Tumor tissue identification method and device and electronic equipment

Abstract

The invention relates to the technical field of tumor tissue discrimination, and discloses a tumor tissue identification method and device and electronic equipment, and the method comprises the steps: obtaining a blood flow index time sequence of a target object; based on the blood flow index time sequence, blood flow characteristics used for reflecting the blood flow dynamic characteristics of the target object are determined; determining a judgment result of the target object based on the blood flow characteristics; and under the condition that the judgment result is the tumor tissue, determining the tumor type and confidence of the tumor tissue. On the basis of the determined target tissue blood flow index, the blood flow index is analyzed from multiple dimensions, the blood flow dynamic characteristics of the target object can be reflected more comprehensively, and the method does not only depend on a single index. The complexity of a blood flow mode can be captured through multi-dimensional feature extraction, so that the tumor can be distinguished more accurately and reliably.

Description

Technical Field

[0001] This invention relates to the field of tumor tissue identification technology, specifically to a tumor tissue identification method, device, and electronic device. Background Technology

[0002] Tumor tissue identification is an important area of ​​modern medical research, and with the development of technology, many new technologies have been applied to this task. Among them, diffusion-correlation spectroscopy (DCS) has attracted widespread attention because it can monitor tissue blood flow in real time. Combining the blood flow index (BFI) obtained from DCS, important information about the physiological state of the tumor and surrounding tissues can be obtained.

[0003] However, current DCS technology for tumor tissue identification typically relies solely on static or single-scale BFI values. This singular analytical approach struggles to comprehensively characterize the complexity and dynamic changes of tissue microcirculation. Particularly in tumor tissue, the stability and response characteristics of blood flow regulation are often influenced by multiple factors, making it difficult to capture these dynamic features using only a single BFI value.

[0004] Therefore, there is an urgent need for a more comprehensive assessment method to better understand and identify the blood flow characteristics of tumor tissue and provide a reliable basis for clinical decision-making. Summary of the Invention

[0005] This invention provides a method, device, and electronic device for tumor tissue identification, in order to solve the problem that existing technologies for tumor tissue identification based on DCS technology only analyze from a single dimension.

[0006] In a first aspect, the present invention provides a method for identifying tumor tissue, the method comprising: Obtain the time series of the blood flow index of the target object; Based on the blood flow index time series, blood flow characteristics that reflect the dynamic characteristics of blood flow in the target object are determined; Based on blood flow characteristics, the discrimination result of the target object is determined; In cases where the diagnosis is tumor tissue, determine the tumor type and confidence level of the tumor tissue.

[0007] In one optional implementation, blood flow characteristics reflecting the dynamic characteristics of blood flow in the target object are determined based on the blood flow index time series, including: The blood flow index time series is divided into windows to obtain multiple blood flow index subsequences; Based on the blood flow index subsequence, at least one of the following is identified: stability index, disorder index, and oscillatory structure index; stability index, disorder index, and oscillatory structure index all belong to blood flow characteristics.

[0008] In one alternative implementation, stability indices are determined based on the blood flow index subsequence, including: The mean blood flow index is calculated based on the blood flow index subsequence; Based on the mean blood flow index, the fluctuation parameters of each blood flow index subsequence are calculated; Based on the fluctuation parameters, determine the stability index; or, Based on the blood flow index subsequence, calculate the relative range or absolute deviation of the median of the blood flow index; Stability indices are determined based on the relative range or the absolute deviation of the median.

[0009] In one alternative implementation, a disorder index is determined based on a blood flow index subsequence, including: Based on the sample entropy algorithm, or the approximate entropy algorithm, or the spectral entropy algorithm, determine the disorder parameters of the blood flow index subsequence; Based on the disorder parameters, disorder indices are determined.

[0010] In one alternative implementation, an oscillatory structure index is determined based on a blood flow index subsequence, including: Determine the power spectral density of the blood flow index subsequence; Determine the target frequency band and analysis frequency band range for the target object; Based on the power spectral density, target frequency band, and analysis frequency band range, the energy percentage parameter of the blood flow index subsequence is determined; Based on the energy ratio parameter, the overall oscillation structure indicator is determined.

[0011] In one optional implementation, the determination of the target object's discrimination result based on blood flow characteristics includes: Based on blood flow characteristics, feature vectors are constructed; The feature vector is input into a pre-constructed discriminant scoring function to determine the discriminant score; The discrimination result is determined based on the discrimination score.

