Early warning screening system for critical and severe patients based on detection of exhaled VOCs by CI-TOFMS

Abstract

The application discloses a kind of early warning screening systems of critical patient based on CI-TOFMS detection exhaled air VOCs, belong to early screening field.It includes sample processing system, CI-TOFMS, data processing module;The sample processing system is used to collect and pre-process exhaled air sample, the CI-TOFMS is used to detect exhaled air sample and obtain eVOCs spectrum, the data processing module is used to collect eVOCs spectrum data, and output early warning result;The data processing module includes data input module, eVOCs-based critical illness prediction model and result output module.The system of the application only needs to collect respiratory air to carry out the health status evaluation of test person, and it is very convenient.The eVOCs-based critical illness prediction model constructed further improves the accuracy and sensitivity of eVOCs detection technology in critical patient diagnosis, and provides more reliable diagnostic basis for clinicians.

Description

Technical Field

[0001] This invention pertains to early disease prediction technology and relates to an early warning and screening system for critically ill patients based on CI-TOFMS detection of exhaled VOCs. Background Technology

[0002] Emergency medicine requires rapid assessment and diagnosis of acute and critical illnesses, followed by emergency treatment, and provides emergency medical rescue services for public health emergencies. Acute and critical illnesses (such as cardiac arrest, acute coronary syndrome, acute pulmonary embolism, aortic dissection, acute respiratory distress syndrome, sepsis, and acute pancreatitis) are characterized by their sudden onset, unpredictability, complexity, rapid progression, high mortality, and poor prognosis, ultimately leading to various types of shock and major organ failure. From a pathophysiological perspective, acute and critical illnesses generally involve internal environmental disturbances, metabolic disorders, and energy supply and utilization disorders, often resulting in impaired cellular function. If not controlled promptly, the condition worsens, eventually leading to multiple organ dysfunction and endangering the patient's life.

[0003] With the advancement of technology, blood and urine tests have greatly facilitated disease diagnosis. However, due to the long testing time, diagnosis may be delayed for some patients. Human exhaled VOCs are produced by cells through metabolic processes, and previous studies have identified physiological links between VOCs and oxidative stress, lipid and amino acid metabolism, and inflammation. Research shows that human respiration contains a complex mixture of hundreds of VOCs, and respiratory analysis has been used to diagnose, treat, or monitor many diseases, such as chronic obstructive pulmonary disease, cancer, asthma, and diabetes. Therefore, VOCs can serve as biomarkers for overall health.

[0004] Current research on breath analysis either lacks specific identification of key VOC species or fails to adequately analyze multidimensional VOC data using advanced algorithms such as machine learning to provide better classification performance. Detecting specific levels of VOCs in exhaled breath as biomarkers for critical illnesses has the potential to provide patients with a low-burden, rapid, and non-invasive tool. Summary of the Invention

[0005] The problem this invention aims to solve is to provide a critical patient early warning and screening system based on chemical ionization time-of-flight mass spectrometry for detecting volatile organic compounds in exhaled breath.

[0006] The technical solution of the present invention is as follows:

[0007] A critical care patient early warning and screening system based on CI-TOFMS detection of exhaled VOCs includes a sample processing system, CI-TOFMS, and a data processing module. The sample processing system is used to collect and preprocess exhaled air samples, the CI-TOFMS is used to detect exhaled air samples to obtain eVOCs spectra, and the data processing module is used to collect eVOCs spectra data and output early warning results.

[0008] The data processing module includes a data input module, an eVOCs-based critical illness prediction model, and a result output module.

[0009] The method for constructing the eVOCs-based critical illness prediction model is as follows:

[0010] S1: Exhaled air samples from patients with acute and critical illnesses and healthy individuals of different degrees were detected by chemical ionization time-of-flight mass spectrometry to obtain eVOCs spectra. The eVOCs markers with significant differences were obtained by comparing different eVOCs spectra.

[0011] S2: Collect spectral data of eVOCs markers with significant differences, process the data, extract features, and obtain the eVOCs feature set;

[0012] S3: The eVOCs feature set, patient basic characteristics and clinical manifestation data are integrated and processed, and machine learning methods are used for training to establish an eVOCs-based prediction model for acute and critical illnesses.

[0013] Furthermore, the machine learning method includes at least one of the following: logistic regression model, XGBoost model, random forest model, support vector model, K-nearest neighbor algorithm model, and Naive Bayes model.

