Prostate cancer recurrence risk prediction system based on multi-modal data

Abstract

The invention relates to the technical field of medical data mining, and provides a prostate cancer recurrence risk prediction system based on multi-modal data, which comprises the following steps: collecting physiological parameters and discomfort moments of a prostate cancer patient after operative treatment, marking a target discomfort moment, and obtaining an adjacent discomfort moment of the target discomfort moment; acquiring a rising and falling ratio of the same type of physiological parameters at the target discomfort moment; dividing and updating abnormal data expression intervals; determining the comprehensive convolution kernel scale of the prostatic cancer patient at the target discomfort moment according to the correlation between any two different types of physiological parameters, the time length of the abnormal data representation interval of the physiological parameters and the rising and falling ratio of all types of physiological parameters at the target discomfort moment; and predicting the prostate cancer recurrence risk of the prostate cancer patient according to the comprehensive convolution kernel scale. According to the invention, effective management of the recurrence risk of prostate cancer can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of medical data mining, and particularly to a prostate cancer recurrence risk prediction system based on multi-modal data. Background Art

[0002] For prostate cancer patients, the surgically removable visible tumors can be removed through surgery. However, cancer cells can spread to different locations such as bones and lymph nodes through the blood or lymphatic system, that is, there are micrometastases of residual cancer cells. At the same time, the postoperative recovery status of prostate cancer patients is affected not only by clinical heterogeneity and physiology, but also by factors such as their own diet, sleep, and medication, resulting in an increased risk of cancer recurrence in prostate cancer. Therefore, timely detection and early intervention can significantly delay the clinical recurrence of tissues and improve the recovery status of patients. Therefore, constructing a prostate cancer recurrence risk prediction model can guide subsequent treatment decisions and improve the prognosis of patients, which has important clinical significance.

[0003] Traditional prediction models such as the MSKC model can be used to extract the characteristics of the physiological parameters of prostate cancer patients after surgery to predict the recurrence risk of prostate cancer. However, the receptive fields of these traditional prediction models have limitations, and there is a problem of single recognition mode. They cannot accurately identify adverse events that have complex effects on the recovery status during the patient's recovery process, nor can they effectively manage the recurrence risk of prostate cancer. Summary of the Invention

[0004] The present invention provides a prostate cancer recurrence risk prediction system based on multi-modal data to solve the problem that when traditional prediction models predict the recurrence risk of prostate cancer, they are affected by the limitation of the receptive field, have a single recognition mode, and cannot effectively manage the recurrence risk of prostate cancer. The specific technical solutions adopted are as follows: An embodiment of the present invention provides a prostate cancer recurrence risk prediction system based on multi-modal data, and the system includes the following modules: A data acquisition module, configured to collect the physiological parameters and discomfort moments of prostate cancer patients after surgical treatment, record any discomfort moment as the target discomfort moment, and obtain the adjacent discomfort moments of the target discomfort moment according to the time interval between the discomfort moments of prostate cancer patients and the target discomfort moment; A physiological parameter analysis module, configured to divide the starting detection stage and the ending detection stage of the target discomfort moment, and obtain the rise and fall ratio of the same type of physiological parameters of prostate cancer patients at the target discomfort moment according to the differences between the same type of physiological parameters of prostate cancer patients during the starting detection stage and the differences during the ending detection stage; An abnormal data performance interval division module, which is used to divide the abnormal data performance interval according to the values of physiological parameters and the corresponding acquisition times of all adjacent discomfort times at the target discomfort time, and update the abnormal data performance interval according to the maximum values of the same type of physiological parameters collected within the abnormal data performance interval; A convolution kernel scale determination module, which is used to determine the comprehensive convolution kernel scale of a prostate cancer patient at the target discomfort time according to the correlation between any two different types of physiological parameters, the time length of the abnormal data performance interval of the physiological parameters, and the rise and fall ratio of all types of physiological parameters at the target discomfort time; A recurrence risk prediction module, which is used to predict the recurrence risk of prostate cancer in a prostate cancer patient according to the comprehensive convolution kernel scale.

[0005] Furthermore, the method for obtaining the adjacent discomfort times of the target discomfort time is as follows: Select the discomfort time that is closest to the target discomfort time in terms of time, and record the time interval between the selected discomfort time and the target discomfort time as the short duration of the target discomfort time; Respectively select each discomfort time included within the short duration of the target discomfort time, and regard all discomfort times whose time interval from the selected discomfort time is less than or equal to the short duration as the selected discomfort time. Repeat the method of selecting discomfort times with a time interval less than or equal to the short duration for the selected discomfort times until no new discomfort times can be selected, and record all selected discomfort times as the adjacent discomfort times of the target discomfort time.

[0006] Furthermore, the specific method for dividing the start detection stage and the end detection stage of the target discomfort time includes: The time period corresponding to minutes before the earliest adjacent discomfort time among all adjacent discomfort times of the prostate cancer patient at the target discomfort time is recorded as the start detection stage of the target discomfort time, where represents the second time threshold; The time period corresponding to minutes after the latest adjacent discomfort time among all adjacent discomfort times of the prostate cancer patient at the target discomfort time is recorded as the end detection stage of the target discomfort time.