[0012] In one alternative implementation, determining the tumor type and confidence level of the tumor tissue includes: Determine the rate of regulatory mutation events, which reflects the difference in mutation frequency in blood flow regulation; The multi-scale mismatch parameters were determined by the mean local blood flow index at different time scales of the blood flow index time series. Determine the disorder enhancement parameters; the disorder enhancement parameters are determined by the disorder parameters. Based on the regulation of mutation event rate, multi-scale mismatch parameters, and disorder enhancement parameters, tumor type and confidence level are determined.

[0013] In one alternative implementation, determining the rate of regulatory mutation events includes: Determine the mean change in blood flow index between adjacent windows; Determine the number of mutations in which the mean change in blood flow index exceeds a preset mutation threshold; Determine the rate of regulatory mutation events based on the number of mutations; or, Calculate the mean of the absolute values ​​of the first differences of the blood flow index subsequences; The rate of regulatory mutation events is determined based on the mean of the absolute values ​​of the first-order differences; or, Calculate the standard deviation of the window mean change in the blood flow index subsequence; The rate of regulatory mutation events is determined based on the standard deviation of the change in the window mean.

[0014] In a second aspect, the present invention provides a tumor tissue identification device, the device comprising: The acquisition module is used to collect the blood flow index of the target object; The algorithm module is used to execute the tumor tissue identification method described in any of the above embodiments; The output module is used to output the discrimination result of the target object; if the discrimination result is tumor tissue, it is also used to output the tumor type and confidence level of the tumor tissue.

[0015] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the tumor tissue identification method of the first aspect or any corresponding embodiment described above.

[0016] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the tumor tissue identification method of the first aspect or any corresponding embodiment described above.

[0017] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the tumor tissue identification method of the first aspect or any corresponding embodiment described above.

[0018] The beneficial effects of this invention are as follows: This invention analyzes the blood flow index from multiple dimensions based on the identified target tissue blood flow index, enabling a more comprehensive reflection of the dynamic blood flow characteristics of the target object, rather than relying solely on a single indicator. This multi-dimensional feature extraction captures the complexity of blood flow patterns, making tumor identification more accurate and reliable, and ultimately providing more precise diagnostic references for clinical practice.

[0019] This invention also employs a hierarchical tissue identification strategy. The first level identifies "normal tissue / tumor tissue," and the second level is triggered if the first level determines the tissue to be tumor tissue, for further identification of whether the tumor is "benign / malignant." This trigger-based two-level structure reduces misjudgments caused by category aliasing and improves the stability and interpretability of the overall identification process. Attached Figure Description

[0020] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the first process of a tumor tissue identification method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a second process for a tumor tissue identification method according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0024] In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0025] In related technologies, the blood flow index is typically calculated using a diffusion correlation model based on the light intensity autocorrelation function obtained from DCS measurements, and this index is used as an important parameter reflecting the activity of tissue blood flow. However, in these technologies, the blood flow index is usually used in analysis as a single value or an average value within a fixed time window, and is directly used for tissue status assessment or discrimination decisions. This static approach is meaningful in reflecting the overall blood flow level, but it still has significant shortcomings in tissue identification tasks. First, the absolute value of the blood flow index is easily affected by factors such as individual differences, detection location, optical path coupling conditions, and tissue heterogeneity, making it difficult to unify the discrimination threshold for different individuals or under different measurement conditions. Second, in complex physiological environments, the differences in blood flow regulation mechanisms among different types of tissues often exhibit dynamic changes, and relying solely on static or single-scale blood flow index values ​​is insufficient to fully characterize the stability and response characteristics of tissue microcirculation regulation.

[0026] Furthermore, in tissue identification applications, normal tissues and tumor tissues, as well as benign and malignant tumors, often correspond to different microcirculation regulation states and dynamic blood flow behaviors. In related technologies, directly using the blood flow index as a single discriminant for tissue type differentiation easily leads to instability in the discrimination results over time or with changes in measurement conditions, making it difficult to simultaneously address the multi-level discrimination needs, such as distinguishing between normal and abnormal tissues and identifying benign and malignant tumors.

[0027] Therefore, although DCS technology and the blood flow index obtained from it have been widely used in reflecting tissue blood flow status, existing technologies still lack a technical solution for tissue identification tasks that can further mine the dynamic information contained in the blood flow index over time while maintaining DCS and the blood flow index as basic detection methods, thereby improving the stability and reliability of tissue type identification. How to construct an analysis and discrimination method more suitable for tissue identification needs without changing the existing DCS hardware system and the basic calculation method of the blood flow index remains to be further researched and resolved.

