Cancer progression assessment method and system thereof

Abstract

The present disclosure relates to a method for cancer progression assessment and a system thereof. The method for cancer progression assessment includes the following steps: inputting a plurality of immune cell population data through an input device, and storing the immune cell population data in a storage device; calculating a plurality of immune entropy values of the immune cell population data, and normalizing the immune entropy values through a processor accessing the storage device, and then applying the normalized immune entropy values to establish a cancer progression assessment model with a decision tree algorithm; obtaining an assessed immune cell population data through the input device and using the processor to perform an interpretation program to obtain an interpretation result of cancer progression; and outputting the interpretation result of cancer progression through an output device. The system includes an input device, a storage device, a processor, and an output device.

Description

[0001] Cancer Progression Assessment Method and System Thereof

[0002] BACKGROUND OF THE INVENTION

[0003] 1. FIELD OF THE PRESENT DISCLOSURE

[0004] The present disclosure relates to a method for cancer progression assessment and a system thereof, more particularly, to a method and a system thereof for assessing immune cell population data obtained through a flow cytometry analysis with a decision tree algorithm, thereby being capable of predicting cancer progression correctly.

[0005] 2.BACKGR0UND

[0006] Due to the changes in diet and rest habits of modem people, or the environmental pollution or radiation caused by the high development of civilization, modem people are prone to various diseases. In serious cases, it may cause mutations in human cells, leading to the abnormal growth of the mutant cells, and then form tumors, such as the continued deterioration and formation of malignant tumors, which is called cancer.

[0007] The World Health Organization (WHO) pointed out that cancer is one of the top causes of death in the world, claiming nearly ten million lives in 2021. In Taiwan, cancer has been the top ten causes of death for decades in a row, and the number of cancer cases is expected to continue to rise due to the rapid aging of the population and unhealthy lifestyle.

[0008] In recent years, with the rapid development of machine learning and the continuous integration of medicine and artificial intelligence, the use of computational methods to assist in the research of related problems in the field of medicine and biology is a powerful tool.

[0009] At present, cancer treatment methods are mainly divided into surgical resection, radiotherapy, chemotherapy, target therapy, and cell therapy. Among them, cytokine-

[0010] 1

[0011] SUBSTITUTE SHEET ( RULE 26) induced killer (CIK) is a treatment with a relatively high rate of success in the current special management method, which is superior to the traditional cancer therapy mentioned above, but there is still only 10 to 30% of the treatment success rate in the world, and there is no evaluable cell therapy-related efficacy prediction system on the market.

[0012] To sum up, there is an urgent need for effective evaluation methods to intervene in the clinical practice of cancer treatment, providing physicians with additional data related to patient immunity as a reference to speed up the selection of medical decisions, improve patients’ treatment willingness and reduce unnecessary waste of resources.

[0013] SUMMARY OF THE INVENTION

[0014] In view of the evaluation problems of conventional cancer treatment mentioned above, the purpose of the present disclosure is directed to a method for cancer progression assessment and a system thereof to reduce the problem of a miscarriage of judgment caused by human evaluation and the difficulty of quickly selecting medical decisions.

[0015] According to one of the purposes of the present disclosure, an embodiment of the present disclosure is directed to a method for cancer progression assessment, which comprises the following steps: inputting a plurality of immune cell population data through an input device, and storing the immune cell population data in a storage device; calculating a plurality of immune entropy values of the immune cell population data, and normalizing the immune entropy values through a processor accessing the storage device, and then applying the normalized immune entropy values of the immune cell population data to establish a cancer progression assessment model with a decision tree algorithm; obtaining an assessed immune cell population data through the input device and using the processor to perform an interpretation program to obtain an interpretation

[0016] 2

[0017] SUBSTITUTE SHEET ( RULE 26) result of cancer progression; and outputting the interpretation result of cancer progression through an output device accessing the storage device.

