invention, this step focuses on using a combination of inducing factors containing transforming growth factor β (TGF-β) and epidermal growth factor (EGF). These two factors exist in the parameter library as target concentration ranges and combinations, and are associated with specific indications and target differentiation directions. The induction culture configuration subsystem inputs the batch number, total cell count, recommended inoculation density, and proposed culture medium components and inducing factor combinations from the high-purity umbilical cord blood mesenchymal mononuclear cell structure into the induction culture configuration table. It also specifies the culture container type, initial culture volume, planned medium change time, maximum passage number, and estimated total culture time, assigning a unique induction culture configuration number to this table. Subsequently, the system generates specific operating instructions based on this configuration table. These instructions include when and by which technician or automated liquid handling device to add the basal culture medium to which containers, when to add the inducing factor, and how to complete the inoculation and mixing process. The operating instructions also specify rules for recording anomalies and retry strategies in case of timeouts or parameter deviations.

[0088] After setting the inoculation concentration and assembling the induction culture system, the cell preparation management subsystem integrates the inoculation density, culture medium components, induction factor parameters, culture container configuration, planned culture timeline, and corresponding batch number related to this batch to form a structured input data record for subsequent multi-generation induction culture execution. This record is named the induction culture input set. Each record in the induction culture input set is bound to a high-purity umbilical cord blood mesenchymal mononuclear cell structure batch number, indicating that the corresponding cell batch was used for the dual-pathway induction culture inoculation protocol. The induction culture input set is stored in the cell preparation database and is explicitly specified as the input field of step S320, called the induction culture input set, in the data stream. It is directly called by the multi-generation culture execution subsystem when starting the induction culture process. At the same time, the key information in this input set will be used in step S330 to trace back the induction condition versions corresponding to different therapeutic cell batches.

[0089] S320. Extract cell batch information and induction factor parameters from the induction culture input set, perform multi-generation culture process execution processing and culture time node recording processing, and generate an induction culture process record set that records the culture conditions and cell state markers of each generation;

[0090] Specifically, this step comprises a highly automated execution chain consisting of a multi-generation culture execution subsystem, a culture environment monitoring subsystem, and a cell state assessment subsystem. Upon receiving the induction culture input set from the previous step, the multi-generation culture execution subsystem establishes an execution plan for the current batch of cell induction culture based on the inoculation protocol and culture container configuration recorded in the input set. This plan includes the inoculation time, the expected time interval for reaching a specific confluence, the planned medium change time, the expected passage time, and the sampling and testing tasks for each generation of cells. When the system detects that the culture room environment meets the initial inoculation requirements and that the corresponding high-purity umbilical cord blood mesenchymal mononuclear cell structure has completed room temperature equilibration and pretreatment, the multi-generation culture execution subsystem, through a human-computer interface, reminds the laboratory technicians to perform the actual inoculation according to the inoculation volume and cell density recorded in the induction culture input set. After inoculation, the subsystem registers the operation time, operator identification, and container number in the system, forming the start time node for the first generation of induction culture.

[0091] During the multi-generation culture process, the culture environment monitoring subsystem continuously records environmental parameters for each culture container by connecting to temperature, humidity, and gas concentration detectors within the incubator. These parameters are compared to the target environment specified in the induction culture input set, such as constant temperature, constant humidity, and constant carbon dioxide (CO2) volume fraction. When deviations from preset ranges are detected, an alarm is triggered, and the corresponding time point and degree of deviation are recorded. For automatically adjustable incubators, the environmental monitoring subsystem can also issue adjustment commands, allowing the equipment control module to correct temperature and gas flow within defined ranges. For incubators without automatic adjustment capabilities, the system primarily performs monitoring and alarm functions, requiring manual adjustments. Based on the medium change time points set in the induction culture input set, the multi-generation culture execution subsystem prompts technicians to change the culture medium in the corresponding containers as the target time window approaches. The old culture medium recovery time, new culture medium batch number, and operator information during the medium change operation are all recorded in the system for subsequent analysis of the relationship between cell generation status and culture medium batch. In one technical solution, the multi-generation culture execution subsystem can be integrated with an automated culture medium replacement device. When a container is detected to have reached a preset fusion threshold, the automated device completes part or all of the medium replacement steps, and the system automatically records the execution status. In another technical solution, the multi-generation culture execution subsystem only provides reminder and recording functions, and technicians perform all operations according to the prompts.

[0092] At each critical juncture of cell culture, the multi-generation culture execution subsystem invokes the cell state assessment subsystem to observe and record cell morphology and proliferation status. The cell state assessment subsystem typically includes an inverted microscope imaging unit and an image analysis module. Technicians observe cells in the culture container at designated time points according to the plan, acquiring representative field-of-view images through image acquisition. The image analysis module provides qualitative or quantitative descriptions of cell morphological characteristics, confluence, and potential abnormalities, such as recording whether cells exhibit a spindle shape, degeneration, or excessive dense growth. For projects requiring higher resolution, continuous tracking can be performed in conjunction with a real-time cell imaging system; this is an optional extended function. In a preferred embodiment of the invention, the system also rigorously records the passage operations for each generation of cells, including initial cell density, harvest time, digestive enzyme treatment conditions, passage ratio, and post-passage cell density. This information, along with environmental parameters and observation results, is stored in a database, forming a time-series record of the induced culture process.