[0007] Furthermore, the method for obtaining the rise and fall ratio of the same type of physiological parameters of the prostate cancer patient at the target discomfort time is as follows: Calculate the first-order difference of the same type of physiological parameters of prostate cancer patients at the initial detection stage. Denote the ratio of the mean of the first-order differences of the same type of physiological parameters of prostate cancer patients at the initial detection stage to the mean of the same type of physiological parameters as the first increment ratio of the same type of physiological parameters of prostate cancer patients at the initial detection stage; Denote the sum of the first increment ratio of the same type of physiological parameters and the constant 1 as the first increment sum ratio of the same type of physiological parameters; Denote the product of the mean of the same type of physiological parameters of prostate cancer patients at the initial detection stage and the first increment sum ratio as the first product of the same type of physiological parameters of prostate cancer patients at the initial detection stage. Calculate the first-order difference of the same type of physiological parameters of prostate cancer patients at the end detection stage. Denote the ratio of the mean of the first-order differences of the same type of physiological parameters of prostate cancer patients at the end detection stage to the mean of the same type of physiological parameters as the second increment ratio of the same type of physiological parameters of prostate cancer patients at the end detection stage; Denote the sum of the second increment ratio of the same type of physiological parameters and the constant 1 as the second increment sum ratio of the same type of physiological parameters; Denote the product of the mean of the same type of physiological parameters of prostate cancer patients at the end detection stage and the second increment sum ratio as the second product of the same type of physiological parameters of prostate cancer patients at the end detection stage. Obtain the rise and fall ratio of the same type of physiological parameters of prostate cancer patients at the target discomfort moment according to the first product of the same type of physiological parameters of prostate cancer patients at the initial detection stage and the second product of the same type of physiological parameters of prostate cancer patients at the end detection stage.

[0008] Furthermore, the method for dividing the abnormal data performance interval is as follows: Obtain the first-order difference values of the physiological parameters of all adjacent discomfort moments at the target discomfort moment. Denote the maximum value, the second largest value among the first-order difference values of the same type of physiological parameters, and the acquisition moment corresponding to the maximum value of the same type of physiological parameters as the prominent acquisition moments. Denote the longest time interval divided by all prominent acquisition moments as the abnormal data performance interval.

[0009] Furthermore, the method for updating the abnormal data performance interval according to each maximum value among the same type of physiological parameters collected within the abnormal data performance interval includes the following specific methods: Denote any maximum value among the physiological parameters of the same type collected within the abnormal data manifestation interval as the target maximum value. According to the collection moment corresponding to the target maximum value, divide the abnormal data manifestation interval into two abnormal data manifestation sub-intervals. Denote the average value of the differences between adjacent extreme values of the physiological parameters of the same type collected within the abnormal data manifestation sub-interval as the first average value of the physiological parameters of the same type within the abnormal data manifestation sub-interval. For the physiological parameters of the same type, denote the absolute value of the difference between the first average values of the two abnormal data manifestation sub-intervals divided by the target maximum value as the first absolute value of the target maximum value; Denote the set composed of all physiological parameters of the same type collected within the abnormal data manifestation sub-interval as the same-type parameter set of the abnormal data manifestation sub-interval. Denote the Jaccard correlation coefficient of the same-type parameter sets of the physiological parameters of the same type corresponding to the two abnormal data manifestation sub-intervals divided by the target maximum value as the correlation value of the target maximum value; Denote the average value of the time intervals between all adjacent extreme values of the physiological parameters collected within the abnormal data manifestation interval as the time interval average value of the abnormal data manifestation interval; Determine the division basis degree of the target maximum value for the physiological parameters of the same type within the abnormal data manifestation interval according to the first absolute value, correlation value of the target maximum value, and the time interval average value of the abnormal data manifestation interval; Update the abnormal data manifestation interval according to the division basis degrees of all types of physiological parameters.

[0010] Furthermore, the updating of the abnormal data manifestation interval according to the division basis degrees of all types of physiological parameters includes: Take the maximum value corresponding to the division basis degree as the division position, divide the abnormal data manifestation interval into two abnormal data manifestation sub-intervals, and take the abnormal data manifestation sub-interval where the maximum value of the physiological parameters corresponding to the division basis degree is located as the new abnormal data manifestation interval.

[0011] Furthermore, the method for determining the comprehensive convolution kernel scale of a prostate cancer patient at the target discomfort moment according to the correlation relationship between any two different types of physiological parameters, the time length of the abnormal data manifestation interval of the physiological parameters, and the rise and fall ratio of all types of physiological parameters at the target discomfort moment includes the following specific methods: Determine the expansion coefficients of two different types of physiological parameters according to the correlation relationship between the two different types of physiological parameters; Denote any two different types of physiological parameters as the first type of physiological parameter and the second type of physiological parameter respectively. Denote the product of the expansion coefficient of the first type of physiological parameter and the second type of physiological parameter and the time length of the abnormal data representation interval of the second type of physiological parameter as the convolution kernel scale of the second type of physiological parameter determined based on the first type of physiological parameter. Determine the comprehensive convolution kernel scale of a prostate cancer patient at the target discomfort moment according to the rise and fall ratio of all types of physiological parameters at the target discomfort moment and the convolution kernel scale of the second type of physiological parameter determined by different types of physiological parameters.

