Breath gas data analysis method for breast cancer based on non-invasive breath biopsy technology

Abstract

The application discloses a breast cancer respiratory gas data analysis method, system, device and computer readable storage medium based on non-invasive breath biopsy technology, and relates to the technical field of medical treatment. The method comprises the following steps: obtaining respiratory gas of a sample to be measured; processing the respiratory gas of the sample to be measured to obtain target VOC ion data; inputting the target VOC ion data into a constructed breast cancer prediction model I to obtain a classification result; or: obtaining respiratory gas and clinical data of a sample to be measured, wherein the clinical data comprises one or more of the following: age, BMI, family cancer history, personal cancer history and menopausal status; processing the respiratory gas of the sample to be measured to obtain target VOC ion data; and inputting the target VOC ion data and the clinical data into a constructed breast cancer prediction model II to obtain a classification result.

Description

Technical Field

[0001] The present invention relates to the medical field, and more particularly to a method and system for analyzing breast cancer respiratory gas data based on non-invasive exhaled breath biopsy technology. Background Art

[0002] Breast cancer is currently one of the most common cancers and a leading cause of death worldwide. Early detection is crucial to improving survival rates for breast cancer patients. Widespread screening—primarily population-based screening using imaging-based approaches, including mammography and ultrasound—is leading to an increasing number of breast cancers being diagnosed at an early stage; however, the proportion of breast cancers detected through screening varies from 1% to 50% across countries, partially limited by individuals' willingness to undergo active medical examinations and the availability of breast cancer screening.

[0003] The accuracy and reproducibility of imaging-based breast cancer screening depend on the experience of the diagnostic medical operator and the reader. In addition, the lack of a consistent supply of experienced radiologists and sonographers limits the consistency of breast cancer screening; this challenge is exacerbated by the high screening volume and costs in each country. Maintaining the cost-effectiveness of imaging-based breast cancer screening is also challenging, given that reducing breast cancer deaths may come at the expense of overdiagnosis and the consequences of false-positive or false-negative results. It has been reported that the cumulative risk of a false-positive result for a woman aged 40 to 50 years who undergoes annual mammography for 10 years exceeds 60%; in particular, the high-density breast tissue of Asian women further reduces the image quality of mammographic examinations. Therefore, there is an urgent need for an accurate and easy-to-perform breast cancer screening method that does not rely on clinicians.

[0004] Along with alterations in the genome and transcriptome, reprogrammed cellular metabolism is considered a "new hallmark of cancer." Metabolic alterations and accumulation of abnormal metabolites have also been observed in different molecular subtypes and histological types of breast cancer. Among the cancer-derived metabolites that diffuse into the alveoli, volatile organic compounds (VOCs) can be detected in exhaled breath. Therefore, analyzing VOCs in exhaled breath is a promising strategy for noninvasive screening and early diagnosis of breast cancer. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention provides a method and system for analyzing breast cancer respiratory gas data based on non-invasive breath biopsy technology. The method utilizes volatile organic compounds (VOCs) or integrates VOCs with clinical data to construct a breast cancer prediction model based on non-invasive breath biopsy technology for non-invasive screening and early diagnosis of breast cancer. By targeting VOCs for breast cancer screening, the method continuously improves the early diagnosis of breast cancer patients, deeply explores the life laws hidden behind the data, and solves related life science problems.

[0006] The first aspect of the present application discloses a method for constructing a breast cancer prediction model based on non-invasive breath biopsy technology, comprising:

[0007] Obtaining respiratory gas and classification labels of training set samples, wherein the classification labels include breast cancer patients and non-breast cancer patients;

[0008] Detecting and analyzing the respiratory gas of the training set samples to obtain processed mass spectrometry data;

[0009] converting the discrete signals of the processed mass spectrometry data into standard VOC features;

[0010] Sorting the standard VOC features to obtain target VOC ion data;

[0011] The target VOC ion data is input into the machine learning model to obtain the predicted classification results, which are compared with the classification labels. The model is optimized based on the comparison results to obtain the constructed breast cancer prediction model 1.

[0012] Sorting the screened VOC features according to feature importance or coefficient;

[0013] Optionally, VOC features with a<m / z<b are screened from the standard VOC features, and the VOC features are sorted to obtain target VOC ion data, wherein the range of a is: 0-50, preferably 20; the range of b is: 100-500, preferably 320;

[0014] Optionally, the method of converting the discrete signal of the mass spectrum data into a standard VOC feature includes: calculating the area of ​​the most important peak in the range of [x-0.1, x+0.1] and treating it as the VOC feature with m / z x, and the VOC feature with m / z x as the standard VOC feature;

[0015] The method for constructing the breast cancer prediction model further includes:

[0016] Acquire clinical data of training set samples, perform feature extraction on the clinical data, and obtain clinical data features;

[0017] The target VOC ion data and clinical data features are subjected to feature fusion processing to obtain a feature set; the feature set is input into a machine learning model to obtain a prediction classification result, which is compared with the classification label, and the model is optimized according to the comparison result to obtain a constructed breast cancer prediction model 2;

[0018] Optionally, the clinical data includes one or more of the following: age, BMI, family cancer history, personal cancer history and menopausal status.

