Cerebral hematoma injury time prediction method based on Fourier infrared spectrum technology

Abstract

The invention discloses a cerebral hematoma injury time prediction method based on a Fourier infrared spectrum technology, and relates to the technical field of cerebral hematoma injury time prediction methods. In order to ensure that the speculation result is more objective and reliable, the method comprises the following steps: S1, collecting human body samples which need to be subjected to cerebral hematoma injury time prediction; s2, collecting spectral data on the basis of an FTIR-ATR (Fourier Transform Infrared-Attenuated S3, preprocessing the spectral data to reduce the influence of a plurality of physical factors such as sample concentration, uniformity and particle scattering; s4, analyzing the preprocessed spectral data based on a chemometrics method and software; and S5, discussing and summarizing analysis results. According to the invention, through thorough analysis of spectral changes of central and peripheral areas of intracerebral hematoma, comparable spectral trends of the areas are determined, and dynamic changes of proteins, phospholipids and nucleic acids in different post-injury stages are proved.

Description

Technical Field

[0001] The invention relates to the technical field of methods for predicting cerebral hematoma injury time, and in particular to a method for predicting cerebral hematoma injury time based on Fourier infrared spectroscopy technology. Background Art

[0002] In the prior art, injuries are common phenomena in forensic practice. The main task of forensic doctors is to determine the morphology of wounds, the time of injury, the cause of injury, and the inference of the object causing the injury. In the field of forensic medicine, estimating the time of injury is a complex task, especially in the case of injury or death that occurs without witnesses, when the external signs of trauma are not obvious or the medical history is unclear. This task is particularly difficult. The main reason is the lack of objective, reliable and portable inference methods.

[0003] Brain hemorrhage is a common condition in forensic investigations. Accidents may be caused by traffic accidents, falls or violent acts. The formation of hematomas is directly related to the time of injury, because after cerebral blood vessels are damaged, blood quickly infiltrates into brain tissue. The occurrence of hemorrhage is usually closely related to the time of injury. The survival time of intracerebral hematomas can be estimated by in vivo and in vitro methods. Radiological techniques such as computed tomography (CT) and magnetic resonance imaging (MRI) play a key role in the diagnosis of in vivo hematomas. However, although CT and MRI are widely used, they are difficult to be used as ideal tools due to their lack of portability and rapid detection capabilities, especially in crime scene investigations.

[0004] Attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR), as a variant of vibrational spectroscopy, provides a fast, economical, minimally invasive and non-destructive method that can characterize substances based on their molecular vibration characteristics. Attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) can effectively predict the time of cerebral hematoma injury. Therefore, this application proposes a method for predicting the time of cerebral hematoma injury based on Fourier transform infrared spectroscopy technology. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a method for predicting the injury time of cerebral hematoma based on Fourier transform infrared spectroscopy technology.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for predicting the time of cerebral hematoma injury based on Fourier transform infrared spectroscopy technology comprises the following steps:

[0008] S1: Collect human samples that need to predict the time of cerebral hematoma injury;

[0009] S2: Collect spectral data based on FTIR-ATR technology;

[0010] S3: The spectral data is preprocessed to reduce the influence of multiple physical factors such as sample concentration, uniformity and particle scattering;

[0011] S4: Analysis of pre-processed spectral data based on chemometric methods and software;

[0012] S5: Discussion and summary of the analysis results.

[0013] Preferably, the collection of human samples in step S1 includes the following standards:

[0014] 1: The sample is taken from a fresh corpse no more than 48 hours after death, or is frozen;

[0015] 2: The deceased must have a clearly documented history of brain trauma within 0 to 60 days before death;

[0016] 3: The patient must be diagnosed with cerebral hemorrhage by medical imaging (CT or MRI), or have obvious external features of traumatic brain injury;

[0017] 4: It must be possible to determine the exact time of injury and death;

[0018] 5: The diameter of the intracerebral hematoma must be at least 0.5 cm;

[0019] 6: Informed consent must be obtained from each deceased person’s legal representative.