[0093] In the recording and processing of culture time nodes, the multi-generation culture execution subsystem establishes an event timeline around each inoculation, each medium change, each passage, and each planned or unplanned sampling. For each event, it records the event type, occurrence time, involved container number, involved cell batch number, and key environmental parameters at the time. Based on system configuration, version information is added to each event, indicating the version of the induction culture protocol used at that time. Understandably, when a batch of cells exhibits significant abnormalities during culture, such as a significant decrease in proliferation rate or abnormal cell morphology, the multi-generation culture execution subsystem will initiate an anomaly handling process according to preset trigger conditions, prompting technicians to investigate the cause, such as checking the culture medium batch, fluctuations in environmental parameters, or potential deviations in previous steps, and recording the handling process in the event timeline. Through this method, the present invention forms a complete and traceable process record chain in the multi-generation culture execution stage, coupling environmental, operational, and cell state information together.

[0094] Finally, the multi-generation culture execution subsystem and the cell state assessment subsystem integrate the time-series data involving culture conditions, environmental parameters, operational events, and cell state markers for each generation to form a structured induction culture process record set. In the induction culture process record set, each cell generation is treated as a hierarchical unit, including the start time, end time, inoculation or passage conditions, culture medium batch used, all medium change events within the generation, all sampling and detection events within the generation, and key cell state descriptions. Simultaneously, the record set metadata indicates the corresponding high-purity umbilical cord blood mesenchymal mononuclear cell structure batch number and induction culture configuration number. The induction culture process record set is stored in the cell preparation database and explicitly designated as the input field for step S330. It is invoked by subsequent marker detection result parsing and batch screening processing corresponding to treatment thresholds. Furthermore, some statistical information in the induction culture process record set can be fed back to the upstream inoculation concentration setting module when needed to correct the inoculation strategy for subsequent batches, achieving continuous optimization of culture conditions.

[0095] S330. The CK18 and βⅢ tubulin labeling detection results of the induction culture process record set are analyzed and processed, and the batch screening corresponding to the treatment threshold is performed. The resuspension and cooling steps of the cryopreservation solution composed of DMSO, human serum albumin and MEM are registered to generate a therapeutic cell batch structure that records the dual-pathway labeling achievement status and cryopreservation batch number.

[0096] In this step, the labeling detection and batch screening subsystem and the cryopreservation management subsystem work together. The labeling detection and batch screening subsystem is responsible for analyzing the functional marker detection results and determining thresholds for each generation of cells in the induction culture process record set. The cryopreservation management subsystem is responsible for performing cryopreservation resuspension and programmed cooling operations on the selected cell batches and recording relevant parameters. Specifically, the labeling detection and batch screening subsystem first identifies the cell passages and sampling time points for urothelial and neural marker detection based on the preset sampling plan in the induction culture process record set, typically selecting cells from the third generation or those close to the planned passage for therapeutic use. During sampling, the system guides technicians to collect representative cell samples from the corresponding culture containers, part of which is used for immunofluorescence detection and part for flow cytometry detection. The relevant operation time, sample number, and cell batch number are simultaneously recorded in the system.

[0097] During the labeling and detection process, the system uses a specific antibody against CK18 (Cytokeratin 18) to label cells for urothelial differentiation and a specific antibody against β-III tubulin (Class III Beta Tubulin) for neural differentiation. The labeling and batch screening subsystem retrieves currently available immunostaining or flow cytometry protocols from a parameter library, including primary and secondary antibody types, incubation time, incubation temperature, number of washes, and color development or signal acquisition settings. These parameters are then correlated with samples from the corresponding passages in the induction culture process record set. After technicians perform antibody incubation and washing according to the detection protocol, images or signals are acquired using a fluorescence microscope or flow cytometer. The image analysis module or flow cytometry analysis module calculates the proportion of CK18-positive cells and β-III tubulin-positive cells in each cell sample and outputs structured detection results. The labeling and batch screening subsystem also writes these detection results into the corresponding detection event entries in the induction culture process record set, allowing each detection result to be traced back to specific induction conditions and environmental parameters.

[0098] In the batch screening process corresponding to the treatment threshold, the labeling detection and batch screening subsystem compares the above detection results according to the treatment threshold rules pre-set according to the present invention. The treatment threshold rules specify the minimum CK18 positivity rate and the minimum β-III tubulin positivity rate that cell batches used for clinical treatment must meet, and multiple versions can be maintained in the parameter library according to different projects and indications. When the labeling detection result meets the current version of the treatment threshold rules, the system marks the corresponding cell batch as labeled as qualified and includes the batch in the candidate set for cell construction that can be used for therapeutic purposes; when the labeling detection result is lower than the threshold, the system marks the corresponding batch as unqualified and decides whether to continue culturing or use it only for research purposes according to the configuration. In one technical solution, the treatment threshold rules can set a buffer zone, and batches close to the threshold can be manually reviewed. Experienced technicians or clinical experts can make further judgments on whether to include them in treatment by combining cell morphology and additional functional experiment results; in another technical solution, the treatment threshold determination is completely executed automatically by the system according to the digital threshold, and manual intervention is only required when the system prompts an abnormality or multiple consecutive batches fail to meet the standard. After completing the above determination, the labeling detection and batch screening subsystem writes the CK18 positivity rate, β-III tubulin positivity rate, treatment threshold version number, and determination result of each batch into the batch labeling table, providing clear batch labels for subsequent cryopreservation management and treatment planning.