[0018] The immune cell population data is collected from the peripheral blood samples of cancer patients who are performing CIK treatment before entering the course of treatment, and using a flow cytometry to analyze the peripheral blood samples to obtain the immune cell population data of these cancer patients.

[0019] Normalization is capable of improving the training speed and accuracy of the model. According to the feature points of the immune cell population data and experimental comparison, a min-max normalization method is used to scale the immune entropy values of the immune cell population data to [0,1] in proportion without changing the distribution of the original data. Among them, the immune cell population data is the proportion of different types of cells in the whole cell population.

[0020] The interpretation results of cancer progression are classified according to a response evaluation criteria in solid tumors (RECIST), which are classified into four stages according to the change of tumor size during cancer treatment, namely complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD).

[0021] The immune cell population data is specifically 82 immune cell population data, so it has 82 feature points for machine learning. The types of 82 immune cell population data are shown in Table 1.

[0022] Table 1 : List of 82 immune cell population data according to the present disclosure.

[0023] 3

[0024] SUBSTITUTE SHEET ( RULE 26)

[0025]

[0026] 4

[0027] SUBSTITUTE SHEET (RULE 26)

[0028]

[0029] 5

[0030] SUBSTITUTE SHEET (RULE 26)

[0031] Before establishing the cancer progression assessment model, first way is to divide the peripheral blood samples obtained from different cancer patients into four categories (i.e. CR, PR, SD, and PD) according to the RECIST criteria, and then second way is to calculate the immune entropy values of the immune cell population data (i.e. the 82 feature points) in the blood samples of individual cancer patients and to normalize the immune entropy values. Then, the immune entropy values of each of the normalized immune cell population data are used as an index to train with the decision tree algorithm. By supervised learning, the normalized immune entropy values of the immune cell population data are classified into a training group, a verification group, and a test group, and the model is continuously trained by using a loop method. When the test group reaches the best accuracy rate, find out the key feature points of an optimal decision tree (i.e. key immune cell population data) to obtain the cancer progression assessment model.

[0032] The key feature points are extracted by using the classification nodes of the decision tree as weights, and the contribution and importance of the key feature points are verified by Shapley additive explanation (SHAP). SHAP is used by Python to interpret the prediction of the model. By analyzing each of the feature points, the 6

[0033] SUBSTITUTE SHEET ( RULE 26) Shapley value of each of the feature points is calculated to assess the contribution and importance of each of the feature points to the prediction.

[0034] Finally, the assessed immune cell population data is using the cancer progression assessment model to obtain the interpretation result of cancer progression. From the interpretation result of cancer progression, it may be known that which of the four stages of CR, PR, SD, or PD is the interpretation result of cancer progression of the cancer patient to be assessed.

[0035] According to another purpose of the present disclosure, an embodiment of the present disclosure is directed to a cancer progression assessment system, which comprises an input device, a storage device, a processor, and an output device. Among them, the input device is used to input a plurality of immune cell population data and an assessed immune cell population data. The storage device is connected to the input device to store the immune cell population data and the assessed immune cell population data. The output device is connected to the storage device to output an interpretation result of cancer progression. The processor is connected to the storage device and to execute a plurality of instructions to perform the following steps: calculating a plurality of immune entropy values of the immune cell population data, and normalizing the immune entropy values; applying the normalized immune entropy values of the immune cell population data to establish a cancer progression assessment model with a decision tree algorithm; performing an interpretation program to obtain an interpretation result of cancer progression according to the cancer progression assessment model; and outputting the interpretation result of cancer progression through the output device accessing the storage device.

[0036] As mentioned above, the method for cancer progression assessment and the system thereof are capable of being used to quickly and accurately judge the current cancer progression stage of cancer patients, reduce the burden of doctors and the problem of

[0037] 7

[0038] SUBSTITUTE SHEET ( RULE 26) different doctors’ artificial judgment standards, to facilitate follow-up physicians to accurately evaluate the condition of cancer patients and select the corresponding medical decisions as soon as possible.