[0019] The second aspect of the present application discloses a method for analyzing breast cancer respiratory gas data based on non-invasive exhaled breath biopsy technology, the method comprising:

[0020] Obtaining respiratory gas of a sample to be tested;

[0021] Processing the respiratory gas of the sample to be tested to obtain target VOC ion data;

[0022] Inputting the target VOC ion data into the constructed breast cancer prediction model 1 to obtain a classification result;

[0023] Optionally, the method further includes:

[0024] Obtaining respiratory gas and clinical data of the sample to be tested, wherein the clinical data includes one or more of the following: age, BMI, family cancer history, personal cancer history, and menopausal status;

[0025] Processing the respiratory gas of the sample to be tested to obtain target VOC ion data;

[0026] The target VOC ion data and clinical data are input into the constructed breast cancer prediction model 2 to obtain a classification result.

[0027] The third aspect of the present application discloses a method for analyzing breast cancer respiratory gas data based on non-invasive breath biopsy technology, comprising:

[0028] Obtaining respiratory gas of a sample to be tested;

[0029] The respiratory gas of the sample to be tested is processed to obtain target VOC ion data; the target VOC ion data includes one or more of the following: m / z(VOC1)=28.0, m / z(VOC2)=32.2, m / z(VOC3)=40.0, m / z(VOC4)=101.0, m / z(VOC5)=112.0, m / z(VOC6)=123.0, m / z(VOC7)=126.0, m / z(VOC8)=132.0, m / z(VOC9)=92.0, m / z(VOC10)=103.0;

[0030] Inputting the target VOC ion data into the constructed breast cancer prediction model 1 to obtain a classification result;

[0031] Optionally, the method further includes:

[0032] Obtaining respiratory gas and clinical data of the sample to be tested, wherein the clinical data includes one or more of the following: age, BMI, family cancer history, personal cancer history, and menopausal status;

[0033] Processing the respiratory gas of the sample to be tested to obtain target VOC ion data;

[0034] Inputting the target VOC ion data and the clinical data into the constructed breast cancer prediction model 2 to obtain a classification result;

[0035] Optionally, the process of processing the respiratory gas of the sample to be tested includes:

[0036] Detecting and analyzing the respiratory gas of the sample to be tested to obtain processed mass spectrometry data;

[0037] converting the discrete signals of the processed mass spectrometry data into standard VOC features;

[0038] VOC features with a<m / z<b are screened from the standard VOC features; the screened VOC features are sorted according to feature importance or coefficient to obtain the target VOC ion data.

[0039] The fourth aspect of the present application discloses a use of a substance for detecting organic compounds in the preparation of a product for diagnosing or assisting in the diagnosis of breast cancer; the organic compound comprises one or more of the following: m / z (VOC1) = 28.0, m / z (VOC2) = 32.2, m / z (VOC3) = 40.0, m / z (VOC4) =

[0040] =101.0, m / z(VOC5)=112.0, m / z(VOC6)=123.0, m / z(VOC7)=126.0, m / z(VOC8)=132.0, m / z(VOC9)=92.0, m / z(VOC10)=103.0.

[0041] In a fifth aspect, the present application discloses a product for diagnosing or assisting in the diagnosis of breast cancer, comprising a substance for detecting organic compounds; the organic compounds comprise one or more of the following: m / z(VOC1)=28.0, m / z(VOC2)=32.2, m / z(VOC3)=40.0, m / z(VOC4)=101.0, m / z(VOC5)=112.0, m / z(VOC6)=123.0, m / z(VOC7)=126.0, m / z(VOC8)=132.0, m / z(VOC9)=92.0, m / z(VOC10)=103.0.

[0042] The breast cancer respiratory gas data analysis system based on non-invasive exhaled breath biopsy technology includes:

[0043] an acquisition unit, used for acquiring the respiratory gas of the sample to be tested;

[0044] a first processing unit, configured to process the respiratory gas of the sample to be tested to obtain target VOC ion data;

[0045] The second processing unit is used to input the target VOC ion data into the constructed breast cancer prediction model 1 for processing to obtain a classification result.

[0046] A breast cancer respiratory gas data analysis device based on non-invasive exhaled breath biopsy technology, the device comprising:

[0047] memory and processor;

[0048] The memory is used to store program instructions; the processor is used to call the program instructions, and when the program instructions are executed, it is used to execute the above-mentioned breast cancer respiratory gas data analysis method based on non-invasive exhaled breath biopsy technology.