[0012] Further, the step of determining the comprehensive convolution kernel scale of a prostate cancer patient at the target discomfort moment according to the rise and fall ratio of all types of physiological parameters at the target discomfort moment and the convolution kernel scale of the second type of physiological parameter determined by different types of physiological parameters includes: Denote the product of the convolution kernel scale of the second type of physiological parameter determined based on the first type of physiological parameter and the rise and fall ratio of the first type of physiological parameter of the prostate cancer patient at the target discomfort moment as the fourth product of the second type of physiological parameter determined by the first type of physiological parameter. Denote the mean value of the fourth products of the second type of physiological parameter determined by all physiological parameters different from the second type of physiological parameter as the adjusted convolution kernel scale of the second type of physiological parameter. Denote the product of the normalized value of the rise and fall ratio of the second type of physiological parameter at the target discomfort moment and the adjusted convolution kernel scale of the second type of physiological parameter as the fifth product of the second type of physiological parameter at the target discomfort moment. Denote the sum of the fifth products of all types of physiological parameters at the target discomfort moment as the first sum at the target discomfort moment. Denote the odd number closest to the first sum at the target discomfort moment as the comprehensive convolution kernel scale of the prostate cancer patient at the target discomfort moment.

[0013] Further, the method for predicting the recurrence risk of prostate cancer in a prostate cancer patient according to the comprehensive convolution kernel scale includes the following specific method: Take the value of the comprehensive convolution kernel scale at the discomfort moment as the value of the convolution kernel scale, and use the MSKC model to predict the recurrence risk of prostate cancer in the prostate cancer patient, and obtain the predicted value of the recurrence risk of prostate cancer in the prostate cancer patient at the discomfort moment.

[0014] The beneficial effects of the present invention are: This application separately collects the physiological parameters and discomfort moments of prostate cancer patients after surgical treatment. First, according to the change trend of the physiological parameters of prostate cancer patients and the distribution of the adjacent discomfort moments of the patients, it evaluates the significance of the fluctuations in the physiological parameters of the same type when discomfort occurs in prostate cancer patients, obtains the rise and fall ratio of the physiological parameters of the same type, and avoids the subjectivity of prostate cancer patients recording based on their own physical feelings. Among the physiological parameters of prostate cancer patients after surgical treatment, the physiological parameter with the most obvious data change amplitude often reflects the characteristics of the impact on the patient's body. According to the value of the physiological parameter, the abnormal data performance interval is divided. Since the recurrence risk of prostate cancer in patients after prostate cancer surgery is affected by many factors such as daily diet, sleep, and drugs, and there are potential associations between these different types of influencing factors, this application conducts a correlation analysis based on the abnormal data performance intervals of different types of physiological parameters, and obtains the comprehensive convolution kernel scale of prostate cancer patients at any discomfort moment according to the correlation analysis results. Finally, based on the comprehensive convolution kernel scale, the prediction of the recurrence risk of prostate cancer in prostate cancer patients is realized, solving the problem that when the traditional prediction model predicts the recurrence risk of prostate cancer, it is affected by the limitation of the receptive field, the recognition pattern is single, and the effective management of the recurrence risk of prostate cancer cannot be achieved. Among them, the comprehensive convolution kernel scale is the size of the receptive field determined by this application. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0016] Figure 1 It is a schematic flowchart of a prostate cancer recurrence risk prediction system based on multi-modal data provided by an embodiment of the present invention; Figure 2 It is a flowchart for obtaining the rise and fall ratio provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0018] Please refer to Figure 1, which shows the flowchart of the prostate cancer recurrence risk prediction system based on multi-modal data provided by an embodiment of the present invention. The system includes: a data acquisition module, a physiological parameter analysis module, an abnormal data performance interval division module, a convolution kernel scale determination module, and a recurrence risk prediction module.

[0019] The data acquisition module collects the physiological parameters and discomfort moments of prostate cancer patients after surgical treatment, records any discomfort moment as the target discomfort moment, and obtains the adjacent discomfort moments of the target discomfort moment according to the time interval between the discomfort moments of prostate cancer patients and the target discomfort moment.

[0020] After surgical treatment of prostate cancer patients, continuously collect the physiological parameters of prostate cancer patients. Among them, the physiological parameters of prostate cancer patients include the heart rate, blood pressure, and emotional state score of prostate cancer patients; the collection time interval of the physiological parameters of prostate cancer patients is 1 minute; the physiological parameters of prostate cancer patients can be automatically extracted by the bracelet worn by prostate cancer patients.

[0021] In the actual application process, as other implementation manners, the implementer can determine the types of data included in the physiological parameters of prostate cancer patients and the value of the collection time interval of the physiological parameters of prostate cancer patients according to the actual situation, and this application does not make special restrictions.

[0022] It should be noted that, for the convenience of calculation, all the physiological parameters of prostate cancer patients involved in the calculation in this embodiment have undergone data preprocessing, thereby canceling the influence of the dimension. This embodiment uses the Z-Score standard normalization method to perform dimension reduction processing on the same type of physiological parameters of prostate cancer patients. In the actual application process, the implementer can use other methods such as the existing technology of the maximum-minimum normalization method to perform dimension reduction processing, which is not limited here.