[0028] In view of this, according to an embodiment of the present invention, a method for identifying tumor tissue is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0029] This embodiment provides a tumor tissue identification method, which can be used on servers, terminals, and mobile terminals, such as mobile phones and tablets. Figure 1 This is a flowchart of a tumor tissue identification method according to an embodiment of the present invention, such as... Figure 1As shown, the process includes the following steps: Step S101: Obtain the blood flow index time series of the target object.

[0030] The target object in this invention refers to biological tissues composed of cell populations, such as breast tissue, liver tissue, and lung tissue.

[0031] In this embodiment, the target object can be continuously detected based on DCS technology, acquiring continuous tissue scattering light signals, and calculating the corresponding blood flow index values ​​at multiple consecutive time points. The obtained blood flow index is expressed in time series form, such as: ; in, Indicates at time The blood flow index value obtained at point N represents the total number of sampling points, and the sampling time interval is [value missing]. .

[0032] The sampling time interval in this embodiment It can be configured according to the detection hardware capabilities, tissue location, and application scenario.

[0033] Blood flow index time series can be represented as Alternatively, it can be represented in vector form as follows: ,in, .

[0034] Step S102: Based on the blood flow index time series, determine the blood flow characteristics used to reflect the dynamic characteristics of blood flow in the target object.

[0035] In some optional implementations, the blood flow index time series is first divided into windows to obtain multiple blood flow index subsequences; then, based on the blood flow index subsequences, at least one of the following blood flow characteristics is determined: stability index, disorder index, and oscillatory structure index; wherein, stability index, disorder index, and oscillatory structure index are all blood flow characteristics.

[0036] To characterize the dynamic behavior of the blood flow index in different time intervals, this embodiment uses a sliding time window method to perform segmented analysis of the blood flow index time series.

[0037] Set the time window length to Then the sampling point is the th sampling point. The blood flow index subsequence within a time window can be represented as: Where k = 1, 2, ..., N M+1.

[0038] Alternatively, as an optional implementation, the sliding window can employ methods such as "point-by-point sliding," "fixed-step sliding," or "non-overlapping segmentation." The window step size is denoted as... (Sampling point unit), then the window sequence is: .

[0039] By dividing the time window, the local variation characteristics of the blood flow index can be analyzed while ensuring the continuity of time, which helps to improve the stability of tissue identification results.

[0040] After obtaining multiple blood flow index subsequences, blood flow features are constructed. In this embodiment, only stability indicators, disorder indicators, or oscillatory structure indicators can be constructed, but stability indicators, disorder indicators, and oscillatory structure indicators are preferred as blood flow features.

[0041] Among these, stability indices primarily reflect the regularity and consistency of blood flow dynamics. Highly stable blood flow characteristics indicate minimal variation in blood flow over time, demonstrating good predictability. This is typically associated with the blood supply of healthy tissues, as blood flow in healthy tissues usually maintains a relatively constant state. Disorder indices measure the randomness and unpredictability of blood flow dynamics. Highly disordered blood flow characteristics indicate large fluctuations and a lack of regularity in blood flow, which may be related to pathological conditions such as tumors or tissue damage. Oscillatory structure indices describe periodic fluctuations or oscillation patterns present in blood flow index sequences. Oscillatory structure reflects the dynamic characteristics of blood flow and can reveal the physiological state of biological tissues. Therefore, in practical applications, it is preferable to combine these three indices to comprehensively assess the blood flow characteristics of the target object.

[0042] Step S103: Based on blood flow characteristics, determine the discrimination result of the target object.

[0043] Specifically, blood flow features are converted into feature vectors. When blood flow features contain multiple features, these different features can be integrated to form a comprehensive feature vector. Then, machine learning, deep learning, or statistical analysis methods are used to determine the final discrimination result. Machine learning models can be pre-built using support vector machines, decision trees, etc., while deep learning models can be pre-built using neural networks, etc. In this embodiment, the discrimination result includes: normal tissue or tumor tissue.

[0044] Step S104: If the determination result is tumor tissue, determine the tumor type and confidence level of the tumor tissue.

[0045] Specifically, the identified blood flow characteristics can be further analyzed, and combined with labeled tumor type data, a multi-classification machine learning model (such as random forest, support vector machine, or deep learning model) can be used to determine the specific type of tumor. In this embodiment, the tumor type can include benign or malignant tumors. If the tumor type is malignant, it can be further identified as sarcoma, lymphoma, etc., and a confidence value for each tumor type can be given.