[0049] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned breast cancer respiratory gas data analysis method based on non-invasive exhaled breath biopsy technology.

[0050] This application has the following beneficial effects:

[0051] 1. This application innovatively discloses a method for analyzing breast cancer breath gas data based on non-invasive exhaled breath biopsy technology. A model is constructed based on target volatile organic compounds screened from breath gas to obtain breast cancer prediction model 1, or a model is constructed using breath gas and related clinical data to obtain breast cancer prediction model 2. Breast cancer prediction model 1 and breast cancer prediction model 2 are used for non-invasive screening and early diagnosis of breast cancer. The method standardizes the collection method of breath gas, simplifies the breathomics detection procedure, improves the sensitivity of VOC detection, and demonstrates the clinical effectiveness of VOC analysis in breast cancer screening and early diagnosis. This method deeply explores the life laws hidden behind the data and conducts in-depth analysis from multiple dimensions, greatly improving the accuracy and depth of data analysis, and improving the efficiency and accuracy of breath-based breast cancer screening. It also overcomes the problems of existing imaging-based breast cancer screening and early diagnosis technologies, including mammography and ultrasound, which are prone to high reliance on radiologists and ultrasound physicians, cost, and underdiagnosis / overdiagnosis rates.

[0052] 2. This application innovatively achieves performance and high detection rates for early breast cancer and DCIS that are comparable to those of traditional imaging-based strategies based on the data of the 10 best target VOC ions, or based on the data of the 10 best target VOC ions and clinical data. It provides a complementary or alternative screening strategy to reduce false positive rates, reliance on medical imaging experts, and screening costs, simplify detection procedures, provide early intervention treatment for breast cancer patients, and improve the life and quality of life of the subjects. This method deeply explores the life laws hidden behind medical data and solves related life science problems.

[0053] 3. This application can make good use of target VOC ion data to detect whether the subject is a breast cancer patient; at the same time, it also proposes the use of substances for detecting organic compounds in the preparation of products for diagnosing or assisting in the diagnosis of breast cancer, as well as products for diagnosing or assisting in the diagnosis of breast cancer, to achieve more accurate and faster treatment of subjects. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0055] Figure 1 This is a schematic flow chart of a method for analyzing respiratory gas data of breast cancer based on non-invasive breath biopsy technology provided in the second aspect of an embodiment of the present invention;

[0056] Figure 2 Schematic diagram of a breast cancer respiratory gas data analysis device based on non-invasive breath biopsy technology provided by an embodiment of the present invention;

[0057] Figure 3 This is a schematic flow chart of a breast cancer respiratory gas data analysis system based on non-invasive breath biopsy technology provided by an embodiment of the present invention;

[0058] Figure 4 This is a subject registration and study design diagram provided by an embodiment of the present invention;

[0059] Figure 5 This is a distribution diagram of target VOC ion data in the data analysis workflow and model construction provided by an embodiment of the present invention;

[0060] Figure 6 Schematic diagram of the performance of the breast cancer detection models of the BreathBC model and BreathBC-Plus model provided in the embodiments of the present invention in an internal validation cohort, a test cohort, and three external validation cohorts. DETAILED DESCRIPTION

[0061] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0062] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., and do not represent the order of precedence, nor do they limit "first" and "second" to be different types.

[0063] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0064] Figure 1This is a schematic flow chart of a method for analyzing respiratory gas data of breast cancer based on non-invasive breath biopsy technology according to a second aspect of an embodiment of the present invention. Specifically, the method comprises the following steps:

[0065] 101: Obtaining respiratory gas of a sample to be tested;

[0066] In one embodiment, standardized sampling requirements and protocols were established for the test samples to minimize the influence of lifestyle factors such as diet, smoking, alcohol consumption, and environmental factors. After a deep nasal inhalation, the participant exhaled completely into a 1.2L polyetheretherketone sampling bag (a self-designed collector and air bag were used to collect breath samples) through a disposable gas nipple. HPPI-TOFMS was used to detect and analyze the breath samples. Breath samples were directly introduced through a stainless steel capillary (250 μm diameter, 0.60 m length); the test sample was the subject.

[0067] In one embodiment, time-of-flight mass spectrometry (HPPI-TOFMS) analyzes human exhaled breath and achieves highly sensitive detection of substances such as acetone, isoprene, dimethyl sulfide, ethanol, acetic acid, etc. in human exhaled breath within a sampling time of 1 minute. Under 100% relative humidity conditions, this method can achieve direct online detection of volatile organic compounds (VOCs) with concentrations as low as tens of pptv, and has ultra-high sensitivity and selectivity for benzene series, furans, sulfur-containing and nitrogen-containing compounds. Compared with other exhaled breath detection methods, the biggest advantage of this technology is that it is less affected by humidity. HPPI-TOFMS is applied to human exhaled breath analysis. It is simple, fast, and does not require any pre-treatment. It is expected to become a high-throughput disease screening tool and can be used for large data sample analysis in clinical practice. Moreover, the accumulation of low-cost, high-throughput, and large-sample data is conducive to the discovery of new disease biomarkers.