[0023] There is a potential correlation between the subjective physical discomfort of prostate cancer patients after surgery and cancer recurrence. After prostate cancer surgery, especially after prostatectomy, patients may experience symptoms such as frequent urination, urgency, dysuria, and pain in the pelvic area. These discomforts may be related to the recovery, complications, or tumor recurrence after prostate surgery. For example, pain or discomfort in the pelvic area may indicate local recurrence of the tumor or metastasis to the bone. At the same time, symptoms such as postoperative difficulty in urination or blood in the urine may also indicate abnormalities in the urinary system and potentially suggest local cancer recurrence.

[0024] Therefore, to fully evaluate the physical condition of prostate cancer patients after surgery, when prostate cancer patients feel uncomfortable, prostate cancer patients record the moments of physical discomfort through a smart portable recorder, and all the moments of physical discomfort are recorded as discomfort moments.

[0025] Evaluate the persistence of physical discomfort experienced by prostate cancer patients based on the continuity and time intervals of their discomfort moments over time.

[0026] Denote any discomfort moment of a prostate cancer patient as the target discomfort moment, and count the number of all discomfort moments within the time period composed of the moments whose time intervals from the target discomfort moment are . Select the discomfort moment that is closest in time to the target discomfort moment, and denote the time interval between the selected discomfort moment and the target discomfort moment as the short persistence duration of the target discomfort moment. Select the discomfort moments included within the short persistence duration of the target discomfort moment. For each selected discomfort moment, consider all discomfort moments whose time intervals from the selected discomfort moment are less than or equal to the short persistence duration as the selected discomfort moment. Repeat the selection method with a time interval less than or equal to the short persistence duration for the selected discomfort moments until no new discomfort moments can be selected. Denote all the selected discomfort moments as the adjacent discomfort moments of the target discomfort moment.

[0027] Among them, represents the first time threshold. In this embodiment, the value of the first time threshold is 12 hours. For the sake of easy understanding, when the target discomfort moment is 8:00 on May 4th, the moments whose time intervals from the target discomfort moment are are 20:00 on May 3rd and 20:00 on May 4th, and the composed time period is from 20:00 on May 3rd to 20:00 on May 4th.

[0028] It can be understood that generally, there is only one discomfort moment that is closest in time to the target discomfort moment. In this case, there is only one discomfort moment included within the short persistence duration of the target discomfort moment. However, there may also be two discomfort moments that are closest in time to the target discomfort moment, that is, there is one discomfort moment both before and after the target discomfort moment at the short persistence duration. Therefore, when selecting the discomfort moments included within the short persistence duration of the target discomfort moment, it is necessary to analyze each selected discomfort moment separately, that is, when there are two discomfort moments that are closest in time to the target discomfort moment, analyze these two closest discomfort moments separately.

[0029] It can be understood that the target discomfort moment may correspond to one adjacent discomfort moment or multiple adjacent discomfort moments.

[0030] The adjacent discomfort moments of a prostate cancer patient at any discomfort moment can be obtained in the same way.

[0031] Thus far, obtain the physiological parameters of the prostate cancer patient and the adjacent discomfort moments of all discomfort moments.

[0032] A physiological parameter analysis module divides the start detection stage and the end detection stage of a target discomfort moment, and obtains the rise and fall ratio of the same type of physiological parameter of a prostate cancer patient at the target discomfort moment according to the differences between the same type of physiological parameters of the prostate cancer patient in the start detection stage and the differences between those in the end detection stage.

[0033] When a prostate cancer patient experiences physical discomfort, the discomfort sensation often corresponds to the fluctuations of physiological parameters. For example, when a prostate cancer patient experiences pain, anxiety, fever or other stress responses, data such as the heart rate and blood pressure of the prostate cancer will increase. As the prostate cancer patient's body recovers, data such as the heart rate and blood pressure will gradually decline. In particular, at the start and end moments when a prostate cancer patient experiences physical discomfort, the physiological parameters often show an upward trend and a downward trend. Therefore, according to the change trend of the physiological parameters of a prostate cancer patient, the approaching discomfort moments of the discomfort moment can be further divided to avoid the subjectivity of the prostate cancer patient's record based on their own physical feelings.

[0034] The time period corresponding to the number of minutes before the earliest approaching discomfort moment among all the approaching discomfort moments of a prostate cancer patient at the target discomfort moment is recorded as the start detection stage of the target discomfort moment. The time period corresponding to the number of minutes after the latest approaching discomfort moment among all the approaching discomfort moments of a prostate cancer patient at the target discomfort moment is recorded as the end detection stage of the target discomfort moment.

[0035] Among them, represents the second time threshold, and the value of the second time threshold in this embodiment is 30. For example, when the earliest approaching discomfort moment among all the approaching discomfort moments of a prostate cancer patient at the target discomfort moment is 2:00, 1:30 - 2:00 is the start detection stage. When the latest approaching discomfort moment among all the approaching discomfort moments of a prostate cancer patient at the target discomfort moment is 13:00, 13:00 - 13:30 is the end detection stage.

[0036] Calculate the first-order difference of the same type of physiological parameter of a prostate cancer patient in the start detection stage. The ratio of the mean value of the first-order difference of the same type of physiological parameter of a prostate cancer patient in the start detection stage to the mean value of the same type of physiological parameter is recorded as the first increment ratio of the same type of physiological parameter of the prostate cancer patient in the start detection stage. The sum of the first increment ratio of the same type of physiological parameter and the constant 1 is recorded as the first increment sum ratio of the same type of physiological parameter. The product of the mean value of the same type of physiological parameter of a prostate cancer patient in the start detection stage and the first increment sum ratio is recorded as the first product of the same type of physiological parameter of the prostate cancer patient in the start detection stage.