[0099] For cell batches marked as compliant, the cryopreservation management subsystem, upon receiving the batch list from the labeling detection and batch screening subsystem, initiates the registration process for the cryopreservation solution resuspension and cooling steps. The cryopreservation management subsystem reads currently valid cryopreservation system configurations from the parameter library. This invention preferably uses a cryopreservation solution combination containing DMSO (Dimethyl Sulfoxide), human serum albumin, and MEM (Minimum Essential Medium), and records the volume fraction range and preparation order of each component in the parameter library. These components and their proportions constitute the minimum set of core parameters for this step. Based on the total number of cells in the batch to be cryopreserved and the planned number of dispensing tubes, the system calculates the required number of cells and cryopreservation solution volume for each cryopreservation tube. Through the user interface, it guides technicians to resuspend the compliant cell pellet in the pre-prepared cryopreservation solution under aseptic conditions. The centrifugation conditions used in the resuspension operation, the cell density after resuspension, and the short-term storage temperature are all recorded. Subsequently, the cryopreservation management subsystem places the cryovials containing cell suspension into the programmed cooling device according to the programmed cooling scheme. The programmed cooling device performs temperature reduction operations at each time interval according to the preset cooling curve until the predetermined low-temperature endpoint is reached. Then, technicians transfer the cryovials to a liquid nitrogen tank or cryogenic freezer for long-term storage. The system records key time points, temperature curves, device numbers, and operator information throughout the entire cooling process. Abnormal situations (such as cooling rates deviating from the preset range) are separately marked in the log.

[0100] After the resuspension and cooling steps of the cryopreservation solution are completed, the cryopreservation management subsystem registers each successfully cryopreserved cell tube as a cryopreservation unit and generates a corresponding cryopreservation batch number for each batch. The cryopreservation batch number code includes the cell batch identifier, cryopreservation date, cryopreservation device number, and cryopreservation protocol version number. Subsequently, the labeling detection and batch screening subsystem and the cryopreservation management subsystem work together to integrate the labeling status of the cells in this batch (including the positive ratio of CK18 and βIII tubulin and whether it meets the treatment threshold), the cryopreservation batch number, the cryopreservation location index (such as the cryopreservation rack number and cryopreservation tube location), the total number of usable cells, and the recommended thawing use (e.g., for multi-point injection in the bladder wall or for tail vein infusion) to form a structured therapeutic cell batch record. The data management module names this record as the therapeutic cell batch structure. In the overall process of this invention, the therapeutic cell batch structure serves as a key intermediate product connecting induction culture and clinical application. It is stored in the cell preparation database and managed through a triple index of patient identification, umbilical cord blood collection batch number, and cryopreservation batch number. In the data flow, it is explicitly designated as the therapeutic cell batch structure as the input field of step S410. The personalized combination dosing regimen generation and closed-loop application module driven by functional feedback are invoked to achieve a direct connection from cell quality control to treatment-level dosage planning.

[0101] In summary, the technical effects of this step are as follows: By analyzing the labeling detection results of CK18 and β-IIItubulin based on the induction culture process record set and screening the batches corresponding to the treatment threshold, and combining the resuspension of cryopreservation solution composed of DMSO, human serum albumin and MEM and the programmed cooling registration process, this invention constructs a structured therapeutic cell batch structure, establishes a stable and traceable correspondence between cell functional status and cryopreservation batch information, and provides quality-controllable cell resources for subsequent personalized combination drug administration.

[0102] Step S400 includes at least steps S410-S430:

[0103] S410. Obtain the batch structure of therapeutic cells, perform treatment-level cell dose allocation processing, and combine bladder wall multi-point injection and tail vein infusion pathway processing to obtain an initial set of combined dosing regimens that records the injection target parameters and intravenous dose parameters for each cycle.

[0104] Specifically, this step is primarily executed by the combined drug delivery regimen scheduling subsystem, in conjunction with the parameter configuration of the patient management subsystem, image navigation subsystem, and cell preparation management subsystem. The combined drug delivery regimen scheduling subsystem first retrieves the target cell batch to be applied from the therapeutic cell batch structure generated in step S330. This therapeutic cell batch structure records at least the cryopreservation batch number, total number of usable cells, recommended thawed cell density, maximum distributable dosage after thawing, the corresponding induction culture configuration number, and previous functional marker detection results for that batch. Once the combined drug delivery regimen scheduling subsystem detects that a particular therapeutic cell batch has been designated for treatment planning by a clinician in the information system, it automatically triggers the treatment-level cell dose allocation process. It then associates this batch with the corresponding urinary control function typing results and target area weight map structure through the patient identifier. Both the urinary control function typing results and the target area weight map structure are outputs of step S130 and have already been stored and version-identified in the patient management subsystem.