[0068] 102: Processing the respiratory gas of the sample to be tested to obtain target VOC ion data;

[0069] In one embodiment, the process of processing the respiratory gas of the sample to be tested includes:

[0070] Detecting and analyzing the respiratory gas of the sample to be tested to obtain processed mass spectrometry data;

[0071] converting the discrete signals of the processed mass spectrometry data into standard VOC features;

[0072] VOC features with a < m / z < b are screened from the standard VOC features; the screened VOC features are sorted according to feature importance or coefficient to obtain the target VOC ion data. The range of a is 0-50, preferably 20; the range of b is 100-500, preferably 320.

[0073] In one embodiment, m / z is the ratio of the number of protons to the number of charges. The mass-to-charge ratio is an important parameter in mass spectrometry. The radius r of the arc trajectory formed by ions of different m / e values ​​under a certain acceleration voltage V and a certain magnetic field strength E is proportional to the m / e ratio.

[0074] In the 1990s, IUPAC mandated that the mass-to-charge ratio be replaced by m / z instead of m / e. Mass spectrometry is an identification technique that plays a crucial role in the identification of organic molecules. It can rapidly and accurately determine the molecular weight of biological macromolecules, enabling proteomics research to move beyond protein identification to the study of higher-order structures and interactions between proteins.

[0075] 103: Inputting the target VOC ion data into the constructed breast cancer prediction model 1 to obtain a classification result;

[0076] In one embodiment, a first aspect discloses a method for constructing a breast cancer prediction model based on non-invasive breath biopsy technology, comprising:

[0077] Obtaining respiratory gas and classification labels of training set samples, wherein the classification labels include breast cancer patients and non-breast cancer patients;

[0078] Detecting and analyzing the respiratory gas of the training set samples to obtain processed mass spectrometry data;

[0079] converting the discrete signals of the processed mass spectrometry data into standard VOC features;

[0080] Sorting the standard VOC features to obtain target VOC ion data;

[0081] The target VOC ion data is input into the machine learning model to obtain the predicted classification results, which are compared with the classification labels. The model is optimized based on the comparison results to obtain the constructed breast cancer prediction model 1.

[0082] In one embodiment, the screened VOC features are sorted according to feature importance or coefficient; optionally, VOC features with a < m / z < b are screened from the standard VOC features, and the VOC features are sorted to obtain target VOC ion data, wherein the range of a is: 0-50, preferably 20; the range of b is: 100-500, preferably 320; the screened VOC features include 1500, and the time interval is 0.2;

[0083] Optionally, the method of converting the discrete signal of the mass spectrum data into a standard VOC feature includes: calculating the area of ​​the most important peak in the range of [x-0.1, x+0.1] and treating it as the VOC feature with m / z x, and the VOC feature with m / z x as the standard VOC feature;

[0084] In one embodiment, the method for detecting and analyzing the respiratory gas of the training set samples includes: introducing the respiratory gas through a catheter (a stainless steel capillary (250 micrometers in diameter and 0.60 meters in length)), detecting and analyzing the respiratory gas using a HPPI-TOFMS to obtain mass spectrometry data, and accumulating the data for 60 seconds; detecting the mass spectrometry data using a time-to-digital converter, screening for peaks with m / z < 350, and accumulating the data for 60 seconds; and performing noise reduction and baseline correction on the mass spectrometry data of the peaks with m / z < 350 using the Python software package pywavelets to obtain processed mass spectrometry data;

[0085] In one embodiment, the peak area of ​​each target VOC ion data is also calculated; the peak areas of 8 VOCs are increased in breast cancer patients, and the peak areas of 2 VOCs (m / z=92 and 103) are decreased in the breast cancer group.

[0086] In one embodiment, the method for constructing a breast cancer prediction model further comprises:

[0087] Acquire clinical data of training set samples, perform feature extraction on the clinical data, and obtain clinical data features;

[0088] The target VOC ion data and clinical data features are subjected to feature fusion processing to obtain a feature set; the feature set is input into a machine learning model to obtain a prediction classification result, which is compared with the classification label, and the model is optimized according to the comparison result to obtain a constructed breast cancer prediction model 2;

[0089] Optionally, the clinical data includes one or more of the following: age, BMI, family cancer history, personal cancer history and menopausal status.