[0014] Furthermore, in S2, the mass spectrometry data is reduced in dimensionality, denoised, and standardized by principal component analysis of the significantly different eVOCs marker spectra, resulting in an eVOCs feature set.

[0015] Furthermore, in S3, the patient's basic characteristics and clinical manifestations data include age, gender, BMI, blood oxygen saturation, heart rate, and blood pressure.

[0016] Furthermore, in S3, the target variable Rj is calculated by the eVOCs feature set, patient basic characteristics and clinical manifestation data through the following formula. The target variable Rj dataset is trained using machine learning methods to establish an eVOCs-based critical illness prediction model.

[0017]

[0018] in,

[0019] Cj represents the patient's basic characteristics;

[0020] Sj represents the patient's clinical data;

[0021] Vj represents the feature set data of eVOCs;

[0022] T response The time from the onset of symptoms to the patient taking a response;

[0023] T symptom The time when disease symptoms appear, manifested as signs perceived by the patient;

[0024] P VOCs The rate of change of eVOCs concentration;

[0025] Lj represents the patient's personalized information;

[0026] α1, α2, and α3 are scaling factors, determined through hyperparameter adjustment during model training;

[0027] δ is the disease type factor; δ = 1 before disease classification.

[0028] Furthermore, Lj represents the patient's personalized information, which is assigned by clinical medical staff based on abnormal fluctuations in eVOCs.

[0029] Furthermore, the chemical ionization time-of-flight mass spectrometry is photo-induced chemical ionization time-of-flight mass spectrometry.

[0030] Furthermore, the light source for the photochemical ionization time-of-flight mass spectrometer is a radio frequency ultraviolet lamp with an energy of 10.6 eV, and the injection line temperature is set to 150 degrees Celsius.

[0031] Furthermore, the sample processing system includes an exhaled air collection channel, a CO2 sensor disposed on the inner wall of the exhaled air collection channel, a sampling switch electrically connected to the CO2 sensor, and a sampling container connected to the exhaled air collection channel; the sampling switch is disposed at the connection between the exhaled air collection channel and the sampling container, and is used to control the opening and closing of the sampling container.

[0032] Furthermore, it also includes S4: classifying the output of the eVOCs-based critical illness prediction model into critical illness types, standardizing the results based on the disease mortality rate, and obtaining the disease type factor δ.

[0033] This invention also provides the application of a CI-TOFMS-based early warning and screening system for critically ill patients in predicting the progression of critical illnesses and assisting in determining mortality and disease type.

[0034] The beneficial effects of this invention are:

[0035] 1) Easy to use. The system of this invention only requires the collection of a person's exhaled breath sample to achieve health status assessment and determine the patient's acute and critical condition. In addition, the model algorithm can be adjusted according to the disease type. For example, in the case of cardiovascular diseases, the exhaled breath composition includes elevated acetone levels in the exhaled breath of patients with heart failure.

[0036] 2) Rapid detection. Collected samples can be detected in just 1 second, and test results can be obtained within 1 minute, enabling rapid clinical testing.

[0037] 3) High representativeness of the test results. The results of this invention demonstrate good diagnostic performance in both the training and test sets, with an area under the ROC curve (AUC) exceeding 0.95. This invention constructs a predictive model for acute and critical illnesses based on eVOCs, further improving the accuracy and sensitivity of eVOCs detection technology in the diagnosis of critically ill patients, providing clinicians with more reliable diagnostic evidence. Furthermore, the application of volatile organic compound detection technology is not limited to the diagnosis of acute and critically ill patients but can be extended to other fields. Combined with the detection of components in exhaled breath, it also holds promise for disease risk assessment and the development of personalized treatment plans. Attached Figure Description

[0038] Figure 1 The influence of X46, X58, X64, X55, and X48 volatile organic compounds on the prediction model is shown in Figure 1. A is a detailed view of the influence of each feature on a single prediction; B is the mean absolute SAP value of the volatile organic compounds; C through G show the relationship between the actual values ​​of each volatile organic compound (X46, X58, X64, X55, and X48) and their corresponding SAP values; and H is a heatmap of interactions for each volatile organic compound.