[0039] BRIEF DESCRIPTION OF THE DRAWINGS

[0040] To describe the technical solutions in the embodiments of this application more clearly, the following briefly introduces the accompanying drawings required for describing the embodiments. Apparently, the accompanying drawings in the following description show merely some embodiments of this application, and a person of ordinary skill in the art may still derive other drawings from these accompanying drawings without creative efforts.

[0041] Fig. 1 is a flow chart of a method for cancer progression assessment according to one embodiment of the present disclosure.

[0042] Fig. 2 is a schematic diagram of an optimal decision tree structure according to one embodiment of the present disclosure.

[0043] Fig. 3 is a schematic diagram of using SHAP to verify the contribution and importance of the key feature points of the optimal decision tree structure according to one embodiment of the present disclosure, where Xo represents Tc, Xi represents NK, and X2 represents Inter. Monocyte.

[0044] Fig. 4 shows the accuracy of different algorithms for cancer progression assessment.

[0045] Fig. 5 is a schematic diagram of the cancer progression assessment system according to one embodiment of the present disclosure.

[0046] DESCRIPTION OF THE EMBODIMENTS

[0047] In order to help the reviewers understand the technical features, contents, and

[0048] 8

[0049] SUBSTITUTE SHEET ( RULE 26) advantages of the present disclosure and the effects it can achieve, the present disclosure is hereby described in detail with the accompanying drawings and the expression form of the embodiment as follows. The main purpose of the schema used is only for the purpose of schematic and auxiliary explanation and is not necessarily the true scale and accurate configuration of the present disclosure after implementation. Therefore, the relationship between the scale and configuration of the attached schema should not be interpreted the scope of rights of the present disclosure in practical implementation is limited, which will be described in advance.

[0050] Unless otherwise defined, all terms used herein (including technical and scientific terms) have the meanings commonly understood by those who are generally knowledgeable in the technical field to which the present disclosure belongs. It will be further understood that terms such as those defined in commonly used dictionaries should be interpreted as having meanings consistent with their meanings in the context of the relevant technology and the present disclosure, and will not be interpreted as idealized or overly formal unless explicitly defined herein.

[0051] Please refer to Fig. 1, which is a flow chart of a method for cancer progression assessment according to one embodiment of the present disclosure. As shown in Fig. 1, the method for cancer progression assessment comprises the following steps (S1-S4):

[0052] SI: inputting a plurality of immune cell population data through an input device, and storing the immune cell population data in a storage device.

[0053] Input the collected immune cell population data into the storage device of the system through the input device. The input device described here is not limited to the flow cytometer that obtains the immune cell population data. The immune cell population data stored in the database of the medical hospital may also be transmitted through the physical line and the file of the storage device. Or, the file of the immune cell population data may be transmitted through wired or wireless network and input

[0054] 9

[0055] SUBSTITUTE SHEET ( RULE 26) into the system database as training data for model construction.

[0056] Among them, the immune cell population data is collected from the peripheral blood samples of cancer patients who are performing CIK treatment before entering the course of treatment, and using a flow cytometry to analyze the peripheral blood samples to obtain the immune cell population data of these cancer patients. Therefore, the peripheral blood samples of the cancer patients have a plurality of immune cell population data, which is specifically 82 immune cell population data.

[0057] In addition, before establishing a cancer progression assessment model, the peripheral blood samples obtained from different cancer patients were first classified into four stages (i.e. CR, PR, SD, and PD) according to the RECIST criteria.

[0058] S2: calculating a plurality of immune entropy values of the immune cell population data, and normalizing the immune entropy values through a processor accessing the storage device, and then applying the normalized immune entropy values of the immune cell population data to establish the cancer progression assessment model with a decision tree algorithm.