[0090] In one embodiment, the method further comprises:

[0091] Obtaining respiratory gas and clinical data of the sample to be tested, wherein the clinical data includes one or more of the following: age, BMI, family cancer history, personal cancer history, and menopausal status;

[0092] Processing the respiratory gas of the sample to be tested to obtain target VOC ion data;

[0093] The target VOC ion data and clinical data are input into the constructed breast cancer prediction model 2 to obtain a classification result.

[0094] In one embodiment, the training set samples included women who underwent breast cancer screening at six hospitals in China between December 19, 2020, and January 21, 2022. Participants were divided into a discovery cohort, which we used to identify candidate VOCs and build a diagnostic model, and an external validation cohort, which we used to test the diagnostic value of the model. The discovery cohort recruited women who underwent opportunistic breast cancer screening at the Cancer Hospital of the Chinese Academy of Medical Sciences and Peking Union Medical College Hospital, as well as women from the Beijing Urban Breast Cancer Screening Program, a population-based program in Yanqing and Daxing districts of Beijing (northern China). The external validation cohort recruited women who underwent opportunistic breast cancer screening at Yantai Yuhuangding Hospital Affiliated to Qingdao University (Yantai cohort in East China) and the First Affiliated Hospital of Wenzhou Medical University (Wenzhou cohort in Southeast China), as well as women from the Guiyang China Urban Population Breast Cancer Screening Program (Guiyang cohort in Southwest China). The study was reviewed and approved by the ethics committee of each participating hospital. Written informed consent was obtained from each participant. This study adhered to the reporting guidelines of the Standards for Reporting Diagnostic Accuracy. Participants and clinical assessment: Women aged ≥18 years who underwent opportunistic breast cancer screening and women aged ≥30 years who participated in a population-based breast cancer screening program were included. Women with a history of malignancy or anticancer treatment within the past 5 years were excluded. Information on breast cancer risk factors was collected for each participant, including age, body mass index (BMI), family cancer history, personal cancer history, and menopausal status. Each center performed standard mammography and ultrasound examinations and used the Breast Imaging-Reporting and Data System (BI-RADS) for classification screening. Women with suspected malignant breast lesions underwent surgery or core needle biopsy. Each breast cancer patient was diagnosed based on pathological results. Women with no suspicious breast lesions or negative biopsy results at 6 months of follow-up were excluded from breast cancer. The stage and molecular subtype of breast cancer were also assessed.

[0095] In one embodiment, a machine learning method is used to construct a model of the respiratory gas to obtain the constructed prediction model;

[0096] Optionally, the machine learning method includes one or more of the following: partial least squares regression, least squares support vector machine, neural network, random forest method, linear regression, logistic regression, linear discriminant analysis, classification and regression tree, naive Bayes, KNN, learning vector quantization, support vector machine, LightGBM, extreme gradient boosting; the machine learning method is preferably: random forest method.

[0097] In one embodiment, during the model building process, the discovery dataset was randomly divided into training, internal validation, and test datasets in a ratio of 5:2:3. The internal validation dataset was used to validate the model and confirm that the optimal sensitivity and specificity cutoff values ​​of >60% were achieved. The test dataset was used to test label anonymization. In this study, we constructed two breast cancer detection models, BreathBC (breast cancer prediction model one) and BreathBC-Plus (breast cancer prediction model two), the former using only breath VOC markers and the latter using both breath VOC markers and risk factors. Both models were validated using three external validation cohorts.

[0098] The third aspect of the present application discloses a method for analyzing breast cancer respiratory gas data based on non-invasive breath biopsy technology, comprising:

[0099] Obtaining respiratory gas of a sample to be tested;

[0100] The respiratory gas of the sample to be tested is processed to obtain target VOC ion data; the target VOC ion data includes one or more of the following: m / z(VOC1)=28.0, m / z(VOC2)=32.2, m / z(VOC3)=40.0, m / z(VOC4)=101.0, m / z(VOC5)=112.0, m / z(VOC6)=123.0, m / z(VOC7)=126.0, m / z(VOC8)=132.0, m / z(VOC9)=92.0, m / z(VOC10)=103.0;

[0101] Inputting the target VOC ion data into the constructed breast cancer prediction model 1 to obtain a classification result;

[0102] Optionally, the method further includes:

[0103] Obtaining respiratory gas and clinical data of the sample to be tested, wherein the clinical data includes one or more of the following: age, BMI, family cancer history, personal cancer history, and menopausal status;

[0104] Processing the respiratory gas of the sample to be tested to obtain target VOC ion data;

[0105] Inputting the target VOC ion data and the clinical data into the constructed breast cancer prediction model 2 to obtain a classification result;

[0106] Optionally, the process of processing the respiratory gas of the sample to be tested includes:

[0107] Detecting and analyzing the respiratory gas of the sample to be tested to obtain processed mass spectrometry data;

[0108] converting the discrete signals of the processed mass spectrometry data into standard VOC features;

[0109] VOC features with a<m / z<b are screened from the standard VOC features; the screened VOC features are sorted according to feature importance or coefficient to obtain the target VOC ion data.