[0020] Further: The aspects that do not meet the collection criteria are:

[0021] 1: History of mental illness, head trauma or intracranial lesions;

[0022] 2: The time of injury or death is not clearly recorded;

[0023] 3: suspected spontaneous cerebral hemorrhage, delayed hemorrhage or secondary brain trauma;

[0024] 4: the presence of subarachnoid hemorrhage, epidural hematoma, or subdural hematoma;

[0025] 5. Pregnant women;

[0026] 6: Infants and young children.

[0027] Further: the collection of spectral data in step S2 comprises the following steps:

[0028] S21: Collect several groups of formalin-fixed hematoma samples;

[0029] S22: During analysis, each group of samples was rinsed with tap water for 6-7 hours;

[0030] S23: After rinsing, each group of samples is homogenized and stored at -75-80°C;

[0031] S24: After pre-operation of the hematoma sample and the collection equipment, the Nicolet IS50 FTIR spectrometer, equipped with a diamond crystal ATR accessory, was used to analyze the hematoma sample using OMNIC TM v8.2 software combined with FTIR-ATR technology for data acquisition, recording wavenumber range from 4200 to 900 cm -1 ;

[0032] S25: Each sample was analyzed three times, 32 scans were performed each time, and the resolution was set to 4 cm -1 .

[0033] As a preferred solution of the present invention: when the acquisition equipment in step S24 is pre-treated, anhydrous ethanol is used to carefully clean the diamond crystal surface of the ATR probe, and the air is scanned as a background reference.

[0034] As a further solution of the present invention: During the sample preparation process, a vortex mixer was used to thoroughly mix the solution, and then 0.5 μL of the sample was deposited on the surface of the diamond crystal, and the sample was gently dried using a hair dryer set at a low temperature until a uniform film was obtained, and the time was 5-10 minutes.

[0035] As a further solution of the present invention: the preprocessing of the spectral data in step S3 comprises the following steps:

[0036] S31: Standard normal variation (SNV) transformation was used to mitigate the effects of surface scattering, particle size, and optical range variation. After the SNV transformation, the second-order derivative transformation was performed using the Savitzky-Golay algorithm, which used a window size of 9 points and a second-order polynomial.

[0037] S32: In terms of identifying and excluding outliers, Q residuals and leverage values ​​are used as standard indicators to promote the development of robust models and improve the reliability of predictions;

[0038] S33: Four spectra were finally identified as outliers and excluded from further data analysis for several groups of hematoma samples collected, of which 70% were allocated for the calibration dataset and the remaining 30% for the prediction dataset.

[0039] On the basis of the above scheme: the chemometric method in step S4 is based on statistical modeling, including supervised and unsupervised categories, among which the unsupervised program is mainly used for data exploration, using principal component analysis (PCA) to find the similarities and differences between data, and the supervised model includes two types: regression and classification, among which the regression method uses partial least squares (PLS).

[0040] On the basis of the above scheme: the software uses MATLAB R2016a and PLS Toolbox 8.2 version to combine, using multiple indicators, including the coefficient of determination (R 2 ), the root mean square error of cross validation (RMSECV), the root mean square error of prediction (RMSEP), and the ratio of RMSECV to RMSEP.

[0041] The beneficial effects of the present invention are:

[0042] 1. A method for predicting the time of brain hematoma injury based on Fourier transform infrared spectroscopy technology. Through a thorough analysis of the spectral changes in the center and peripheral areas of the intracerebral hematoma, the comparable spectral trends of these areas were determined, and the dynamic changes of proteins, phospholipids and nucleic acids in different post-injury stages were demonstrated.

[0043] 2. A method for predicting the time of cerebral hematoma injury based on Fourier transform infrared spectroscopy technology was developed and compared with full-spectrum PLS and GA-PLS models. The limited number of variables included in the model significantly improved the results. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 The present invention proposes a flow chart of a method for predicting cerebral hematoma injury time based on Fourier transform infrared spectroscopy technology. DETAILED DESCRIPTION

[0045] The technical solution of this patent is further described in detail below in conjunction with specific implementation methods.