[0037] Calculate the first-order difference of the same type of physiological parameters of prostate cancer patients in the final detection stage, and denote the ratio of the mean value of the first-order difference of the same type of physiological parameters of prostate cancer patients in the final detection stage to the mean value of the same type of physiological parameters as the second increment ratio of the same type of physiological parameters of prostate cancer patients in the final detection stage. Denote the sum of the second increment ratio of the same type of physiological parameters and the constant 1 as the second increment sum ratio of the same type of physiological parameters; denote the product of the mean value of the same type of physiological parameters of prostate cancer patients in the final detection stage and the second increment sum ratio as the second product of the same type of physiological parameters of prostate cancer patients in the final detection stage.

[0038] Among them, the calculation of the first-order difference is a well-known technology and will not be elaborated here.

[0039] According to the first product of the same type of physiological parameters of prostate cancer patients in the initial detection stage and the second product of the same type of physiological parameters of prostate cancer patients in the final detection stage, obtain the rise and fall ratio of the same type of physiological parameters of prostate cancer patients at the target discomfort moment.

[0040] Among them, the rise and fall ratio has a positive correlation with the first product and a negative correlation with the second product.

[0041] It can be understood that the positive correlation and negative correlation in this application refer to the relationship between the independent variable and the dependent variable. The positive correlation means that the dependent variable increases (decreases) as the independent variable increases (decreases), which can be an additive relationship, a multiplicative relationship, etc.; the negative correlation means that the dependent variable decreases (increases) as the independent variable increases (decreases), which can be an inverse ratio relationship, a subtractive relationship, etc.

[0042] Some other embodiments of this application can be that the ratio of the first product of the same type of physiological parameters of prostate cancer patients in the initial detection stage to the second product in the final detection stage is denoted as the rise and fall ratio of the same type of physiological parameters of prostate cancer patients at the target discomfort moment.

[0043] The rise and fall ratio reflects the significance of the fluctuations in the physiological parameters corresponding to the rise and fall ratio when prostate cancer patients experience discomfort. The flowchart for obtaining the rise and fall ratio is as Figure 2 shown.

[0044] It can be understood that for any type of physiological parameter of prostate cancer patients at the target discomfort moment, there is a corresponding rise and fall ratio. The rise and fall ratio corresponding to any type of physiological parameter of prostate cancer patients at any discomfort moment can be obtained by the same method.

[0045] So far, the rise and fall ratio corresponding to any type of physiological parameter at any uncomfortable moment of a prostate cancer patient is obtained.

[0046] The abnormal data performance interval division module divides the abnormal data performance interval according to the values of the physiological parameters and the corresponding acquisition times of all adjacent uncomfortable moments of the target uncomfortable moment, and updates the abnormal data performance interval according to each maximum value among the physiological parameters of the same type collected within the abnormal data performance interval.

[0047] The risk of prostate cancer recurrence after surgery for prostate cancer patients is affected by various factors such as daily diet, sleep, and medications, and there are potential associations among these different types of influencing factors. Among the physiological parameters of prostate cancer patients after surgical treatment, the physiological parameters with the most obvious data change amplitude often reflect the impact on the patient's body.

[0048] Obtain the first-order difference values of the physiological parameters of all adjacent uncomfortable moments of the target uncomfortable moment, record the maximum value, the second-largest value among the first-order difference values of the physiological parameters of the same type, and the acquisition time corresponding to the maximum value of the physiological parameters of the same type as the prominent acquisition times, and record the time interval divided from the earliest moment to the latest moment among all prominent acquisition times as the abnormal data performance interval.

[0049] It can be understood that the first-order difference value of the physiological parameter of the adjacent uncomfortable moment is the difference between the physiological parameter of the adjacent uncomfortable moment and the physiological parameter of the same type at the previous acquisition moment of the adjacent uncomfortable moment.

[0050] Record any maximum value among the physiological parameters of the same type collected within the abnormal data performance interval as the target maximum value, use the acquisition time corresponding to the target maximum value as the division position, divide the abnormal data performance interval into two abnormal data performance sub-intervals with the acquisition time corresponding to the target maximum value, record the average value of the differences between adjacent extreme values of the physiological parameters of the same type collected within the abnormal data performance sub-interval as the first average value of the physiological parameter of the same type within the abnormal data performance sub-interval, for the physiological parameter of the same type, record the absolute value of the difference between the first average values of the two abnormal data performance sub-intervals divided by the target maximum value as the first absolute value of the target maximum value; record the set composed of all physiological parameters of the same type collected within the abnormal data performance sub-interval as the same-type parameter set of the abnormal data performance sub-interval, and record the Jaccard correlation coefficient of the same-type parameter sets of the physiological parameters of the same type corresponding to the two abnormal data performance sub-intervals divided by the target maximum value as the correlation value of the target maximum value; record the average value of the time intervals between all adjacent extreme values of the physiological parameters collected within the abnormal data performance interval as the time interval average value of the abnormal data performance interval.

[0051] Among them, the calculation of the Jaccard correlation coefficient is a well-known technology and will not be elaborated here.

[0052] Based on the first absolute value of the target maximum value, the correlation value, and the average time interval of the abnormal data performance interval, determine the degree of division basis of the target maximum value for the same type of physiological parameter within the abnormal data performance interval.