[0110] The fourth aspect of the present application discloses a use of a substance for detecting organic compounds in the preparation of a product for diagnosing or assisting in the diagnosis of breast cancer; the organic compound comprises one or more of the following: m / z (VOC1) = 28.0, m / z (VOC2) = 32.2, m / z (VOC3) = 40.0, m / z (VOC4) =

[0111] =101.0, m / z(VOC5)=112.0, m / z(VOC6)=123.0, m / z(VOC7)=126.0, m / z(VOC8)=132.0, m / z(VOC9)=92.0, m / z(VOC10)=103.0.

[0112] In a fifth aspect, the present application discloses a product for diagnosing or assisting in the diagnosis of breast cancer, comprising a substance for detecting organic compounds; the organic compounds comprise one or more of the following: m / z(VOC1)=28.0, m / z(VOC2)=32.2, m / z(VOC3)=40.0, m / z(VOC4)=101.0, m / z(VOC5)=112.0, m / z(VOC6)=123.0, m / z(VOC7)=126.0, m / z(VOC8)=132.0, m / z(VOC9)=92.0, m / z(VOC10)=103.0.

[0113] Figure 2 yes A schematic diagram of a device for analyzing breast cancer respiratory gas data based on noninvasive breath biopsy technology provided in an embodiment of the present invention includes: a memory and a processor; the memory is used to store program instructions; the processor is used to call the program instructions, and when the program instructions are executed, it is used to perform the above-mentioned method for analyzing breast cancer respiratory gas data based on noninvasive breath biopsy technology.

[0114] Figure 3 yes The schematic flow chart of the breast cancer respiratory gas data analysis system based on non-invasive breath biopsy technology provided in an embodiment of the present invention includes:

[0115] An acquisition unit 301 is used to acquire the respiratory gas of a sample to be tested;

[0116] The first processing unit 302 is used to process the respiratory gas of the sample to be tested to obtain target VOC ion data;

[0117] The second processing unit 303 is configured to input the target VOC ion data into the constructed breast cancer prediction model 1 for processing to obtain a classification result.

[0118] The method of processing the respiratory gas of the sample to be tested to obtain target VOC ion data includes:

[0119] The respiratory gas of the sample to be tested is detected and analyzed to obtain processed mass spectrometry data; discrete signals of the processed mass spectrometry data are converted into standard VOC features; VOC features with a<m / z<b are screened from the standard VOC features; and the screened VOC features are sorted according to feature importance or coefficient to obtain the target VOC ion data.

[0120] An embodiment of the present invention provides a breast cancer respiratory gas data analysis system based on non-invasive breath biopsy technology, comprising:

[0121] an acquisition unit, configured to acquire respiratory gas and clinical data of a sample to be tested, wherein the clinical data includes one or more of the following: age, BMI, family cancer history, personal cancer history, and menopausal status;

[0122] a first processing unit, configured to process the respiratory gas of the sample to be tested to obtain target VOC ion data;

[0123] The second processing unit is used to input the target VOC ion data and clinical data into the constructed breast cancer prediction model 2 to obtain a classification result.

[0124] Figure 4 yesThe present invention provides a subject enrollment and study design diagram, wherein this study consecutively recruited women undergoing breast cancer screening at six hospitals in China. Participants were divided into a discovery cohort, used to identify candidate VOCs and construct a diagnostic model, and an external validation cohort, used to independently test the model's diagnostic value. The discovery cohort recruited 498 women undergoing opportunistic breast cancer screening at the Cancer Hospital of the Chinese Academy of Medical Sciences and 2,755 women from the Beijing population undergoing breast cancer screening in northern China. For model construction, the discovery dataset was randomly divided into training, internal validation, and testing datasets in a 5:2:3 ratio. The external validation cohort included women undergoing opportunistic breast cancer screening in the Yantai cohort in eastern China and the Wenzhou cohort in southeastern China, as well as women from the China Urban Population Breast Cancer Screening Project in Guiyang (the Guiyang cohort in southwestern China). For each participant, clinical data (risk factor information) and a breath sample for breast cancer were collected before standard mammography and ultrasound examinations. Women with breast lesions suspected of malignancy underwent surgical or core needle biopsy after screening. Women without suspicious breast lesions or with negative biopsy results were excluded from breast cancer follow-up after 6 months of follow-up. The final diagnosis was based on pathological findings and follow-up. Seventy-eight patients who were lost to follow-up were excluded. Abbreviations: BC, breast cancer; CAMS, Chinese Academy of Medical Sciences.