[0046] In this embodiment, based on the determined blood flow index of the target tissue, the blood flow index is analyzed from multiple dimensions, which can more comprehensively reflect the dynamic blood flow characteristics of the target object, rather than relying solely on a single indicator. This multi-dimensional feature extraction can capture the complexity of blood flow patterns, making tumor identification more accurate and reliable, and ultimately providing more precise diagnostic references for clinical practice.

[0047] Additionally, refer to Figure 2 As shown, this embodiment also employs a hierarchical tissue identification strategy. The first level identifies "normal tissue / tumor tissue," and the second level is triggered if the first level determines the tissue to be tumor tissue, for further identification of whether the tumor is "benign / malignant." This trigger-based two-level structure reduces misjudgments caused by category aliasing and improves the stability and interpretability of the overall identification process.

[0048] This embodiment provides a tumor tissue identification method, which can be used on servers, terminals, and mobile terminals, such as mobile phones and tablets. The process includes the following steps: Step S201: Obtain the blood flow index time series of the target object. For details, please refer to [link to relevant documentation]. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0049] Step S202: Based on the blood flow index time series, determine the blood flow features used to reflect the dynamic characteristics of blood flow in the target object. In this embodiment, the blood flow features include: stability indicators, disorder indicators, and oscillatory structure indicators.

[0050] In this embodiment, instead of using only the absolute value of the blood flow index or a fixed threshold as the criterion, the blood flow index is constructed as a time series x(t), and dynamic discriminant features are extracted from the perspectives of stability, disorder, and regulatory structure, based on the dynamic nature of regulation, to form a criterion system for tissue identification, thus better reflecting the dynamic nature of tissue microcirculation regulation.

[0051] Furthermore, the absolute value of the blood flow index in traditional methods is easily affected by factors such as individual differences, tissue heterogeneity, and probe coupling state, making it difficult to standardize the threshold. In this embodiment, the dynamic behavioral characteristics of x(t) are used for discrimination, which reduces the dependence on the absolute scale and improves the consistency and repeatability of recognition results across individuals and measurement conditions.

[0052] In some alternative implementations, stability indices are determined based on blood flow index subsequences, including: Step a1: Divide the blood flow index time series into windows to obtain multiple blood flow index subsequences.

[0053] Set the time window length to Then the sampling point is the th sampling point. The blood flow index subsequence within each time window is represented as follows: Where k = 1, 2, ..., N M+1.

[0054] Step a2: Calculate the mean blood flow index based on the blood flow index subsequence.

[0055] Step a3: Calculate the fluctuation parameters of each blood flow index subsequence based on the mean blood flow index.

[0056] Step a4: Determine the stability index based on the fluctuation parameters.

[0057] In this embodiment, the stability index is used to characterize the stability of blood flow regulation, which is one of the important physiological characteristics that distinguishes normal tissue from abnormal tissue. To quantitatively describe the stability of the blood flow index within a local time window, this embodiment calculates the average level of the blood flow index and its relative fluctuations within each time window.

[0058] First, in the Calculate the mean blood flow index within each time window: ; And calculate the corresponding normalized fluctuation parameters: ; The normalized fluctuation parameter is used to characterize the degree of fluctuation of the blood flow index relative to its average level.

[0059] Secondly, the fluctuation parameters of all time windows are statistically analyzed to obtain the overall stability index: ; in, Indicates the total number of time windows. The larger the value, the more obvious the relative fluctuation of the blood flow index over time, and the more unstable the microcirculation regulation of the corresponding tissue.

[0060] In some alternative implementations, stability indices are determined based on blood flow index subsequences, including: Step b1: Calculate the relative range or absolute deviation of the median of the blood flow index based on the blood flow index subsequence.

[0061] Step b2: Determine the stability index based on the relative range or the absolute deviation of the median.

[0062] The relative range can be expressed as: The median absolute deviation (MAD) after normalization can be expressed as: .

[0063] In some alternative implementations, disorder indices are determined based on blood flow index subsequences, including: Step c1: Divide the blood flow index time series into windows to obtain multiple blood flow index subsequences.

[0064] Step c2: Determine the disorder parameters of the blood flow index subsequence based on the sample entropy algorithm, or the approximate entropy algorithm, or the spectral entropy algorithm (based on the entropy of the power spectrum normalized distribution).

[0065] Step c3: Determine the disorder index based on the disorder parameter.

[0066] Besides stability, the predictability or disorder of blood flow index time series is also an important indicator reflecting the state of tissue microcirculation. Therefore, this invention introduces a complexity analysis method to quantify the disorder of blood flow index time series.