[0053] It can be understood that the degree of division basis is positively correlated with the first absolute value and the average time interval respectively, and negatively correlated with the correlation value of the target maximum value.

[0054] Preferably, as an embodiment of the present application, the product of the first absolute value of the target maximum value and the average time interval of the abnormal data performance interval is denoted as the third product of the target maximum value for the same type of physiological parameter within the abnormal data performance interval, and the ratio of the third product of the target maximum value for the same type of physiological parameter within the abnormal data performance interval to the correlation value of the target maximum value is denoted as the degree of division basis of the target maximum value for the same type of physiological parameter within the abnormal data performance interval.

[0055] According to the same method, the degree of division basis of any maximum value of the same type of physiological parameter collected within the abnormal data performance interval can be obtained.

[0056] Take the maximum value corresponding to the degree of division basis as the division position, divide the abnormal data performance interval into two abnormal data performance sub-intervals, and take the abnormal data performance sub-interval where the maximum value of the physiological parameter corresponding to the degree of division basis is located as the new abnormal data performance interval to realize the update of the abnormal data performance interval.

[0057] So far, obtain and update the abnormal data performance intervals corresponding to all types of physiological parameters.

[0058] Convolution kernel scale determination module, according to the correlation relationship between any two different types of physiological parameters, the time length of the abnormal data performance interval of the physiological parameter, and the rise and fall ratio of all types of physiological parameters at the target discomfort moment, determine the comprehensive convolution kernel scale of the prostate cancer patient at the target discomfort moment.

[0059] Furthermore, perform correlation analysis according to the abnormal data performance intervals of different types of physiological parameters.

[0060] Arrange the same type of physiological parameters of the same prostate cancer patient in the order of the acquisition time, and obtain the physiological parameter sequence of the same type of physiological parameters of the same prostate cancer patient. Denote the Pearson correlation coefficient between the physiological parameter sequences of two different types of physiological parameters as the overall correlation degree of the two different types of physiological parameters. Denote the Jaccard correlation coefficient of the two sets respectively composed of two different types of physiological parameters as the interval intersection degree of the two different types of physiological parameters.

[0061] When the interval intersection degree of two different types of physiological parameters is smaller, the values of these two different types of physiological parameters deviate from the change trend of all physiological parameters of the prostate cancer patient. Therefore, it is more necessary to expand the interval of the physiological parameters corresponding to the smaller interval intersection degree in order to capture more information on potential human discomfort data.

[0062] Denote the ratio of the interval intersection degree of two different types of physiological parameters to the overall correlation degree as the first ratio of the two different types of physiological parameters; when the first ratio of two different types of physiological parameters is greater than or equal to , assign the expansion coefficient of the two different types of physiological parameters to 0, and when the first ratio of the two different types of physiological parameters is less than , assign the expansion coefficient of the two different types of physiological parameters to the first ratio of the two different types of physiological parameters.

[0063] Among them, represents the third time threshold, and the value of the third time threshold in this embodiment is 1.

[0064] Denote any two different types of physiological parameters as the first type of physiological parameter and the second type of physiological parameter respectively, and denote the product of the expansion coefficient of the first type of physiological parameter and the second type of physiological parameter and the time length of the abnormal data performance interval of the second type of physiological parameter as the convolution kernel scale of the second type of physiological parameter determined based on the first type of physiological parameter.

[0065] So far, obtain the convolution kernel scale of other different types of physiological parameters determined by any one type of physiological parameter.

[0066] According to the rise and fall ratio of all types of physiological parameters at the target discomfort moment, and the convolution kernel scale of the second type of physiological parameter determined by different types of physiological parameters, determine the comprehensive convolution kernel scale of the prostate cancer patient at the target discomfort moment.

[0067] Multiply the convolution kernel scale of the second type of physiological parameter determined based on the first type of physiological parameter by the rise and fall ratio of the first type of physiological parameter of the prostate cancer patient at the target discomfort moment, and denote it as the fourth product of the second type of physiological parameter determined by the first type of physiological parameter. Denote the mean value of the fourth products of the second type of physiological parameter determined by all physiological parameters different from the second type of physiological parameter as the adjusted convolution kernel scale of the second type of physiological parameter.

[0068] Multiply the normalized value of the rise and fall ratio of the second type of physiological parameter at the target discomfort moment by the adjusted convolution kernel scale of the second type of physiological parameter, and denote it as the fifth product of the second type of physiological parameter at the target discomfort moment. Denote the sum of the fifth products of all types of physiological parameters at the target discomfort moment as the first sum at the target discomfort moment. Denote the odd number closest to the first sum at the target discomfort moment as the comprehensive convolution kernel scale of the prostate cancer patient at the target discomfort moment.

[0069] The comprehensive convolution kernel scale of the prostate cancer patient at any discomfort moment can be obtained by the same method.

[0070] The recurrence risk prediction module predicts the recurrence risk of prostate cancer in prostate cancer patients based on the comprehensive convolution kernel scale.

[0071] Take the value of the comprehensive convolution kernel scale at the discomfort moment as the value of the convolution kernel scale, use the MSKC model to predict the recurrence risk of prostate cancer in prostate cancer patients, and obtain the predicted value of the recurrence risk of prostate cancer in prostate cancer patients at the discomfort moment.