[0105] In the treatment-level cell dose allocation process, the combined dosing regimen orchestration subsystem retrieves the current patient's treatment configuration parameters from the patient management subsystem. These parameters include at least the planned number of treatment cycles (e.g., four consecutive weeks), the upper and lower limits of the allowed cell dose per cycle, the baseline dose allocation ratio between bladder wall multi-point injection and tail vein infusion, the range of cell quantities allowed per injection target, and recommended dose adjustment coefficients for different case subtypes. These parameters constitute the minimum set of parameters for achieving individualized dose allocation in this step. The planned number of treatment cycles, the total allowed dose per cycle, and the baseline dose allocation ratio are mandatory parameters. Adjustment coefficients for specific subtypes and dose scaling coefficients adjusted for weight or renal function status can be preferred extended parameters, stored in a parameter library in template form according to project requirements. The combined drug delivery system integrates the total number of available cells in the batch structure of the treatment cells and the number of cycles in the treatment course configuration to calculate the total cell dose that can be allocated for the entire treatment course. After multiplying the total number of available cells by the allocation ratio, the cells are evenly or distributed to each treatment cycle according to a set decreasing (increasing) strategy. Within each cycle, the cells are further subdivided according to the basic ratio of multi-point injection in the bladder wall and tail vein infusion to form the initial values ​​of the cycle-level local injection dose and intravenous infusion dose.

[0106] Furthermore, in the combined processing of multi-point injection into the bladder wall and tail vein infusion pathway, the combined drug delivery protocol scheduling subsystem calls the pelvic floor ultrasound or magnetic resonance imaging templates stored in the image navigation subsystem, and overlays the target area weight map structure generated in step S130 onto the standardized bladder outlet and periurethral anatomical coordinate system. The spatial coordinates and functional repair weights of each region recorded in the target area weight map structure are converted by the system into a set of candidate injection targets. Each candidate injection target has standardized spatial coordinates, a label of its corresponding anatomical region, a functional repair weight value associated with the case subtype, and an injectable depth range. Within each treatment cycle, the combined drug delivery protocol scheduling subsystem selects and sorts the candidate injection targets based on the total local injection dose in the treatment course-level cell dose allocation results, combined with the target area weights. It prioritizes allocating injection doses to targets with higher weights, while controlling the total number of injection targets within the same cycle to within a preset range (e.g., four to six targets). The spatial distance between adjacent targets is checked using the spatial measurement module of the image navigation subsystem to avoid excessive clustering of injection points. For each target included in the current injection plan, the system generates injection target parameters, which include at least the target spatial coordinates, recommended needle insertion direction, recommended needle tip dwell depth, planned injection volume of cell suspension, and injection sequence. If necessary, the corresponding image section markers and local anatomical notes can also be added for intraoperative navigation display.

[0107] For tail vein infusion routes, the combined dosing regimen orchestration subsystem selects a suitable infusion route from the intravenous infusion configuration library based on the baseline dose ratio and the intravenous infusion dose per cycle in the treatment regimen configuration, taking into account the patient's weight, intravenous condition, and previous infusion tolerance. The intravenous infusion route includes at least the recommended infusion route (e.g., upper limb vein or central vein), recommended infusion duration range, recommended upper limit of infusion rate, recommended dilution volume range, and key vital sign monitoring points to be recorded during infusion. These parameters are maintained by the intravenous infusion configuration library and constitute the core set of intravenous dosing parameters. The combined dosing regimen orchestration subsystem binds these intravenous infusion parameters to the intravenous dose per cycle, forming corresponding intravenous dose parameter entries. For scenarios requiring coordination with other supportive treatments, the intravenous dose parameters can also record coordination points and information related to anesthesia, sedation, or bladder function training. This is considered an optional extended field, and the clinical team decides whether to enable it based on the project protocol.

[0108] After completing the treatment-level cell dose allocation and pathway combination, the combination dosing regimen orchestration subsystem summarizes the injection target parameters and intravenous dose parameters for each treatment cycle, forming structured cycle-level dosing entries, and generates a cross-cycle initial combination dosing regimen set for the entire treatment course. In the initial combination dosing regimen set, each cycle records a cycle number, planned execution date, the batch number of therapeutic cells used, a list of bladder wall injection targets for that cycle, and a list of tail vein infusion parameters. Each target entry and each intravenous infusion entry has an independent parameter version number and operational recommendations. The initial combination dosing regimen set is stored in the clinical information system and explicitly designated as the input field for step S420 in the data stream. The initial combination dosing regimen set is called sequentially during dosing by the ultrasound-guided bladder wall multi-point injection execution record processing and tail vein infusion execution record processing. Simultaneously, the cycle number information in the initial combination dosing regimen set will be used in step S430 to correlate the dosing execution record with the functional evaluation results, thereby constructing a cross-cycle closed-loop adjustment basis.

[0109] S420. Extract single-cycle injection parameters and intravenous infusion parameters from the initial combined dosing regimen set, perform ultrasound-guided bladder wall multi-point injection execution record processing and tail vein infusion execution record processing, and generate a set of periodic dosing execution records that record the dosing time axis and dosage implementation status.

[0110] In this step, the drug administration management subsystem serves as the core control unit, working in conjunction with the ultrasound navigation terminal, infusion pump control device, and electronic medical record system to guide on-site operations and record the process. As the planned execution date of a treatment cycle approaches, the drug administration management subsystem automatically extracts the corresponding single-cycle injection parameters and intravenous infusion parameters from the initial combined drug administration protocol set, based on the clinical scheduling table and cycle number in the initial combined drug administration protocol set. These parameters are then pushed to the ultrasound navigation terminal and infusion pump control device during the preoperative preparation phase, triggering the drug administration process for that cycle. After the patient enters the room and completes identity verification, the system matches the patient identifier with the cycle number. Once a match is successful, the drug administration management subsystem internally establishes the initial node of the drug administration timeline for that cycle, writing the combination relationship of treatment number – cycle number – patient identifier – therapeutic cell batch number into the log, laying the foundation for subsequent tracking.