[0039] Figure 2 The results of screening for significantly different eVOCs are shown in Figure 1; A represents volatile organic compounds with an AUC value greater than 0.9; B represents the Gini coefficient comparison of ten key eVOCs; C represents the five volatile organic compounds (X46, X58, X64, X55, and X48) identified by the VSOLassoBag algorithm; and D represents the distribution of expression levels of each VOC in the two study groups. Detailed Implementation

[0040] The following detailed embodiments further illustrate the above-described content of the present invention. However, this should not be construed as limiting the scope of the present invention to the following examples. All technologies implemented based on the above-described content of the present invention fall within the scope of the present invention. Obviously, based on the above-described content of the present invention, and according to ordinary technical knowledge and common practices in the art, various other modifications, substitutions, and alterations can be made without departing from the basic technical concept of the present invention.

[0041] Breath detection of exhaled volatile organic compounds (eVOCs) is a non-invasive technique with potential value in disease diagnosis. In some embodiments of this invention, a prospective study was conducted, consecutively enrolling 121 critically ill patients who visited the emergency department of our hospital from January to March 2024, with 728 healthy volunteers as controls. A self-developed chemical ionization time-of-flight mass spectrometry (CI-TOFMS) platform was used to detect volatile organic compounds (VOCs) in exhaled breath. Receiver operating characteristic (ROC) curves, ensemble learning algorithms, and LASSO regression were used to screen key VOC biomarkers. Based on these key VOC biomarkers, a diagnostic model was constructed using four machine learning algorithms and formulas incorporating different disease variables.

[0042] In some embodiments of this invention, patients were divided into an acute myocardial infarction subgroup and a pneumonia subgroup based on their initial diagnosis. Propensity score matching was used to select controls, and the model's performance in specific disease populations was evaluated. Based on VOC detection results from healthy controls, a reference range for exhaled VOCs in healthy individuals was established, and standard curves for nine aldehyde compounds were plotted. Using various statistical methods, five VOCs (X46, X58, X64, X55, and X48) were identified as having the best diagnostic performance in distinguishing critically ill patients from healthy controls. Four diagnostic models constructed based on these five biomarkers showed good diagnostic performance on both the training and test sets, with area under the ROC curve (AUC) exceeding 0.95. In conclusion, this study utilized CI-TOFMS technology to detect eVOC spectra, preliminarily exploring its application potential in the rapid screening and evaluation of critically ill patients. This non-invasive diagnostic strategy is simple to operate and can improve the diagnostic efficiency for critically ill patients.

[0043] Example 1

[0044] The critical patient early warning screening system of the present invention:

[0045] The present invention relates to an early warning and screening system for critically ill patients based on CI-TOFMS detection of exhaled VOCs, comprising a sample processing system, a CI-TOFMS, and a data processing module; the sample processing system is used to collect and preprocess exhaled air samples, the CI-TOFMS is used to detect exhaled air samples to obtain eVOCs spectra, and the data processing module is used to collect eVOCs spectra data and output early warning results;

[0046] The patient's exhaled breath sample is introduced into a photochemical ionization time-of-flight mass spectrometer using a sample processing system to detect volatile organic compounds (eVOCs) in the exhaled breath and obtain eVOC composition data for different samples.

[0047] The data processing module includes a data input module and a result output module that transmit VOCs composition data from different samples to an eVOCs-based critical illness prediction model.

[0048] The data input module obtains eVOCs composition data of different samples from the sample detection module. It uses principal component analysis (PCA), t-distributed random neighborhood embedding (t-SNE), uniform manifold approximation and projection (UMAP), ensemble learning algorithm XGBoost model and VSOLassoBag algorithm to distinguish the most discriminative eVOCs between critically ill patients and healthy controls, namely the eVOCs feature set.

[0049] The target variable Rj is calculated by the eVOCs feature set, patient basic characteristics and clinical manifestation data through the following formula. The target variable Rj dataset is trained using the XGBoost machine learning method to establish an eVOCs-based critical illness prediction model (XGBoost model).

[0050]

[0051] in,

[0052] Cj represents the patient's basic characteristics;

[0053] Sj represents the patient's clinical data;

[0054] Vj represents the feature set data of eVOCs;

[0055] T response The time from the onset of symptoms to the patient taking a response;

[0056] T symptom The time when disease symptoms appear, manifested as signs perceived by the patient;

[0057] P VOCs The rate of change of eVOCs concentration;

[0058] Lj represents the patient's personalized information;

[0059] α1, α2, and α3 are scaling factors, determined through hyperparameter adjustment during model training;

[0060] δ is the disease type factor; δ = 1 before disease classification.