[0059] The processor reads the immune cell population data stored in the storage device, calculates the immune entropy values of the immune cell population data (that is, the 82 immune cell population data, as shown in Table. 1 above) in the blood samples of individual cancer patients, and normalizes the immune entropy values of the immune cell population data using a min-max normalization method to scale the immune entropy values of the immune cell population data to [0,1] in proportion. The training speed and accuracy of the model are capable of being improved after normalization. The calculation method of immune entropy values is as follows:

[0060] 10

[0061] SUBSTITUTE SHEET ( RULE 26) ( Entropy )

[0062] In the above formula, P represents the probability, Bl, B2, ... to Bn respectively represents the first sample value, the second sample value, ... to the nth sample value of the B cell population, and the rest of the cell populations are the same, which will not be repeated here. There are 82 immune entropy values of each of the immune cell population data calculated by the above formula.

[0063] Next, take the immune entropy values of the normalized immune cell population data (82 in total, so it is equivalent to 82 feature points) as an index, train the model with the decision tree algorithm, and use a loop method to train the model continuously. When a test group reaches the best accuracy rate, find out the key feature points of an optimal decision tree (i.e. key immune cell population data) to obtain the cancer progression assessment model.

[0064] Among them, the immune entropy values of these normalized immune cell population data are classified into a training group, a verification group, and a test group, with the distribution proportion of 64% in the training group, 16% in the verification group, and 20% in the test group, i.e. (training group+verification group):test group=8:2, while training group: test group=8:2.

[0065] 11

[0066] SUBSTITUTE SHEET ( RULE 26) Please refer to Fig. 2, which is a schematic diagram of an optimal decision tree structure according to one embodiment of the present disclosure. As shown in Fig. 2, the key feature points of the optimal decision tree are extracted by using the classification node of the decision tree as weights. Three important key feature points are selected, namely CD3+CD8+Tc, CD3 CD56+CD16+NK, and CD14++CD16+Inter. Monocyte, and use SHAP to verify the contribution and importance of these key features.

[0067] Please refer to Fig. 3, which is a schematic diagram of using SHAP to verify the contribution and importance of the key feature points of the optimal decision tree structure according to one embodiment of the present disclosure, where Xo represents Tc, Xi represents NK, and X2 represents Inter. Monocyte. As shown in Fig. 3, the contribution of Tc, NK, and Inter, monocyte in each node is significantly different. Inter, monocyte is the most important for cancer patients classified as PR stage, NK is the most important for cancer patients classified as SD stage, and Tc is the most important for cancer patients classified as CR stage and PD stage. It can be seen that these key feature points are important judgment factors to judge the four stages of CR, PR, SD, or PD.

[0068] S3 : obtaining an assessed immune cell population data through the input device and using the processor to perform an interpretation program to obtain an interpretation result of cancer progression.

[0069] The cancer progression assessment model established in S2 mentioned above is used to evaluate the assessed immune cell population data and obtain the cancer progression interpretation result. According to the interpretation result of cancer progression, we can know which of the four stages of cancer progression is CR, PR, SD, or PD. The input device described here is the same as that described above, and will not be repeated here.

[0070] 12

[0071] SUBSTITUTE SHEET ( RULE 26) Please refer to Fig.4, which shows the accuracy of different algorithms for cancer progression assessment. As shown in Fig. 4, the immune cell population data (i.e. the 82 immune cell population data) are used to establish a cancer progression assessment model using other machine learning algorithms such as support vector machine (SVM) and K-nearest neighbor algorithm (KNN). The accuracy of the model is far lower than that of the decision tree algorithm. The accuracy of SVM is only 42.85%, the accuracy of KNN is slightly higher than 64.28%, and the accuracy of the decision tree is as high as 85.71%. It can be seen that the cancer progression assessment model established by the decision tree algorithm is superior to other kinds of algorithms.

[0072] S4: outputting the interpretation result of cancer progression through an output device accessing the storage device.

[0073] The cancer progression interpretation result obtained through S3 mentioned above may be further output through the output device. The output device disclosed in this embodiment may include various display interfaces, such as computer screen, display, or handheld device display.