[0111] In the processing of ultrasound-guided multi-point bladder wall injection execution records, the ultrasound navigation terminal reads the list of bladder wall injection targets from the corresponding cycle entries of the initial combined dosing regimen set, loading the injection target parameters one by one, including spatial coordinates, recommended needle insertion direction, needle tip depth, and planned injection volume. The ultrasound navigation terminal overlays these parameters onto real-time acquired transabdominal or transperineal ultrasound images, displaying the approximate location and marker number of each target point in the image through an image overlay, and providing the operator with dynamic adjustment tools, allowing for fine-tuning of target point coordinates and needle insertion paths when there are differences between the patient's position or bladder filling status and the planned state. Each adjustment of the target point parameters is recorded by the ultrasound navigation terminal, including the values ​​before and after the adjustment and the operator's identification, and written back to the dosing execution management subsystem, forming the actual usage version of the target point parameters for this cycle. When using thawed cell suspensions from therapeutic cell batches, the dosing execution management subsystem guides nursing staff to complete syringe preparation and labeling based on the total local injection dose for this cycle and the planned injection volume for each target point, indicating the correspondence between target point numbers and syringe numbers.

[0112] During actual acupuncture injection, the physician, guided by an ultrasound navigation terminal, sequentially punctures and injects into each injection target. The ultrasound navigation terminal receives event signals such as the start of needle insertion, arrival at the target point, commencement of injection, and completion of injection via a hand-controlled button or foot switch, and automatically marks the timestamps and probe positions of each event in conjunction with real-time images. The drug administration management subsystem aggregates these events to form an execution record entry for each target injection. Each entry includes at least the target number, number of punctures, actual injection volume, injection start and end times, and status indicators such as whether blood aspiration, local leakage, or significant patient discomfort occurred during the injection process. If the injection is interrupted or a backup target is used, the physician selects the corresponding reason code through the navigation terminal, and the system writes the reason code and related description into the execution record. The above information constitutes the core record set for the dosage implementation of multi-point injections in the bladder wall in this invention.

[0113] During the tail vein infusion execution recording process, the infusion pump control device reads intravenous dose parameters from the corresponding cycle entry in the initial combined dosing regimen set, including the recommended infusion route, target infusion volume, target cell dose, recommended infusion duration, and recommended upper limit of infusion rate. After connecting the patient's intravenous access, the nursing staff selects the intravenous infusion protocol for the current cycle on the infusion pump control device interface. The system automatically presets the infusion duration and starting rate, and prompts for the input of legally valid double-verified information, including patient identity, therapeutic cell batch number, and pre-prepared cell suspension label number. After the infusion begins, the infusion pump control device propels the cell suspension at the set rate, recording the cumulative infusion volume, current instantaneous rate, and pause or rate adjustment events during the infusion process in real time. If a situation requiring temporary cessation of infusion occurs, such as local venous pain, leakage, or abnormal patient vital signs, the nursing staff executes a pause operation through the control device. The system automatically generates a pause event and corresponding timestamp, and requires the recording of the reason and handling measures. At the end of the infusion, the infusion pump control device uploads the actual infusion volume, actual infusion duration, and final rate curve to the drug administration management subsystem as an objective record of the intravenous dose implementation for this cycle.

[0114] During the dosing timeline generation process, the dosing execution management subsystem integrates the aforementioned bladder wall multi-point injection events and tail vein infusion events in chronological order to construct the dosing timeline for this cycle. The dosing timeline marks each puncture event, the time point at which each target point injection is completed, the start and end times of intravenous infusion, and all abnormal events, pause events, and parameter adjustment events, and associates this information with the treatment course number, cycle number, and therapeutic cell batch number. This timeline can be recorded at minute or even finer time granularity, depending on the specific implementation plan and equipment support capabilities. After the end of this cycle, the dosing execution management subsystem packages the timeline data along with target-level and infusion-level dose implementation data to generate structured cycle dosing execution record entries, ultimately forming a cycle dosing execution record set. Each cycle in the cycle dosing execution record set has a recording unit, which details the actual injection volume, number of punctures, operation sequence, and changes in the actual volume, duration, and rate of intravenous infusion for each target point during this cycle. It also records any dosing-related unexpected or warning events that occur during this cycle. The periodic dosing execution record set is written into the clinical information system and is clearly marked as the input field of step S430 on the data stream. It is then called by subsequent functional assessment aggregation and combined dosing closed-loop adjustment processing.

[0115] S430, Summarize and process changes in the number of urinary incontinence occurrences, changes in urodynamic parameters, and changes in symptom scale scores on the periodic dosing execution record set, generate protocol adjustment parameters, and generate a closed-loop adjustment structure for combined dosing that records the data input requirements for a new round of urinary continence function and the combined dosing adjustment parameters.