[0067] In each time window Here, we take the calculation of the corresponding disorder parameter based on the sample entropy method as an example: ; Where m represents the embedding dimension and r represents the similarity tolerance parameter. The larger the sample entropy value, the more irregular and unpredictable the dynamic behavior of the sequence.

[0068] The overall disorder index is obtained by averaging the sample entropy across all time windows: ; This disorder index can reflect whether there are abnormalities or disorders in the process of tissue blood flow regulation.

[0069] In some alternative implementations, oscillatory structure indicators are determined based on blood flow index subsequences, including: Step d1 involves dividing the blood flow index time series into windows to obtain multiple blood flow index subsequences.

[0070] Step d2: Determine the power spectral density of the blood flow index subsequence.

[0071] Step d3: Determine the target frequency band and analysis frequency band range for the target object.

[0072] Step d4: Based on the power spectral density, target frequency band, and analysis frequency band range, determine the energy percentage parameter of the blood flow index subsequence.

[0073] Step d5: Determine the overall oscillation structure indicator based on the energy ratio parameter.

[0074] Different types of tissues may also exhibit different oscillatory structural characteristics during blood flow regulation. To characterize these structural differences, this invention further analyzes the frequency domain characteristics of the blood flow index time series.

[0075] Within each time window, calculate the power spectral density of the blood flow index sequence. Define the target frequency band energy percentage parameter: ; in, The target frequency band is related to the regulation of tissue microcirculation. To analyze the frequency band range.

[0076] Furthermore, by averaging the energy percentage parameters of each window, the overall oscillation structure index is obtained: ; This oscillation structure index can reflect the regulatory characteristics of the blood flow index within a specific frequency band.

[0077] Target frequency band It can be set according to the sampling frequency and application scenario, or it can be replaced with one of the following structural features: spectrum center frequency, spectrum bandwidth; energy ratio of multiple frequency bands (such as low frequency / medium frequency / high frequency energy ratio vector); autocorrelation peak and principal period parameter (characterizing the oscillation structure in the time domain).

[0078] Step S203: Based on blood flow characteristics, determine the discrimination result of the target object; Specifically, step S203 includes: Step S2031: Construct a feature vector based on blood flow characteristics. When blood flow characteristics include stability indices, disorder indices, and oscillatory structure indices, the feature vector is: ; Among them, V represents the stability index of blood flow regulation, E represents the disorder index of blood flow regulation, and R represents the oscillatory structure index of blood flow regulation.

[0079] Step S2032: Input the feature vector into the pre-constructed discrimination scoring function to determine the discrimination score.

[0080] The pre-constructed discriminant scoring function can be: ; in, These are weighting coefficients, which can be obtained by fitting calibration data or training samples, or by setting them based on human experience. Equal weighting or normalized weighting methods can also be used. or . Step S2033: Determine the discrimination result based on the discrimination score.

[0081] For example, when the discrimination score satisfies When the test tissue is identified as tumor tissue, the discrimination score does not exceed the threshold. When the test tissue is identified as normal, the confidence level of the tissue identification result can be output by performing probability mapping on the discrimination score.

[0082] Among them, threshold It can be obtained through optimization of training samples, or it can be set using statistical quantiles (e.g., by taking samples from normal tissue samples). (A certain quantile is used as the threshold), or it is selected based on the trade-off between false positives and false negatives.

[0083] Regarding probability mapping, in addition to Sigmoid mapping, piecewise linear mapping, Softmax (multi-class extension), or probability output methods based on calibration curves can also be used.

[0084] Step S204: If the determination result is tumor tissue, determine the tumor type and confidence level of the tumor tissue.

[0085] In this embodiment, based on the two-level identification, it can be extended to a three-level identification (for example, by adding a "uncertain / retest required" category), or further subdividing different tumor types in the second level. As long as the triggering logic of "first level first identifies normal / tumor, second level subdivides under tumor conditions" is still followed, it is an equivalent extension of the present invention.

[0086] Specifically, step S204 includes: Step S2041: Determine the rate of regulatory mutation events, which reflects the difference in mutation frequency in blood flow regulation.

[0087] In some alternative implementations, step S2041 includes: Step g1: Determine the mean change in blood flow index between adjacent windows; Step g2: Determine the number of mutations in which the mean change in blood flow index exceeds a preset mutation threshold; Step g3: Determine the rate of regulatory mutation events based on the number of mutations.

[0088] Differences in benign and malignant tumor tissue are often more likely to manifest as differences in the frequency of mutations regulating blood flow. To quantify this difference, the change in the mean between adjacent windows is defined as follows: ; Given a mutation threshold Define a mutation event indicator function: ; in, This is an indicator function; it takes the value 1 if the condition is true, and 0 otherwise.