[0046] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0047] Embodiment 1:

[0048] A method for predicting the time of cerebral hematoma injury based on Fourier transform infrared spectroscopy technology, such as Figure 1 As shown, the following steps are included:

[0049] S1: Collect human samples that need to predict the time of cerebral hematoma injury;

[0050] S2: Collect spectral data based on FTIR-ATR technology;

[0051] S3: The spectral data is preprocessed to reduce the influence of multiple physical factors such as sample concentration, uniformity and particle scattering;

[0052] S4: Analysis of pre-processed spectral data based on chemometric methods and software;

[0053] S5: Discussion and summary of the analysis results;

[0054] The collection of human samples in step S1 includes the following standards:

[0055] 1: The sample is taken from a fresh corpse no more than 48 hours after death, or is frozen;

[0056] 2: The deceased must have a clearly documented history of brain trauma within 0 to 60 days before death;

[0057] 3: The patient must be diagnosed with cerebral hemorrhage by medical imaging (CT or MRI), or have obvious external features of traumatic brain injury;

[0058] 4: It must be possible to determine the exact time of injury and death;

[0059] 5: The diameter of the intracerebral hematoma must be at least 0.5 cm;

[0060] 6: Informed consent must be obtained from each deceased person’s legal representative;

[0061] The aspects that do not meet the collection standards are:

[0062] 1: History of mental illness, head trauma or intracranial lesions;

[0063] 2: The time of injury or death is not clearly recorded;

[0064] 3: suspected spontaneous cerebral hemorrhage, delayed hemorrhage or secondary brain trauma;

[0065] 4: the presence of subarachnoid hemorrhage, epidural hematoma, or subdural hematoma;

[0066] 5. Pregnant women;

[0067] 6: Infants and young children;

[0068] The collection of spectral data in step S2 comprises the following steps:

[0069] S21: Collect several groups of formalin-fixed hematoma samples;

[0070] S22: During analysis, each group of samples was rinsed with tap water for 6-7 hours;

[0071] S23: After rinsing, each group of samples is homogenized and stored at -75-80°C;

[0072] S24: After pre-operation of the hematoma sample and the collection equipment, the Nicolet IS50 FTIR spectrometer, equipped with a diamond crystal ATR accessory, was used to analyze the hematoma sample using OMNIC TM v8.2 software combined with FTIR-ATR technology for data acquisition, recording wavenumber range from 4200 to 900 cm -1 ;

[0073] S25: Each sample was analyzed three times, 32 scans were performed each time, and the resolution was set to 4 cm -1 ;

[0074] When the acquisition device in step S24 is pre-processed, the diamond crystal surface of the ATR probe is carefully cleaned with anhydrous ethanol, and the air is scanned as a background reference to reduce environmental interference;

[0075] During sample preparation, the solution was thoroughly mixed using a vortex mixer, and then 0.5 μL of the sample was deposited on the surface of the diamond crystal. The sample was gently dried using a hair dryer set to low temperature until a uniform film was obtained, which took 5–10 min;

[0076] After the operation is completed, the sample is tested to see if it meets the established standards. If the test fails, the operation is repeated;

[0077] The spectral data is preprocessed in step S3, which includes the following steps:

[0078] S31: Standard normal variation (SNV) transformation is used to mitigate the effects of surface scattering, particle size, and optical range variation. After the SNV transformation, a second-order derivative transformation is performed using the Savitzky-Golay algorithm, which uses a window size of 9 points and a second-order polynomial;

[0079] S32: In terms of identifying and excluding outliers, Q residuals and leverage values ​​are used as standard indicators to promote the development of robust models and improve the reliability of predictions;

[0080] S33: Four spectra were finally identified as outliers and excluded from further data analysis for several groups of hematoma samples collected, of which 70% were allocated for the calibration dataset and the remaining 30% for the prediction dataset;

[0081] The chemometric method in step S4 is based on statistical modeling and includes supervised and unsupervised methods. The unsupervised program is mainly used for data exploration, using principal component analysis (PCA) to find similarities and differences between data. The supervised model includes two types: regression and classification. The regression method uses partial least squares (PLS).