[0061] The output module is responsible for using the trained XGBoost model to predict and analyze eVOCs samples at healthy locations, and to obtain the detection results of the samples.

[0062] The results of this invention demonstrate good diagnostic performance on both the training and test sets, with an area under the ROC curve (AUC) exceeding 0.95.

[0063] Example 2

[0064] Application of eVOCs samples for early warning screening analysis of critically ill patients.

[0065] 1. Method

[0066] The critically ill patient group consisted of 121 patients from Qilu Hospital of Shandong University, all of whom were in critical condition. The healthy control group consisted of 728 patients, all recruited volunteers. The demographic and pathological characteristics of the healthy control group and the critically ill patient group are shown in Table 1.

[0067]

[0068]

[0069] This study was approved by the Ethics Committee of Qilu Hospital, Shandong University, and has been registered on clinicalTrials.gov (NCT06277895). It followed the STARD (Standard for Reporting Diagnostic Accuracy) reporting guidelines. All participants signed written informed consent forms.

[0070] Healthy individuals in the control group were recruited from the Health Examination Center of Qilu Hospital, Shandong University. Inclusion criteria included: 1) being at least 18 years old; 2) no history of cancer, no signs of liver or kidney dysfunction, and no signs of infection; 3) not consuming spicy or irritating foods (such as coffee, wine, garlic, etc.) within the past 24 hours; and 4) comprehensive clinical data. Exclusion criteria included: 1) a history of cancer; 2) signs of active infection; or 3) liver or kidney dysfunction; and 4) incomplete clinical and / or accessory examination records.

[0071] Patients in the experimental group were recruited from the Emergency Department of Qilu Hospital of Shandong University. Inclusion criteria included: 1) being at least 18 years old; 2) having no cardiovascular or pulmonary disease, no liver or kidney dysfunction, and no acute infection or pneumonia but no cardiovascular or pulmonary disease. Data collection for this study was conducted in accordance with relevant local guidelines and indicators, and human data collection was approved by the Ethics Committee of Qilu Hospital of Shandong University (Approval No.: KYLL-202401-047).

[0072] We established standard sampling requirements and protocols, with trained investigators collecting exhaled breath samples. For healthy and conscious individuals: 1) All participants needed to prepare for sampling in advance: refrain from smoking, drinking alcohol, or consuming spicy or irritating foods (such as coffee, garlic, etc.) for 24 hours prior to sampling; 2) rinse their mouths with purified water immediately before sampling; 3) all samples were collected at the same location to minimize the influence of environmental factors. Subjects took a deep breath and completely inhaled air into the sampling bag through a disposable nozzle. The 1L breath sampling bag was made of polytetrafluoroethylene (PTFE).

[0073] For intubated individuals: Connect the inlet of the gas collection bag to the endotracheal tube port. Observe the ventilator waveform interface; open the gas collection bag valve during exhalation and immediately close the inlet valve at the end of exhalation, then reconnect the ventilator tubing.

[0074] The chemical ionization time-of-flight mass spectrometry (CI-TOFMS) in this embodiment mainly consists of a chemical ionization source, an ion transport system, a vacuum system, and a TOF mass spectrometer. The chemical ionization source mainly comprises a VUV krypton lamp, a repulsion electrode (V1), a sample introduction electrode (V2), a segmented quadrupole (SQ), and a differential vacuum orifice (Skimmer 1). Exhaled breath samples are directly introduced into the ionization region through a PEEK capillary with an inner diameter of 0.5 mm, an outer diameter of 1 / 16 inch, and a length of 1.2 m. To eliminate eVOCs condensation and minimize potential surface adsorption, the PEEK capillary is heated to 150 degrees Celsius. The ion extraction period is 40 μs (25 kHz), and the temporal resolution of one spectral data point is 1 s.

[0075] The 2.5 to 97.5 percentile ranges for each compound were used to establish medical reference ranges for 132 volatile organic compounds (eVOCs) detected in healthy controls. Principal component analysis (PCA), t-distributed random neighborhood embedding (t-SNE), and uniform manifold approximation and projection (UMAP) were performed on the eVOC sample data from critically ill patients and healthy individuals to analyze the overall classification trend of VOC expression profiles between groups.