[0116] In this step, the closed-loop adjustment engine and the functional assessment aggregation subsystem jointly perform data aggregation and adjustment parameter generation. The closed-loop adjustment engine first selects record units within the target treatment course or target cycle range from the periodic drug administration execution record set generated in step S420, aligning the timeline of each cycle's drug administration events with the corresponding cycle's follow-up assessment data. This follow-up assessment data includes records of the number of urinary incontinence occurrences, urodynamic examination results, and symptom scale scoring items. Records of the number of urinary incontinence occurrences are collected by the patient management subsystem and may originate from patient-filled diaries, mobile terminal records, or nursing records. The content includes at least the number of involuntary urinary leakage events per day and the leakage situation in specific activity scenarios. Urodynamic examination results are output by a dedicated urodynamic testing device, including multiple curve parameters and numerical parameters such as maximum urinary flow rate, bladder compliance, urethral closure pressure, and residual urine volume. Symptom scale scoring items are derived from standardized questionnaires, covering aspects such as urinary urgency, urinary frequency, and the degree of impact of urinary leakage on quality of life. The functional evaluation aggregation subsystem binds the above three types of data with the unique period number for each period, laying the foundation for subsequent trend analysis.

[0117] In the summary processing of changes in the frequency of urinary incontinence, the functional assessment aggregation subsystem calls the event records in the patient management subsystem, correlates the baseline number of leaks before the start of the treatment course with the number of leaks recorded during each cycle's follow-up, calculates the amount and direction of change in the number of leaks relative to the baseline for each cycle, and internally generates a summary of urinary incontinence changes for each cycle. The summary records the cycle number, baseline reference window, current follow-up window, the difference in average number of leaks between the two, and a classification label, such as decrease, remain largely unchanged, or increase. These classification labels are given by pre-configured threshold rules. The rules can also include statistical methods specifically for nocturnal leaks or leaks triggered by specific activities, as optional extended parameters for more granular analysis of specific symptom types when needed. The functional assessment aggregation subsystem writes the summary of urinary incontinence changes for each cycle into an internal functional evaluation table and establishes a chain link between the treatment course number and the dose implementation status in the cycle's dosing execution record set.

[0118] In the processing of urodynamic parameter changes, the functional assessment aggregation subsystem imports standardized output files from the urodynamic monitoring equipment, extracting key parameters such as maximum urinary flow rate, voiding intravesical pressure, bladder compliance, urethral closure pressure, and residual urine volume, and comparing them with the baseline examination results for this treatment course. The system generates parameter change entries for each cycle, categorized by parameter type. Each entry includes the examination time for that cycle, the equipment number, the original parameter value, the difference from the baseline value, and a predefined direction of change marker. For parameters that may be significantly affected by bladder fullness, the functional assessment aggregation subsystem can also record the bladder capacity range and the patient's subjective assessment level during the examination, providing a reference for subsequent manual interpretation. All urodynamic parameter change entries are aligned temporally with the dosing timeline in the cycle dosing execution record, providing a temporal basis for subsequently determining the correlation between different dosing strategies and parameter changes.

[0119] In the processing of symptom scale score changes, the functional assessment aggregation subsystem compares the total score and scores of each dimension (such as urinary urgency, urinary frequency, and impact on quality of life) on the urinary incontinence-related symptom scale completed by the patient for each treatment cycle with the baseline scale before the start of the treatment cycle, calculates the score differences, and indicates the direction of change. The system can generate a comprehensive symptom change index on a cycle, normalizing the differences in multi-dimensional scores and synthesizing them into a comparable indicator to assist the closed-loop adjustment engine in making comprehensive trade-offs. In one technical solution, the functional assessment aggregation subsystem can also support patients to submit simplified scales multiple times within a cycle via mobile terminals. The closed-loop adjustment engine performs time-weighted processing on this additional data to further refine the understanding of symptom fluctuations during the treatment cycle. This part is an optional extension function and can be enabled or disabled depending on the system deployment environment.

[0120] After summarizing the aforementioned changes, the closed-loop adjustment engine enters the parameter generation and processing stage. Utilizing pre-defined adjustment decision rules, the engine jointly analyzes the dosage implementation records from the periodic dosing execution log with the leakage change summary, urodynamic parameter change entries, and symptom scale change information provided by the functional assessment aggregation subsystem. The decision rules include at least the basic adjustment strategies for different case subtypes, adjustment thresholds for the ratio of local injection volume to intravenous infusion volume, fine-tuning step sizes for injection target weights, and criteria for determining whether to extend or shorten the treatment cycle. These rules constitute the core rule set for generating adjustment parameters. Based on the comprehensive response of each cycle, the closed-loop adjustment engine generates several adjustment parameters for the next round or cycle of dosing, such as increasing or decreasing the injection volume for a certain type of target, increasing the local injection ratio or appropriately reducing the intravenous dose while maintaining the total dose, adjusting the injection sequence, and changing the time interval between certain cycles. For cases where efficacy improvement is not significant but safety is good, the system may generate suggestions to strengthen local injection; for cases with adverse reactions or unfavorable changes in urodynamic parameters, the system may generate suggestions to reduce the total dose or extend the cycle interval. These recommendations are presented in the form of structured adjustment parameters, with a summary of the basis for each type of adjustment parameter provided to help clinicians understand the reasons for the system's recommendations during review.