[0089] The rate of regulatory mutation events is defined as: ; Among them, parameters The higher the value, the higher the proportion of significant jumps in the blood flow index during the testing period, reflecting more frequent unstable events in microcirculation regulation.

[0090] In some alternative implementations, step S2041 includes: Step h1: Calculate the mean of the absolute values ​​of the first differences of the blood flow index subsequences; Step h2: Determine the rate of regulatory mutation events based on the mean of the absolute values ​​of the first-order differences.

[0091] That is, the mean of the absolute values ​​of the first-order differences within the window is used as the "abrupt change intensity": .

[0092] In some alternative implementations, step S2041 includes: Step m1: Calculate the standard deviation of the window mean change for the blood flow index subsequence; Step m2: Determine the rate of regulatory mutation events based on the standard deviation of the change in the window mean.

[0093] In other words, the "standard deviation of the window mean change" is used to characterize abrupt fluctuations: .

[0094] Step S2042: Determine the multi-scale mismatch parameters, which are determined by the mean local blood flow index at different time scales of the blood flow index time series.

[0095] The difference between benign and malignant tumors may also manifest as variations in the consistency of blood flow regulation across different time scales. To characterize this difference, a multi-scale expression is constructed based on the local averages of blood flow index sequences at two different time scales (or two different window lengths). Let the local average sequences corresponding to the two scales be... and (Its specific construction can be achieved using moving averages with different window lengths), the multi-scale mismatch parameter is defined as: ; Among them, multi-scale mismatch parameters The larger the value, the more significant the differences in the regulatory behavior of the blood flow index at different time scales, and the worse the multi-scale consistency.

[0096] In addition to using the local average difference between two scales, multi-scale mismatch can also be extended to multi-scale ensembles. We can construct the average mismatch or maximum mismatch as features; alternatively, we can use the complementarity of the multi-scale correlation coefficient as a mismatch index, for example: ; Step S2043: Determine the disorder enhancement parameter, which is determined by the disorder parameter.

[0097] The sample entropy is calculated for each time window. Subsequently, to further reflect the "persistence or proportion of highly disordered dynamic states", a disorder threshold is given. Define the proportion of high disordered windows, i.e., the disorder enhancement parameter: ; in, The larger the value, the higher the proportion of time in the blood flow index sequence that is in a highly disordered state, reflecting more significant dynamic disorder characteristics.

[0098] In addition to a high proportion of disordered windows, a disordered mean can also be used. The extent of enhancement within a tumor segment (e.g., relative to baseline or relative to the overall first-order level) serves as an indicator of malignancy tendency.

[0099] Step S2044: Based on the regulation of mutation event rate, multi-scale mismatch parameters, and disorder enhancement parameters, determine the tumor type and confidence level.

[0100] The controlled mutation event rate, multi-scale mismatch parameter, and disorder enhancement parameter are used to construct a feature vector for distinguishing between benign and malignant diseases. And construct a malignancy tendency discrimination score function: ; in These are weighting coefficients, which can be determined from calibration data or training samples. The weighting coefficients are used to balance the contributions of different features to the distinction between benign and malignant characteristics.

[0101] When the malignant tendency score meets the following: If the tumor tissue is malignant, it is determined to be malignant. If the threshold is reached, the tumor tissue is determined to be a benign tumor. It can be determined based on sample statistics or clinical annotation data.

[0102] To output confidence levels, this invention can map the score to a probability of malignancy: ; in, These are mapping parameters.

[0103] The output includes the benign or malignant determination and the corresponding probability value, which can effectively reflect the reliability of the determination. For window-level output, the final conclusion can be generated by methods such as majority voting, weighted voting, confidence-weighted average, or median fusion.

[0104] In addition, discriminant functions such as quadratic form, logistic regression, or support vector machine can be used, as long as their input is based on... Constructing the second-level feature vector The output should be a good / bad judgment.

[0105] In addition to outputting tissue identification conclusions, this invention can also output confidence information based on the probability mapping of the discrimination score, so that the identification results not only provide category conclusions, but also quantify their reliability, which is convenient for risk warning and decision support in bedside / intraoperative scenarios.

[0106] Furthermore, the tumor tissue identification method provided by this invention is directly compatible with existing DCS hardware systems and mature blood flow index calculation methods. Based on this, the tissue identification function is implemented through an algorithm module, lowering the barriers to system development and clinical translation, and improving the feasibility of device implementation.