[0076] The diagnostic performance of each VOC was evaluated using receiver operating characteristic (ROC) curve analysis. The area under the curve (AUC) for each eVOC was calculated, and VOCs with an AUC greater than 0.9 were selected for further analysis to obtain the eVOC feature set. The XGBoost ensemble learning algorithm was used to rank the importance of the detected VOCs. The intersection of the VOCs selected by the two methods was used for feature selection using the VSOLassoBag algorithm. The final VOCs selected by VSOLassoBag were considered the key VOCs distinguishing critically ill patients from healthy controls, and the target variable Rj was extracted for the key VOCs.

[0077]

[0078] Cj represents the patient's age, gender, BMI, blood oxygen saturation, heart rate, and blood pressure;

[0079] Sj represents the patient's age, gender, BMI, blood oxygen saturation, heart rate, and blood pressure;

[0080] Vj represents the feature set data of eVOCs;

[0081] T response The time from the onset of symptoms to the patient taking a response;

[0082] T symptom The time when disease symptoms appear, manifested as signs perceived by the patient;

[0083] P VOCs The rate of change of eVOCs concentration;

[0084] Lj represents the patient's personalized information;

[0085] The target variable Rj was further calculated by using the eVOCs feature set, patient basic characteristics and clinical manifestation data through the following formula. The target variable Rj dataset was trained using machine learning methods to establish an eVOCs-based prediction model for acute and critical illnesses.

[0086] This embodiment identifies 112 volatile organic compounds (VOCs) that showed differences between healthy individuals and critically ill patients, with an adjusted p-value of fdr < 0.05. The AUC was calculated to assess their individual diagnostic performance. Thirteen of these VOCs had an AUC greater than 0.9. Figure 2 A). We then used an ensemble machine learning strategy to evaluate the importance of these 112 volatile organic compounds in constructing the random forest model to differentiate between the case and control groups. Based on the average decrease in the Gini coefficient, we identified the ten most important volatile organic compounds (VOCs). Figure 2 B). The intersection of these methods yielded 10 key volatile organic compounds: X46, X58, X64, X55, X48, X158, X50, X44, X63, and X56, representing small volatile organic compounds in exhaled breath with molecular weights of 46, 58, 64, 55, 48, 158, 50, 44, 63, and 56, which can be used to characterize differences between critically ill patients and healthy subjects. Subsequently, we applied the VSOLassoBag algorithm with 100 sampling iterations for feature selection to enhance robustness. This algorithm identified five volatile organic compounds (X46, X58, X64, X55, and X48) as the optimal markers for the final diagnostic model. Figure 2C). Standard curves were plotted to identify these organic compounds as acetaldehyde (46), methyl mercaptan (48), 1,3-butadiene (55), acrolein (55), n-pentene (55), cyclopentane (55), cis-trans-2,pentene (55), n-butyraldehyde (55), propionaldehyde (58), butane (58), and 1,2-dichloroethane. Violin plots illustrate the distribution of expression levels of each eVOC in the two study groups. Figure 2 D).

[0087] During model building and evaluation, the data was divided into training and validation sets in a 7:3 ratio. Six models were built, including logistic regression, XGBoost, random forest, support vector machine, K-nearest neighbor algorithm model, and Naive Bayes model, to validate the correctness and classification performance of candidate biomarkers.

[0088] We enhance the interpretability of our Rj-optimized XGBoost model by using the SHAP method to elucidate the output of the prediction by quantifying the contribution of each variable. Figure 1 A provides a detailed view of the impact of each feature on a single prediction. Each point represents a prediction, the color indicates the feature value, and the position on the x-axis indicates the SHAP value. Figure 1 B shows the mean absolute SHAP values ​​of these volatile organic compounds, arranged in descending order of their contribution. The results indicate that X46 has the most significant impact on the model output, while X55 has the least. To further clarify the influence of individual characteristics on model predictions, Figure 1 C through 1G present SHAP dependency plots illustrating the relationship between the actual value of each biomarker and its corresponding SHAP value, thus highlighting the contribution of each biomarker to prediction. For example, subjects with X46 > 2247 or X48 > 17307.6 have SHAP values ​​greater than zero, which pushes the decision towards the "Case" class. Furthermore, interaction heatmaps ( Figure 1 H) shows that, in particular, the main effects of the individual features X46 and X48 (see diagonal) are significantly higher than their interaction effects, indicating that these features mainly influence the model independently.

[0089] The results show that eVOCs characteristics can be used to detect critically ill patients and healthy individuals with high accuracy, sensitivity and specificity.

[0090] In summary, the early warning screening system of the present invention can realize early warning screening for critically ill patients.