[0074] Please refer to Fig. 5, which is a schematic diagram of the cancer progression assessment system according to one embodiment of the present disclosure. As shown in Fig. 5, the cancer progression assessment system 20 may comprise an input device 21, a storage device 22, a processor 23, and an output device 24.

[0075] In this embodiment, the input device 21 is a flow cytometer, which collects and analyzes the peripheral blood samples of cancer patients who are performing CIK treatment before entering the course of treatment to obtain the immune cell population data. In another embodiment, the input device 21 is not limited to a flow cytometer, and the input device 21 may include input interfaces of electronic devices such as personal computers, smart phones, servers, etc., including touch screens, keyboards, mice, etc., through the immune cell population data is transmitted as a file; or the historical data is 13

[0076] SUBSTITUTE SHEET ( RULE 26) uploaded to the memory storage in the storage device 22 through wireless network transmission, wireless communication transmission, or general wired Internet. The memory may include read-only memory, flash memory, disk, cloud database, etc.

[0077] Next, the cancer progression assessment system 20 accesses the storage device 22 through the processor 23. The processor 23 may include the central processing unit, image processor, microprocessor, etc. in the computer or server, and it may include multi-core processing units or a combination of multiple processing units. The processor 23 executes instructions to access the immune cell population data in the storage device 22 for training programs, and simultaneously takes the assessed immune cell population data for interpretation programs. In detail, the training program is to normalize the immune cell population data in the original storage device 22 by calculating the immune entropy values of the immune cell population data, and then calculate the immune entropy values of the normalized immune cell population data with the decision tree algorithm to establish the cancer progression assessment model.

[0078] Next, after the assessed immune cell population data is calculated through the established cancer progression assessment model through the interpretation procedure, the assessed immune cell population data is classified into one of CR, PR, SD and PD according to the RECIST criteria, and the interpretation result of cancer progression is obtained. The output device 24 accesses the storage device 22 and outputs the results of cancer progression interpretation. The output device 24 may include various display interfaces, such as computer screen, display, or handheld device display.

[0079] Through the above cancer progression assessment methods and systems, the workload of doctors is capable of being significantly reduced, and the errors of manual interpretation may be reduced, which will lead to the deviation of cancer progression diagnosis. Furthermore, using the cancer progression assessment method and the system of the present disclosure, the current cancer progression stage of cancer patients 14

[0080] SUBSTITUTE SHEET ( RULE 26) may be quickly and accurately assessed, with an accuracy of more than 80%, so that follow-up physicians can correctly assess the condition of cancer patients and make appropriate medical decisions as soon as possible. The above is only illustrative, not restrictive. Any equivalent modification or change to the present disclosure that does not deviate from the spirit and scope of the present disclosure shall be included in the attached scope of the patent application.

[0081] 15

[0082] SUBSTITUTE SHEET ( RULE 26)

Claims

AMENDED CLAIMS received by the International Bureau on 7 February 2024 (07.02.2024)AMENDMENTS TO THE CLAIMSThis listing of claims will replace all prior versions and listings of claims in the application:Listing of Claims1 . (Currently amended) A method for cancer progression assessment, which comprises:S1 : inputting a plurality of immune cell population data through an input device, and storing the immune cell population data in a storage device;S2: calculating a plurality of immune entropy values of the immune cell population data, and normalizing the immune entropy values through a processor accessing the storage device, and then applying the normalized immune entropy values of the immune cell population data to establish a cancer progression assessment model with a decision tree algorithm;S3: obtaining an assessed immune cell population data through the input device and using the processor to perform an interpretation program to obtain an interpretation result of cancer progression; andS4: outputting the interpretation result of cancer progression through an output device accessing the storage device; wherein, the immune cell population data are classified into four categories according to a response evaluation criteria in solid tumors (RECIST): complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD)[[.]] i wherein the calculating of the immune entropy values comprises:AMENDED SHEET (ARTICLE 19)wherein the cancer progress assessment model comprises an optimal decision tree with three key immune cell populations, and the three key immune cell populations are a cytotoxic T cell (Tc), a natural killer cell (NK), and an intermediate monocyte (Inter. Monocyte).