[0125] Figure 5 yes The data analysis workflow and distribution diagram of target VOC ion data in model construction provided by an embodiment of the present invention, wherein, (A) the workflow of data analysis and model construction. Breath samples were collected using a self-designed collector and air bag through a standardized breath collection procedure, and then analyzed using high-pressure photon ionization-time of flight mass spectrometry (HPPI-TOFMS). Data of 1500 VOC ions were detected from the m / z range [20,320] with an interval of 0.2. Based on the random forest algorithm, the best 10 VOC ions were confirmed according to the feature importance or coefficient in model training. Two breast cancer detection models (BreathBC and BreathBC-Plus) were constructed using breath VOC markers with or without risk factors (clinical data). Both models were validated using three external validation cohorts. (B) Among all participants in this study, the 10 best VOC ions showed significant differences between breast cancer patients and non-cancer women, including 8 elevated VOCs and 2 decreased VOCs. (C) Receiver operating characteristic (ROC) curves and associated area under the curve (AUC) for the diagnostic performance of the 10 best VOC ions. Abbreviations: HPPI-TOFMS, high-pressure photoionization-time-of-flight mass spectrometry; VOC, volatile organic compound; BC, breast cancer; HC, healthy control; AUC, area under the curve.

[0126] Figure 6 yes Schematic diagram of the performance of the breast cancer detection models of the BreathBC model (breast cancer prediction model 1) and BreathBC-Plus (breast cancer prediction model 2) provided by the embodiments of the present invention in the internal validation cohort, the test cohort, and three external validation cohorts. For the BreathBC model using 10 breath VOC markers, the diagnostic AUC was 0.96 (95% CI, 0.94-0.97) in the internal validation cohort, 0.95 (95% CI, 0.93-0.90) in the test cohort (A), and 0.87 (B) in the external validation cohort. For the BreathBC-Plus model that uses both breath VOC markers and risk factors (clinical data), the combined model outperformed the BreathBC model in the internal validation cohort and the test cohort (AUC = 0.97-0.98) (C) and in the external validation cohort (AUC = 0.94) (D). Abbreviations. AUC, area under the curve.

[0127] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned breast cancer respiratory gas data analysis method based on non-invasive exhaled breath biopsy technology.

[0128] The validation results of this validation example show that assigning inherent weights to indications can moderately improve the performance of the method relative to the default setting.

[0129] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0130] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0131] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0132] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0133] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, which may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0134] Those skilled in the art will understand that all or part of the steps in the above-mentioned embodiment method can be implemented by instructing the relevant hardware through a program, and the program can be stored in a computer-readable storage medium. The above-mentioned storage medium can be a read-only memory, a disk or an optical disk, etc.

[0135] The above is a detailed introduction to a computer device provided by the present invention. For those skilled in the art, according to the concept of the embodiments of the present invention, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. Method for constructing a breast cancer prediction model based on non-invasive breath biopsy technology, including: Obtaining respiratory gas and classification labels of training set samples, wherein the classification labels include breast cancer patients and non-breast cancer patients; Detecting and analyzing the respiratory gas of the training set samples to obtain processed mass spectrometry data; Converting the discrete signals of the processed mass spectrometry data into a standard volatile organic compound (VOC) signature; the conversion method includes: calculating the area of ​​the most important peak in the range of [x-0.1, x+0.1] and treating it as the VOC signature of the volatile organic compound with m / z x, and the VOC signature of the volatile organic compound with m / z x as the standard VOC signature of the volatile organic compound; m / z is the ratio of the number of protons to the number of charges; Sorting the standard volatile organic compound (VOC) features to obtain target volatile organic compound (VOC) ion data; Inputting the target volatile organic compound (VOC) ion data into a machine learning model to obtain a predicted classification result, comparing it with the classification label, and optimizing the model based on the comparison result to obtain a constructed breast cancer prediction model 1; The method for constructing the breast cancer prediction model also includes: obtaining clinical data of training set samples, performing feature extraction on the clinical data to obtain clinical data features; performing feature fusion processing on the target volatile organic compound (VOC) ion data and clinical data features to obtain a feature set; inputting the feature set into a machine learning model to obtain a predicted classification result, comparing it with the classification label, and optimizing the model based on the comparison result to obtain a constructed breast cancer prediction model 2.

2. The method for constructing a breast cancer prediction model based on non-invasive breath biopsy technology according to claim 1, characterized in that: Screening volatile organic compound (VOC) features with a < m / z < b from the standard volatile organic compound (VOC) features, sorting the volatile organic compound (VOC) features, and obtaining target volatile organic compound (VOC) ion data, wherein the range of a is 0-50; the range of b is 100-500; The screened volatile organic compound (VOC) features are ranked according to feature importance or coefficient.

3. The method for constructing a breast cancer prediction model based on non-invasive breath biopsy technology according to claim 1, characterized in that: The clinical data include one or more of the following: age, BMI, family cancer history, personal cancer history and menopausal status.