[0082] It is an effective quantitative analysis technique. PLS regression has the advantages of noise reduction and multicollinearity processing. Compared with the traditional linear regression method, it can provide better prediction performance, especially when analyzing high-throughput data. It represents a modern method of multivariate linear regression.

[0083] At the same time, genetic algorithm and partial least squares (GA-PLS) are used to optimize variable selection and introduce evolutionary principles to significantly improve the accuracy and reliability of the model.

[0084] The combination of multiple methods makes the application of chemometrics in spectroscopy more efficient and accurate;

[0085] The software was combined with MATLAB R2016a and PLS Toolbox 8.2, and multiple indicators were used, including the coefficient of determination (R 2 ), the root mean square error of cross validation (RMSECV), the root mean square error of prediction (RMSEP), and the ratio of RMSECV to RMSEP. When the ratio of RMSEP to RMSECV is lower than 1.2, it indicates that the model has good stability and reproducibility. At the same time, R 2 The values ​​provide valuable insights into the explanatory power of the independent variables on the dependent variable, thus further validating the robustness of the model.

[0086] Embodiment 2:

[0087] A method for predicting the time of cerebral hematoma injury based on Fourier infrared spectroscopy technology. In order to detect the spectral difference between fresh samples and formalin-fixed samples, this embodiment is used as a comparative example to experimentally detect the spectral difference between the above two samples. The experimental detection method is:

[0088] 1: Compare the spectral differences of hematoma before and after fixation;

[0089] 2: Analysis results:

[0090] (1) Fresh hematoma and formalin-fixed hematoma have similar average spectral profiles within a specific wavenumber range, but there are significant differences in other areas;

[0091] (2) Compared with the fresh sample, the fixed sample has a -1 and 1080cm -1The intensity and spectral shape at 1645 cm -1 There is a slight offset;

[0092] (3) Conclusion: After continuous rinsing and soaking in tap water for 6 hours, the level of formalin contamination in fixed intracerebral hematoma samples was significantly reduced. This rinsing process effectively removed the residual formalin in the sample, thereby significantly reducing the artifacts that may be introduced in spectral analysis. This improvement helps to improve the accuracy of data interpretation;

[0093] Overall, the main absorption peaks of fixed hematoma samples remained consistent, indicating that fixation did not significantly change their chemical composition;

[0094] The formalin fixation process produces only minor changes to the main infrared spectral peaks of proteins and lipids, but these changes do not affect the suitability of the samples for subsequent studies; usually, after formalin fixation of tissues, the infrared spectrum will have a peak at 907cm -1 、1041cm -1 and 1492cm -1 Obvious absorption peaks were observed at

[0095] In this study, ATR-FTIR spectroscopy combined with chemometric techniques was used to thoroughly analyze the spectral changes in the central and peripheral regions of intracerebral hematomas. Comparable spectral trends in these regions were identified and dynamic changes in proteins, phospholipids, and nucleic acids were demonstrated at different post-injury stages.

[0096] The above is a preferred specific implementation manner of the present invention, and the protection scope of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by any technician familiar with the field within the technical scope disclosed by the present invention in combination with the prior art or public common sense, within the spirit and principle of the present invention, shall be covered by the protection scope of the present invention.

Claims

1. A method for predicting the time of cerebral hematoma injury based on Fourier transform infrared spectroscopy technology, characterized in that: The following steps are involved: S1: Collect human samples that need to predict the time of cerebral hematoma injury; S2: Collect spectral data based on FTIR-ATR technology; S3: The spectral data is preprocessed to reduce the influence of multiple physical factors such as sample concentration, uniformity and particle scattering; S4: Analysis of pre-processed spectral data based on chemometric methods and software; S5: Discussion and summary of the analysis results.