[0121] While generating adjustment parameters, the closed-loop adjustment engine also proposes specific requirements for the next round of urinary continence data input, forming a new round of urinary continence data input requirement field. For example, when the system identifies a significant change in urethral closure pressure after a certain cycle but a lack of significant change in the number of leaks, it can mark the priority of urodynamic re-examination in the input requirements. When it identifies an inconsistency between changes in the patient's subjective symptoms and objective parameters, it can mark an increase in the frequency of symptom scale follow-up or guide the patient to record specific trigger scenarios. These input requirements, together with the adjustment parameters, are encapsulated into a combined drug administration closed-loop adjustment structure. The combined drug administration closed-loop adjustment structure includes at least input configuration suggestions for the next S110 data collection (e.g., specific types of examinations and time points to be collected) and a list of adjustment parameters for the next S410 treatment cycle-level cell dose allocation and pathway combination. The closed-loop adjustment structure for combined drug administration is written into the clinical information system and the cell preparation management system. It is called by step S110 during the next round of data collection and by step S410 during the planning of the combined drug administration regimen in the next round or cycle. Thus, at the methodological level, a closed-loop operation mechanism is formed, which includes urinary continence function assessment, cell preparation, induction culture, individualized drug administration, and functional reassessment.

[0122] In summary, the technical effects of this step are as follows: By jointly summarizing and analyzing the periodic dosing execution record set with the number of urinary incontinence occurrences, urodynamic parameters, and symptom scale scores, this step generates a structured closed-loop adjustment structure for combined dosing. It establishes a mapping relationship between the dosing process and functional response at the data level and provides clear parameter inputs for subsequent rounds of functional data collection and optimization of combined dosing regimens, making the entire treatment process exhibit traceable, iterative, and configurable closed-loop characteristics.

Claims

1. A method for preparing and applying umbilical cord blood mesenchymal mononuclear cells for the treatment of urinary incontinence, characterized in that, include: We acquire raw urodynamic data, bladder and urethral image annotation data, and symptom scale records. We perform time-aligned coding, parameter field organization, case classification, and target weight modeling to generate a target weight map structure carrying case classification labels and target weight parameters. Based on the target area weight map structure, the total cell requirement for treatment is estimated, umbilical cord blood collection and centrifugation preprocessing are performed, and Ficoll Hypaque density gradient stratification combined with CD105 and CD73 immunomagnetic sorting is used to generate high-purity umbilical cord blood mesenchymal mononuclear cell structures with recorded purity, yield and batch number. Based on the high-purity umbilical cord blood mesenchymal mononuclear cell structure, the seeding concentration setting, multi-generation induction culture execution, and CK18 and βⅢ tubulin labeling detection combined with cryopreservation steps were processed to generate a therapeutic cell batch structure that records the dual-pathway labeling achievement status and cryopreservation batch number. The system acquires the batch structure of therapeutic cells, performs treatment-level cell dose allocation, records of multi-point injection into the bladder wall and tail vein infusion, and drives protocol adjustment through functional assessment. It then generates a closed-loop adjustment structure for combined dosing, recording the dosing execution timeline and the input requirements for the new round of urinary continence function data, as well as the combined dosing adjustment parameters.

2. The method according to claim 1, characterized in that, Raw urodynamic data, annotated bladder and urethral imaging data, and symptom scale records include: The raw urodynamic data include bladder pressure time series, urethral pressure time series, urine flow rate time series, and filling volume time series. The bladder pressure time series includes the bladder pressure curve during filling, the intravesical pressure curve during voiding, and the abdominal pressure compensation curve. The urethral pressure time series includes the resting urethral pressure distribution curve, the urethral pressure change curve under stress, and the urethral pressure response waveform during the cough test. The urine flow rate time series includes the maximum urine flow rate, the average urine flow rate, and characteristic points of the flow time curve. The filling volume series records the changes in bladder volume at different time points. The bladder and urethra image annotation data includes dynamic images of the bladder neck position, continuous frames of the urethra course, and images of the sphincter region, acquired by pelvic floor ultrasound or other medical imaging equipment at multiple time phases including resting state, straining state, and before and after urination. The region of interest boundaries of the bladder outlet region, proximal urethra contour, mid-urethral wall structure, distal urethral junction, and surrounding soft tissue support structures are marked on the images in coordinate form by the annotator. The symptom scale records include self-reported answers to the standardized urinary incontinence scale collected via electronic terminals.

3. The method according to claim 1, characterized in that, The process of performing time-aligned encoding and parameter field organization also includes: The time-aligned encoding process includes establishing a timeline for the raw urodynamic data according to the start and end times of the examination, mapping the acquisition time points of each frame of the image annotation data to the urodynamic timeline, encoding the symptom scale records according to the questionnaire completion date and assessment period label, and handling missing timestamps and abnormal sampling intervals. The parameter field processing includes standardizing the naming and unit normalization of peak bladder pressure, resting urethral closure pressure, maximum urinary flow rate, and residual urine volume according to a predefined urinary continence function parameter dictionary; encoding the field for bladder neck displacement distance, urethral angle change, and pelvic floor descent degree; converting urinary incontinence frequency classification, quality of life score, and nocturia frequency; and setting field version identifiers.

4. The method according to claim 1, characterized in that, The process of case subtyping and target weight modeling also includes: The case classification process includes extracting sphincter function indicators, bladder neck mobility indicators, and urethral closure pressure indicators from the urinary continence function dataset; performing rule matching and classification label encoding according to the classification rule set; and generating urinary continence function classification results containing main type labels and auxiliary labels. The target area weight modeling process includes constructing a basic anatomical model of the bladder outlet and the area around the urethra, identifying the bladder neck contour and the urethral lumen centerline through a boundary fitting algorithm, discretizing the area into target area grids, labeling the grids with functional categories based on the urinary control function classification results, and calculating the functional repair weight value for each grid according to the weight rule set.