[0107] Compared to identification methods that rely solely on a single-dimensional blood flow index, this invention identifies tumor tissue by acquiring time series data of the blood flow index and extracting three core blood flow features: stability, disorder, and oscillatory structure. It further determines the tumor type and confidence level, overcoming the limitations of static or single-scale indicators. This allows for a comprehensive capture of the complex biological information of tumor tissue microcirculation from multiple dimensions, including dynamic regulation, spatiotemporal heterogeneity, and functional response characteristics. The synergistic analysis of multi-dimensional features not only compensates for the limitations of single indicators in characterizing microcirculation function but also enhances the differentiation between tumor tissue and normal tissue, as well as different types of tumor tissue, through complementary correlations between features. This effectively reduces the risk of misjudgment due to insufficient exploration of the dynamic characteristics of microcirculation, ultimately achieving more accurate and reliable tumor tissue identification and classification assessment.

[0108] This embodiment also provides a tumor tissue identification device for implementing the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0109] This embodiment provides a tumor tissue identification device, including: The acquisition module is used to collect the blood flow index of the target object; The algorithm module is used to execute the tumor tissue identification method in any of the above embodiments; The output module is used to output the discrimination result of the target object; if the discrimination result is tumor tissue, it is also used to output the tumor type and confidence level of the tumor tissue.

[0110] The tumor tissue identification device in this embodiment is based on a diffusion-related spectroscopy (DCS) detection system, which can continuously acquire the dynamic signal of scattered light from the tissue under test, calculate the blood flow index, and generate a time series. Based on this, an organization identification algorithm is executed to output the organization type identification result and confidence information.

[0111] The data acquisition module should include at least the following modules: Optical detection module: used to emit near-infrared light toward the target object and receive tissue-scattered light signals; Signal acquisition module: used for sampling, digitizing and time-synchronized management of scattered light signals.

[0112] The algorithm module in this embodiment is used to calculate and output the blood flow index based on the existing DCS model. This invention does not limit the specific calculation method; it is used for... The analysis was performed to determine the tissue identification results, including normal / tumor identification, and benign / malignant identification when the tumor was identified.

[0113] The output module in this embodiment is used to display or output the recognition results and confidence information, and can store or transmit data.

[0114] The above modules can be integrated into the same device or deployed in a distributed manner between embedded processors and host / edge computing devices.

[0115] The module connection relationships and data flow of the device can be summarized as follows: The optical detection module performs near-infrared optical detection on the target object and outputs a scattered light signal to the signal acquisition module. The signal acquisition module samples, digitizes, and manages the time synchronization of the scattered light signal, and transmits the digitized data to the algorithm module. The algorithm module calculates the blood flow index based on the received data and generates a time series. Based on this, the system completes the organization identification and generates corresponding confidence information. The output module receives the identification results and confidence information output by the algorithm module and displays, stores, or transmits them.

[0116] The output module should at least include the following: Tissue identification result: normal tissue or tumor tissue; When identified as tumor tissue, it is further classified as either benign or malignant tumor. Confidence information (e.g., probability value, score value, or level prompt) corresponding to the above identification results; Recorded data related to the testing process, including timestamps and blood flow index sequences. And identification result logs, used for traceability and verification.

[0117] The results can be displayed on a screen, a host computer interface, or a mobile terminal, or output to an external system via wired or wireless means.

[0118] In this embodiment, the optical detection module, signal acquisition module, algorithm module and output module can be integrated into the same device, or a split structure of "front-end acquisition + host computer algorithm + terminal display" can be adopted; the algorithm module can also be deployed on edge computing devices or servers.

[0119] In addition to displaying the results, the output module can also output to an external system via a wired / wireless interface. The output content, besides the identification conclusions and confidence levels, may also include... Sequence and discrimination scores Log information facilitates traceability and verification.

[0120] The device can also be expanded to a multi-probe or multi-channel DCS synchronous detection mode, respectively obtaining multiple blood flow index sequences. At this point, the algorithm module can fuse and output the results from multiple channels, such as averaging or weighted summing the scores of each channel to obtain a comprehensive score, which is an extended implementation of the present invention.

[0121] The device of this invention adopts a modular structure of "optical detection module - signal acquisition module - algorithm module - output module", with clear data flow and functional boundaries. It can realize an integrated portable device or a split deployment of front-end acquisition + host computer / edge computing. At the same time, it can be expanded into a multi-channel / multi-probe form to meet the needs of different application scenarios.

[0122] The device of the present invention can be used for rapid identification of tumor tissue and normal tissue, and can further be used for tumor benign or malignant assessment. It is suitable for intraoperative tissue boundary assistance, bedside rapid assessment, follow-up monitoring and other scenarios, and has good practicality and promotion value.