[0072] Among them, the MSKC model is a well-known prediction model, and using the MSKC model to predict the recurrence risk of prostate cancer in prostate cancer patients is a well-known technology, which will not be elaborated here.

[0073] Thus, the prediction of the recurrence risk of prostate cancer in prostate cancer patients is realized.

[0074] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A prostate cancer recurrence risk prediction system based on multimodal data, characterized in that: The system includes the following modules: A data acquisition module is used to collect physiological parameters and discomfort moments of prostate cancer patients after surgical treatment, record any discomfort moment as a target discomfort moment, and obtain the discomfort moment adjacent to the target discomfort moment according to the time interval between the discomfort moment of the prostate cancer patient and the target discomfort moment; A physiological parameter analysis module, for dividing the target discomfort moment into an initial detection phase and an end detection phase, and obtaining a rise-fall ratio of the same type of physiological parameters of the prostate cancer patient at the target discomfort moment according to the difference between the same type of physiological parameters of the prostate cancer patient at the initial detection phase and the difference between the same type of physiological parameters at the end detection phase; The abnormal data performance interval division module is used to divide the abnormal data performance interval according to the values ​​of all physiological parameters close to the target discomfort moment and the corresponding collection time, and update the abnormal data performance interval according to the maximum values ​​of the same type of physiological parameters collected within the abnormal data performance interval; A convolution kernel scale determination module is used to determine the comprehensive convolution kernel scale of a prostate cancer patient at a target discomfort moment according to the correlation between any two different types of physiological parameters, the time length of the abnormal data performance interval of the physiological parameters, and the rise and fall ratio of all types of physiological parameters at the target discomfort moment; The recurrence risk prediction module is used to predict the risk of prostate cancer recurrence in prostate cancer patients based on the comprehensive convolution kernel scale.

2. The prostate cancer recurrence risk prediction system based on multimodal data according to claim 1, characterized in that: The method for obtaining the discomfort moment adjacent to the target discomfort moment is: Select the discomfort moment that is closest to the target discomfort moment in time, and record the time interval between the selected discomfort moment and the target discomfort moment as the short duration of the target discomfort moment; Select each discomfort moment contained in the short duration of the target discomfort moment respectively, and take all discomfort moments whose time interval with the selected discomfort moment is less than or equal to the short duration as the selected discomfort moments, and repeat the selection according to the method of time interval less than or equal to the short duration for the selected discomfort moments until there is no new discomfort moment to be selected, and record all the selected discomfort moments as the adjacent discomfort moments of the target discomfort moment.

3. The prostate cancer recurrence risk prediction system based on multimodal data according to claim 1, characterized in that: The specific method of dividing the target discomfort moment into the initial detection stage and the final detection stage includes: The earliest approaching discomfort moment among all the approaching discomfort moments of the target discomfort moment is set for the prostate cancer patient The time period corresponding to minutes is recorded as the initial detection phase of the target discomfort moment, where: represents a second time threshold; The latest adjacent discomfort moment of all adjacent discomfort moments of the target discomfort moment is set after the latest adjacent discomfort moment of the prostate cancer patient The time period corresponding to the target discomfort moment is recorded as the end detection phase.

4. The prostate cancer recurrence risk prediction system based on multimodal data according to claim 1, characterized in that: The method for obtaining the rise-fall ratio of the same type of physiological parameters of the prostate cancer patient at the target discomfort moment is: Calculate the first-order differences of the same type of physiological parameters of prostate cancer patients at the initial detection stage, record the ratio of the mean of the first-order differences of the same type of physiological parameters of prostate cancer patients at the initial detection stage to the mean of the same type of physiological parameters as the first incremental ratio of the same type of physiological parameters of the prostate cancer patients at the initial detection stage; record the sum of the first incremental ratio of the same type of physiological parameters and a constant 1 as the first incremental sum ratio of the same type of physiological parameters; record the product of the mean of the same type of physiological parameters of prostate cancer patients at the initial detection stage and the first incremental sum ratio as the first product of the same type of physiological parameters of the prostate cancer patients at the initial detection stage; Calculate the first-order differences of the same type of physiological parameters of the prostate cancer patients at the final detection stage, record the ratio of the mean of the first-order differences of the same type of physiological parameters of the prostate cancer patients at the final detection stage to the mean of the same type of physiological parameters as the second incremental ratio of the same type of physiological parameters of the prostate cancer patients at the final detection stage; record the sum of the second incremental ratio of the same type of physiological parameters and a constant 1 as the second incremental sum ratio of the same type of physiological parameters; record the product of the mean of the same type of physiological parameters of the prostate cancer patients at the final detection stage and the second incremental sum ratio as the second product of the same type of physiological parameters of the prostate cancer patients at the final detection stage; Based on the first product of the same type of physiological parameters of the prostate cancer patient in the initial detection stage and the second product of the same type of physiological parameters in the final detection stage, the rise and fall ratio of the same type of physiological parameters of the prostate cancer patient at the target discomfort moment is obtained.

5. The prostate cancer recurrence risk prediction system based on multimodal data according to claim 1, characterized in that: The method for dividing the abnormal data performance interval is as follows: Obtaining the first-order difference values ​​of all physiological parameters adjacent to the target discomfort moment, recording the maximum value and the second largest value of the first-order difference values ​​of the same type of physiological parameters, and the collection time corresponding to the maximum value of the same type of physiological parameters as the prominent collection time; The longest time interval divided by all prominent collection moments is recorded as the abnormal data performance interval.