5. The method according to claim 1, characterized in that, The process of estimating total cell requirements for treatment, collecting umbilical cord blood, and pre-centrifuging also includes: The total cell requirement estimation process includes sorting and prioritizing each target area according to the functional repair weight in the target area weight map structure, merging multiple target areas that are too close anatomically or highly related in function into the same injection target group, calculating the theoretical total number of cells required by weighted summation based on the number of basic cells injected per injection and the number of basic cells infused per tail vein infusion in the treatment course design parameters and the number of treatment cycles, and introducing process loss prediction coefficient and quality inspection sampling reserve coefficient. The cord blood collection and centrifugation pretreatment process includes calculating the required cord blood volume range based on the estimated total cell demand for treatment, generating a cord blood collection demand list, collecting cord blood under the guidance of the cord blood collection execution terminal and monitoring the collection volume and anticoagulation parameters, and performing centrifugation pretreatment steps, including retrieving the pretreatment centrifugation protocol for centrifugation, removing the red blood cell layer and plasma portion while retaining the white membrane layer, performing cell counting and viability assessment, and generating a mononuclear cell suspension data package.

6. The method according to claim 1, characterized in that, The process of Ficoll Hypaque density gradient stratification combined with CD105 and CD73 immunomagnetic sorting also includes: The Ficoll Hypaque density gradient stratification combined with CD105 and CD73 immunomagnetic sorting process includes reading the sample volume and total number of cells from a mononuclear cell suspension data package, calculating the Ficoll Hypaque density gradient liquid volume and cell suspension volume ratio, performing density gradient stratification centrifugation, identifying and aspirating the white membrane layer for washing, and then performing CD105 and CD73 antibody incubation and magnetic bead sorting, including adding CD105 and CD73 monoclonal antibodies for incubation, adding magnetic beads coupled with secondary antibodies for magnetic labeling, separating CD105 / CD73 double-positive cells using a magnetic sorting device, and evaluating purity and yield.

7. The method according to claim 1, characterized in that, The process of setting the inoculum concentration and performing multiple generations of induction culture also includes: The inoculation concentration setting process includes reading the CD105 / CD73 double-positive ratio, total number of cells and number of cells per unit volume from the high-purity umbilical cord blood mesenchymal mononuclear cell structure, calculating the target starting number of cells in combination with the treatment course configuration parameters, dynamically fine-tuning the inoculation density according to the basic inoculation density strategy and calculating the inoculation volume and number of inoculated cells. The multi-generation induction culture execution process includes establishing an execution plan based on the induction culture input set, monitoring culture environment parameters, performing medium change and passage operations, observing cell morphology and proliferation status through the cell state assessment subsystem, and generating an induction culture process record set by recording the conditions, environmental parameters and operation events of each generation of culture.

8. The method according to claim 1, characterized in that, The process of registering and processing CK18 and βⅢ tubulin labeling combined with cryopreservation steps also includes: The registration and processing of CK18 and βⅢ tubulin labeling detection combined with cryopreservation steps includes: identifying sampling time points from the induction culture process record set; collecting cell samples for immunofluorescence or flow cytometry detection; labeling with CK18 and βⅢ tubulin specific antibodies; analyzing the proportion of CK18-positive cells and βⅢ tubulin-positive cells; determining the labeling achievement status by batch screening according to treatment threshold rules; performing cryopreservation resuspension and programmed cooling operations on the labeled cell batches, including resuspension using cryopreservation solution containing DMSO, human serum albumin, and MEM; cooling according to the programmed cooling protocol; and registering the cryopreservation batch number and cryopreservation location.

9. The method according to claim 1, characterized in that, The process of recording the execution of treatment-level cell dose distribution, multi-point injection into the bladder wall, and tail vein infusion also includes: The treatment-level cell dose allocation process includes reading the total number of available cells and the recommended thawed cell density from the therapeutic cell batch structure, allocating the total cell dose for the entire treatment course to each treatment cycle in combination with the treatment course configuration parameters, and further splitting the local injection dose and intravenous infusion dose within each cycle according to the basic dose allocation ratio of multi-point injection in the bladder wall and tail vein infusion. The bladder wall multi-point injection and tail vein infusion execution record processing includes extracting injection target parameters and intravenous infusion parameters from the initial combined dosing regimen set, performing bladder wall multi-point injection under the guidance of an ultrasound navigation terminal and recording the number of needle pricks, actual injection volume and event timestamp for each target point, performing tail vein infusion under the control of an infusion pump control device and recording the actual infusion volume, duration and rate changes, and generating a periodic dosing execution record set.

10. The method according to claim 1, characterized in that, The process of functional evaluation-driven solution adjustment also includes: The functional assessment-driven protocol adjustment process includes aligning the timeline of dosing events with follow-up assessment data from the periodic dosing execution record set, summarizing changes in the number of urinary incontinence occurrences, changes in urodynamic parameters and symptom scale scores, generating combination dosing adjustment parameters and new round of urinary continence function data input requirements according to adjustment decision rules, and forming a closed-loop adjustment structure for combination dosing.

Images