Claims

1. A critical care patient early warning and screening system based on CI-TOFMS detection of exhaled VOCs, characterized in that, It includes a sample processing system, a CI-TOFMS, and a data processing module; the sample processing system is used to collect and preprocess exhaled air samples, the CI-TOFMS is used to detect exhaled air samples to obtain eVOCs spectra, and the data processing module is used to collect eVOCs spectra data and output early warning results; The data processing module includes a data input module, an eVOCs-based critical illness prediction model, and a result output module. The method for constructing the eVOCs-based critical illness prediction model is as follows: S1: Exhaled air samples from patients with acute and critical illnesses and healthy individuals of different degrees were detected by chemical ionization time-of-flight mass spectrometry to obtain eVOCs spectra. The eVOCs markers with significant differences were obtained by comparing different eVOCs spectra. S2: Collect spectral data of eVOCs markers with significant differences, process the data, extract features, and obtain the eVOCs feature set; S3: Integrate and process the eVOCs feature set, patient basic characteristics and clinical manifestation data, and use machine learning methods for training to establish an eVOCs-based prediction model for acute and critical illnesses. In S3, the patient's basic characteristics and clinical manifestation data include age, gender, BMI, blood oxygen saturation, heart rate, and blood pressure; the target variable Rj is calculated by the eVOCs feature set, the patient's basic characteristics, and clinical manifestation data using the following formula, and the target variable Rj dataset is trained using machine learning methods to establish an eVOCs-based critical illness prediction model. ， in, Cj represents the patient's basic characteristics; Sj represents the patient's clinical data; Vj represents the feature set data of eVOCs; T response The time from the onset of symptoms to the patient taking a response; T symptom The time when disease symptoms appear, manifested as signs perceived by the patient; P VOCs The rate of change of eVOCs concentration; Lj represents the patient's personalized information; α1, α2, and α3 are scaling factors, determined through hyperparameter adjustment during model training; δ is the disease type factor; δ = 1 before disease classification.

2. The critical care patient early warning and screening system based on CI-TOFMS detection of exhaled VOCs as described in claim 1, characterized in that, The machine learning methods include at least one of the following: logistic regression model, XGBoost model, random forest model, support vector model, K-nearest neighbor algorithm model, and Naive Bayes model.

3. The critical care patient early warning and screening system based on CI-TOFMS detection of exhaled VOCs as described in claim 1, characterized in that, In S2, the mass spectrometry data are reduced in dimensionality, denoised, and standardized by principal component analysis of the significantly different eVOCs marker spectra, resulting in the eVOCs feature set.

4. The critical care patient early warning and screening system based on CI-TOFMS detection of exhaled VOCs as described in claim 1, characterized in that, The chemical ionization time-of-flight mass spectrometry is photo-induced chemical ionization time-of-flight mass spectrometry.

5. The critical care patient early warning and screening system based on CI-TOFMS detection of exhaled VOCs as described in claim 4, characterized in that, The light source for the photochemical ionization time-of-flight mass spectrometry is an RF ultraviolet lamp with an energy of 10.6 eV, and the injection line temperature is set to 150 degrees Celsius.

6. The critical patient early warning screening system based on CI-TOFMS detection of exhaled VOCs as described in claim 1, wherein the sample processing system includes an exhaled air collection channel, a CO2 sensor disposed on the inner wall of the exhaled air collection channel, a sampling switch electrically connected to the CO2 sensor, and a sampling container connected to the exhaled air collection channel; the sampling switch is disposed at the connection between the exhaled air collection channel and the sampling container, and is used to control the opening and closing of the sampling container.

7. The critical care patient early warning and screening system based on CI-TOFMS detection of exhaled VOCs as described in claim 1, characterized in that, It also includes S4: classifying the output of the eVOCs-based acute and critical illness prediction model into acute and critical illness types, standardizing the results based on the disease mortality rate, and obtaining the disease type factor δ.

8. The critical care patient early warning and screening system based on CI-TOFMS detection of exhaled VOCs as described in claim 1, characterized in that, Lj represents the patient's personalized information, which is assigned by clinical medical staff based on abnormal fluctuations in eVOCs.

9. The application of the CI-TOFMS-based exhaled VOCs detection early warning screening system for critically ill patients as described in claim 1 in predicting the progression rate of critically ill diseases and assisting in the determination of mortality and disease type.

Images