2. The method for cancer progression assessment as claimed in claim 1 , wherein the input device is a flow cytometry.

3. (Currently amended) The method for cancer progression assessment as claimed in claim 1 , wherein the immune cell population data is specifically 82 immune cell population data[[.]] , which comprises:B cell, T cell, T helper cell, Active T helper cell, Naive T helper cell, Memory T helper cell, Type 1 T helper cell, Naive Type 1 T helper cell, Memory Type 1 T helper cell, Type 2 T helper cell, Naive Type 2 T helper cell, Memory Type 2 T helper cell, regulatory T cell, Naive regulatory T cell, Memory regulatory T cell, Type 17 T helper cell, Naive Type 17 T helper cell, Memory Type 17 T helper cell, Type 22 T helper cell,20AMENDED SHEET (ARTICLE 19)Naive Type 22 T helper cell, Memory Type 22 T helper cell, follicular helper T cell, Naive follicular helper T cell, Memory follicular helper T cell, cytotoxic T cell, Active cytotoxic T cell, Naive cytotoxic T cell, Memory cytotoxic T cell, Regulatory cytotoxic T cell, double positive T cell, double negative T cell, natural killer cell, CD56b natural killer cell, HLADR+CD56b natural killer cell, T-bet+CD56b natural killer cell, IFNG+CD56b natural killer cell, Perf+CD56b natural killer cell, GranB+CD56b natural killer cell, CD11a+CD56b natural killer cell, CD94+CD56b natural killer cell, CD107a+CD56b natural killer cell, CD56d natural killer cell, HLA-DR+CD56d natural killer cell, T-bet+CD56d natural killer cell, IFNG+CD56d natural killer cell, Perf natural killer cell, GranB natural killer cell, CD11a+CD56d natural killer cell, CD94+CD56d natural killer cell, CD107a+CD56d natural killer cell, CD56b cell, CD56d cell, natural killer T cell, natural killer T cell CD8, natural killer T cell CD4, natural killer T cell double positive, natural killer T cell double negative, CD56b natural killer T cell, HLA- DR+CD56b natural killer T cell, T-bet+CD56b natural killer T cell, IFNG+CD56b natural killer T cell, Perf+CD56b natural killer T cell, GranB+CD56b natural killer T cell, CD11a+CD56b natural killer T cell, CD94+CD56b natural killer T cell, CD107a+CD56b natural killer T cell, CD56d natural killer T cell, HLA-DR+CD56d natural killer T cell, T- bet+CD56d natural killer T cell, IFNG+CD56d natural killer T cell, Perf+CD56d natural killer T cell, GranB+CD56d natural killer T cell, CD11a+CD56d natural killer T cell, CD94+CD56d natural killer T cell, CD107a+CD56d natural killer T cell, CD56b CD16' cell, CD56d CD16' cell, mature dendritic cell, monocyte, classical monocyte, non-classical monocyte, and intermediate monocyte.21AMENDED SHEET (ARTICLE 19)4. The method for cancer progression assessment as claimed in any one of claims 1 to 3, wherein the method for normalizing the immune entropy values of the immune cell population data is to use a min-max normalization method.

5. The method for cancer progression assessment as claimed in any one of claims 1 to 3, wherein the normalized immune entropy values of the immune cell population data are classified into a training group, a verification group, and a test group, and a distribution proportion is 64% in the training group, 16% in the verification group, and 20% in the test group.

6. (Currently amended) The method for cancer progression assessment as claimed in claim [[7]] 1 , wherein the cytotoxic T cell (Tc) is the most important for classifying complete response (CR), and progressive disease (PD); the natural killer cell (NK) is the most important for classifying stable disease (SD); and the intermediate monocyte (Inter. Monocyte) is the most important for classifying partial response (PR).AMENDED SHEET (ARTICLE 19)

Images