4. A method for analyzing breast cancer respiratory gas data based on non-invasive breath biopsy technology, characterized in that: The method comprises: Obtaining respiratory gas of a sample to be tested; Processing the breath gas of the sample to be tested to obtain target volatile organic compound (VOC) ion data; The target volatile organic compound (VOC) ion data is input into a breast cancer prediction model 1 constructed by the method according to any one of claims 1 to 3 to obtain a classification result.

5. The breast cancer respiratory gas data analysis method based on non-invasive breath biopsy technology according to claim 4, characterized in that: The method further comprises: Obtaining respiratory gas and clinical data of the sample to be tested, wherein the clinical data includes one or more of the following: age, BMI, family cancer history, personal cancer history, and menopausal status; Processing the breath gas of the sample to be tested to obtain target volatile organic compound (VOC) ion data; The target volatile organic compound (VOC) ion data and clinical data are input into the second breast cancer prediction model constructed by the method according to any one of claims 1 to 3 to obtain a classification result.

6. Breast cancer respiratory gas data analysis method based on non-invasive breath biopsy technology, including: Obtaining respiratory gas of a sample to be tested; Processing the breath gas of the sample to be tested to obtain target volatile organic compound (VOC) ion data; The target volatile organic compound (VOC) ion data includes one or more of the following: m / z=28.0, m / z=32.2, m / z=40.0, m / z=101.0, m / z=112.0, m / z=123.0, m / z=126.0, m / z=132.0, m / z=92.0, m / z=103.0; The target volatile organic compound (VOC) ion data is input into a breast cancer prediction model 1 constructed by the method according to any one of claims 1 to 3 to obtain a classification result.

7. The method for analyzing breast cancer respiratory gas data based on non-invasive breath biopsy technology according to claim 6, characterized in that: The method further comprises: Obtaining respiratory gas and clinical data of the sample to be tested, wherein the clinical data includes one or more of the following: age, BMI, family cancer history, personal cancer history, and menopausal status; Processing the breath gas of the sample to be tested to obtain target volatile organic compound (VOC) ion data; The target volatile organic compound (VOC) ion data and the clinical data are input into a second breast cancer prediction model constructed by the method according to any one of claims 1 to 3 to obtain a classification result.

8. The method for analyzing breast cancer respiratory gas data based on non-invasive breath biopsy technology according to claim 6 or 7, characterized in that: The process of processing the respiratory gas of the sample to be tested includes: Detecting and analyzing the respiratory gas of the sample to be tested to obtain processed mass spectrometry data; Converting the processed mass spectrometry data into a standard volatile organic compound (VOC) signature. Volatile organic compound (VOC) features with a < m / z < b are screened from the standard volatile organic compound (VOC) features; and the screened volatile organic compound (VOC) features are sorted according to feature importance or coefficient to obtain the target volatile organic compound (VOC) ion data.

9. Use of a substance for detecting an organic compound obtained according to the method of any one of claims 1 to 3 in preparing a product for diagnosing or assisting in the diagnosis of breast cancer; the substance of the organic compound is the target volatile organic compound (VOC) ion data, including one or more of the following: m / z=28.0, m / z=32.2, m / z=40.0, m / z=101.0, m / z=112.0, m / z=123.0, m / z=126.0, m / z=132.0, m / z=92.0, m / z=103.

0.

10. A product for diagnosing or assisting in the diagnosis of breast cancer, comprising a substance for detecting an organic compound obtained according to the method of any one of claims 1 to 3; the substance of the organic compound is the target volatile organic compound (VOC) ion data, including one or more of the following: m / z=28.0, m / z=32.2, m / z=40.0, m / z=101.0, m / z=112.0, m / z=123.0, m / z=126.0, m / z=132.0, m / z=92.0, m / z=103.

0.

11. A breast cancer respiratory gas data analysis system based on non-invasive breath biopsy technology, including: an acquisition unit, used for acquiring the respiratory gas of the sample to be tested; a first processing unit, configured to process the respiratory gas of the sample to be tested to obtain target volatile organic compound (VOC) ion data; The second processing unit is used to input the target volatile organic compound (VOC) ion data into the breast cancer prediction model 1 constructed by the method according to any one of claims 1 to 3 for processing to obtain a classification result.

12. A breast cancer respiratory gas data analysis device based on non-invasive exhaled breath biopsy technology, the device comprising: memory and processor; The memory is used to store program instructions; The processor is used to call program instructions, and when the program instructions are executed, is used to execute the breast cancer respiratory gas data analysis method based on non-invasive exhaled breath biopsy technology according to any one of claims 4 to 8.

13. A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method for analyzing respiratory gas data of breast cancer based on non-invasive breath biopsy technology according to any one of claims 4 to 8 is implemented.

Images