6. The prostate cancer recurrence risk prediction system based on multimodal data according to claim 1, characterized in that: The specific method of updating the abnormal data performance interval according to the maximum values ​​of the same type of physiological parameters collected within the abnormal data performance interval includes: Any maximum value of the same type of physiological parameters collected within the abnormal data performance interval is recorded as the target maximum value, and the abnormal data performance interval is divided into two abnormal data performance sub-intervals according to the collection time corresponding to the target maximum value, and the average value of the difference between adjacent extreme values ​​of the same type of physiological parameters collected within the abnormal data performance sub-interval is recorded as the first average value of the same type of physiological parameters in the abnormal data performance sub-interval, and for the same type of physiological parameters, the absolute value of the difference between the first average values ​​of the two abnormal data performance sub-intervals divided by the target maximum value is recorded as the first absolute value of the target maximum value; The set of all physiological parameters of the same type collected in the abnormal data performance sub-interval is recorded as the parameter set of the same type in the abnormal data performance sub-interval, and the Jaccard correlation coefficient of the parameter set of the same type corresponding to the physiological parameters of the same type in the two abnormal data performance sub-intervals divided by the target maximum value is recorded as the correlation value of the target maximum value; The mean of the time intervals between all adjacent extreme values ​​of the physiological parameters collected within the abnormal data performance interval is recorded as the mean time interval of the abnormal data performance interval; Determine, according to the first absolute value and the correlation value of the target maximum value and the time interval mean of the abnormal data performance interval, the basis degree for dividing the target maximum value within the abnormal data performance interval into the physiological parameters of the same type; According to the classification criteria of all kinds of physiological parameters, the abnormal data performance interval is updated.

7. The prostate cancer recurrence risk prediction system based on multimodal data according to claim 6, characterized in that: The updating of abnormal data performance intervals according to the classification basis of all types of physiological parameters includes: The maximum value corresponding to the maximum value of the division basis degree is used as the division position, the abnormal data performance interval is divided into two abnormal data performance sub-intervals, and the abnormal data performance sub-interval where the maximum value of the physiological parameter of the type corresponding to the division basis degree is located is used as the new abnormal data performance interval.

8. The prostate cancer recurrence risk prediction system based on multimodal data according to claim 1, characterized in that: The method of determining the comprehensive convolution kernel scale of the prostate cancer patient at the target discomfort moment according to the correlation between any two different types of physiological parameters, the time length of the abnormal data performance interval of the physiological parameters, and the rise and fall ratio of all types of physiological parameters at the target discomfort moment includes the following specific methods: determining expansion coefficients of the two different types of physiological parameters according to a correlation between the two different types of physiological parameters; Any two different types of physiological parameters are recorded as first type physiological parameters and second type physiological parameters respectively, and the product of the expansion coefficient of the first type physiological parameters and the second type physiological parameters and the time length of the abnormal data performance interval of the second type physiological parameters is recorded as the convolution kernel scale of the second type physiological parameters determined based on the first type physiological parameters; The comprehensive convolution kernel scale of the prostate cancer patient at the target discomfort moment is determined according to the rise-fall ratio of all types of physiological parameters at the target discomfort moment and the convolution kernel scale of the second type of physiological parameters determined by different types of physiological parameters.

9. The prostate cancer recurrence risk prediction system based on multimodal data according to claim 8, characterized in that: The method of determining the comprehensive convolution kernel scale of the prostate cancer patient at the target discomfort moment according to the rise-fall ratio of all types of physiological parameters at the target discomfort moment and the convolution kernel scale of the second type of physiological parameters determined by different types of physiological parameters includes: The product of the convolution kernel scale of the second type of physiological parameter determined based on the first type of physiological parameter and the rise-fall ratio of the first type of physiological parameter of the prostate cancer patient at the target discomfort moment is recorded as the fourth product of the second type of physiological parameter determined by the first type of physiological parameter, and the average of the fourth products of the second type of physiological parameter determined by all physiological parameters different from the second type of physiological parameter is recorded as the adjusted convolution kernel scale of the second type of physiological parameter; The product of the normalized value of the rise-fall ratio of the second type of physiological parameter at the target discomfort moment and the adjusted convolution kernel scale of the second type of physiological parameter is recorded as the fifth product of the second type of physiological parameter at the target discomfort moment, the sum of the fifth products of all types of physiological parameters at the target discomfort moment is recorded as the first cumulative sum at the target discomfort moment, and the odd number closest to the first cumulative sum at the target discomfort moment is recorded as the comprehensive convolution kernel scale of the prostate cancer patient at the target discomfort moment.

10. The prostate cancer recurrence risk prediction system based on multimodal data according to claim 1, characterized in that: The method for predicting the risk of prostate cancer recurrence in prostate cancer patients based on the comprehensive convolution kernel scale includes the following specific methods: The value of the comprehensive convolution kernel scale at the discomfort moment is taken as the value of the convolution kernel scale, and the MSKC model is used to predict the risk of prostate cancer recurrence in prostate cancer patients, so as to obtain the predicted value of the risk of prostate cancer recurrence in prostate cancer patients at the discomfort moment.

Images