[0123] The tumor tissue identification device provided in this embodiment of the invention can execute the tumor tissue identification method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units described above are the same as in the corresponding embodiments described above, and will not be repeated here.

[0124] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0125] The following is a detailed reference. Figure 3 The diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 301, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 302 or a program loaded from memory 308 into random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device. The processor 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0126] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0127] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 309, or installed from a memory 308, or installed from a ROM 302. When the computer program is executed by the processor 301, it performs the functions defined in the tumor tissue identification method of the embodiments of the present invention.

[0128] Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0129] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the tumor tissue identification method shown in the above embodiments is implemented.

[0130] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0131] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A tumor tissue recognition method characterized by, The method comprises: acquiring a blood flow index time sequence of a target object; determining a blood flow feature reflecting blood flow dynamic characteristics of the target object based on the blood flow index time sequence; determining a discrimination result of the target object based on the blood flow feature; in the case that the discrimination result is a tumor tissue, determining a tumor type and a confidence level of the tumor tissue.

2. The method of claim 1, wherein, The determination of the blood flow feature reflecting the blood flow dynamic characteristics of the target object based on the blood flow index time sequence comprises: window partitioning the blood flow index time sequence to obtain a plurality of blood flow index subsequences; determining at least one of a stability indicator, a disorder indicator and a shock structure indicator based on the blood flow index subsequences; the stability indicator, the disorder indicator and the shock structure indicator all belong to the blood flow feature.

3. The method of claim 2, wherein, The determination of the stability indicator based on the blood flow index subsequences comprises: calculating a blood flow index mean value based on the blood flow index subsequences; calculating a fluctuation parameter of each blood flow index subsequence based on the blood flow index mean value; determining the stability indicator based on the fluctuation parameter; or, calculating a blood flow index relative range or a median absolute deviation based on the blood flow index subsequences; determining the stability indicator based on the relative range or the median absolute deviation.

4. The method of claim 2, wherein, The determination of the disorder indicator based on the blood flow index subsequences comprises: determining a disorder parameter of the blood flow index subsequences based on a sample entropy algorithm, an approximate entropy algorithm or a spectral entropy algorithm; determining the disorder indicator based on the disorder parameter.

5. The method of claim 2, wherein, The determination of the shock structure indicator based on the blood flow index subsequences comprises: determining a power spectral density of the blood flow index subsequences; determining a target frequency band and an analysis frequency band range of the target object; determining an energy proportion parameter of the blood flow index subsequences based on the power spectral density, the target frequency band and the analysis frequency band range; determining an overall shock structure indicator based on the energy proportion parameter.

6. The method of claim 2, wherein, The determination of the discrimination result of the target object based on the blood flow feature comprises: constructing a feature vector based on the blood flow feature; inputting the feature vector into a pre-constructed discrimination score function to determine a discrimination score; determining the discrimination result based on the discrimination score.

7. The method of claim 2, wherein, The determination of the tumor type and the confidence level of the tumor tissue comprises: determining a regulatory mutation event rate reflecting a difference in mutation frequency of blood flow regulation; determining a multi-scale mismatch parameter determined by local blood flow index mean values of the blood flow index time sequence at different time scales; determining a disorder enhancement parameter determined by a disorder parameter; determining a tumor type and a confidence level based on the regulatory mutation event rate, the multi-scale mismatch parameter and the disorder enhancement parameter.

8. The method of claim 7, wherein, The determination of the regulatory mutation event rate comprises: determining a blood flow index mean value change between adjacent windows; determining a mutation number of the blood flow index mean value change exceeding a preset mutation threshold; determining the regulatory mutation event rate based on the mutation number; or, calculating a first-order difference absolute value mean of the blood flow index subsequence; determining the mutation event rate based on the first-order difference absolute value mean; or, calculating a window mean variation standard deviation of the blood flow index subsequence; determining the mutation event rate based on the window mean variation standard deviation.

9. A tumor tissue identification apparatus, comprising: The device comprises: a collection module configured to collect a blood flow index of a target object; an algorithm module configured to execute the tumor tissue recognition method in any one of claims 1 to 8; an output module configured to output a discrimination result of the target object, and further configured to output a tumor type and a confidence level of the tumor tissue if the discrimination result is the tumor tissue.

10. An electronic device, comprising: comprise: a memory and a processor in communication connection with each other, the memory storing computer instructions, and the processor executing the computer instructions to execute the tumor tissue recognition method in any one of claims 1 to 8.

Images