2. The method for predicting cerebral hematoma injury time based on Fourier transform infrared spectroscopy technology according to claim 1, characterized in that: The collection of human samples in step S1 includes the following standards: 1: The sample is taken from a fresh corpse no more than 48 hours after death, or is frozen; 2: The deceased must have a clearly documented history of brain trauma within 0 to 60 days before death; 3: The patient must be diagnosed with cerebral hemorrhage by medical imaging (CT or MRI), or have obvious external features of traumatic brain injury; 4: It must be possible to determine the exact time of injury and death; 5: The diameter of the intracerebral hematoma must be at least 0.5 cm; 6: Informed consent must be obtained from each deceased person’s legal representative.

3. The method for predicting cerebral hematoma injury time based on Fourier transform infrared spectroscopy technology according to claim 2, characterized in that: The aspects that do not meet the collection criteria are: 1: History of mental illness, head trauma or intracranial lesions; 2: The time of injury or death is not clearly recorded; 3: suspected spontaneous cerebral hemorrhage, delayed hemorrhage or secondary brain trauma; 4: the presence of subarachnoid hemorrhage, epidural hematoma, or subdural hematoma; 5. Pregnant women; 6: Infants and young children.

4. The method for predicting cerebral hematoma injury time based on Fourier transform infrared spectroscopy technology according to claim 3, characterized in that: The collection of spectral data in step S2 comprises the following steps: S21: Collect several groups of formalin-fixed hematoma samples; S22: During analysis, each group of samples was rinsed with tap water for 6-7 hours; S23: After rinsing, each group of samples is homogenized and stored at -75-80°C; S24: After pre-operation of the hematoma sample and the collection equipment, the Nicolet IS50 FTIR spectrometer, equipped with a diamond crystal ATR accessory, was used to analyze the hematoma sample using OMNIC TM v8.2 software combined with FTIR-ATR technology for data acquisition, recording wavenumber range from 4200 to 900 cm -1 ; S25: Each sample was analyzed three times, 32 scans were performed each time, and the resolution was set to 4 cm -1 .

5. The method for predicting cerebral hematoma injury time based on Fourier transform infrared spectroscopy technology according to claim 4, characterized in that: When the acquisition equipment in step S24 is pre-treated, the diamond crystal surface of the ATR probe is carefully cleaned with anhydrous ethanol, and the air is scanned as a background reference.

6. The method for predicting cerebral hematoma injury time based on Fourier transform infrared spectroscopy technology according to claim 5, characterized in that: During sample preparation, the solution was thoroughly mixed using a vortex mixer, and then 0.5 μL of the sample was deposited on the surface of the diamond crystal. The sample was gently dried using a hair dryer set to low temperature until a uniform film was obtained, which took 5-10 min.

7. The method for predicting cerebral hematoma injury time based on Fourier transform infrared spectroscopy technology according to claim 6, characterized in that: The spectral data is preprocessed in step S3, which includes the following steps: S31: Standard normal variation (SNV) transformation was used to mitigate the effects of surface scattering, particle size, and optical range variation. After the SNV transformation, the second-order derivative transformation was performed using the Savitzky-Golay algorithm, which used a window size of 9 points and a second-order polynomial. S32: In terms of identifying and excluding outliers, Q residuals and leverage values ​​are used as standard indicators to promote the development of robust models and improve the reliability of predictions; S33: Four spectra were finally identified as outliers and excluded from further data analysis for several groups of hematoma samples collected, of which 70% were allocated for the calibration dataset and the remaining 30% for the prediction dataset.

8. The method for predicting the time of cerebral hematoma injury based on Fourier transform infrared spectroscopy technology according to claim 7, characterized in that: The chemometric method in the S4 step is based on statistical modeling and includes supervised and unsupervised methods. The unsupervised program is mainly used for data exploration, using principal component analysis (PCA) to find similarities and differences between data. The supervised model includes two types: regression and classification. The regression method uses partial least squares (PLS).

9. The method for predicting cerebral hematoma injury time based on Fourier transform infrared spectroscopy technology according to claim 8, characterized in that: The software was combined with MATLAB R2016a and PLS Toolbox 8.2, and multiple indicators were used, including the coefficient of determination (R 2 ), the root mean square error of cross validation (RMSECV), the root mean square error of prediction (RMSEP), and the ratio of RMSECV to RMSEP.

Images