CN-120651796-B - Raman spectrum analysis device for quantitative detection of blood tumor markers

CN120651796BCN 120651796 BCN120651796 BCN 120651796BCN-120651796-B

Abstract

The invention belongs to the technical field of biomedical detection and discloses a Raman spectrum analysis device for quantitatively detecting a blood tumor marker, which comprises a blood collection centrifugal unit with a self-adaptive load adjustment mechanism, an optical fiber Raman spectrometer and an automatic wavelength calibration system, wherein the blood collection centrifugal unit is used for monitoring blood density and components in real time and automatically adjusting centrifugal force and time parameters according to the blood density and components, and the optical fiber Raman spectrometer is provided with the automatic wavelength calibration system and is integrated with an ultrafast laser pulse emitter and a high-sensitivity photon counting detector. This improvement allows the device to accurately detect the presence of tumor markers at an early stage of cancer, providing a powerful support for early treatment of patients.

Inventors

ZHENG GUIXI
LI WEI
SUN MENGTING
AN HAOYUAN
GENG XUEYAN

Assignees

山东大学齐鲁医院

Dates

Publication Date: 20260512
Application Date: 20250619

Claims (10)

1. A Raman spectrum analysis device for quantitatively detecting a blood tumor marker is characterized by comprising the following components: The blood collection centrifugal unit is provided with a self-adaptive load regulation mechanism and is used for monitoring the density and the components of blood in real time and automatically regulating centrifugal force and time parameters according to the density and the components of the blood; The optical fiber Raman spectrometer is provided with an automatic wavelength calibration system, an ultrafast laser pulse emitter and a high-sensitivity photon counting detector are integrated on the optical fiber Raman spectrometer, and the optical fiber Raman spectrometer is based on a quantum entanglement light source and a nonlinear optical crystal to enhance Raman scattering signals and is used for synchronously capturing and analyzing spectrum data of a plurality of wavelengths; The image optimization unit is used for preprocessing the Raman spectrum image based on the self-adaptive noise suppression algorithm and the multi-scale image fusion algorithm, separating a target signal from complex background noise and enhancing the image through the high-resolution image reconstruction algorithm; The machine learning and diagnosis unit is based on an advanced deep learning framework combining a deep reinforcement learning algorithm and a multi-mode learning mechanism and comprises STCN networks for feature learning and mode recognition, and the machine learning and diagnosis unit further comprises an adaptive decision support module for updating and optimizing a detection model in real time according to clinical data and scientific research progress; cancer cell identification and quantification unit, based on multidimensional Raman spectrum characteristic analysis calculation model, combines machine learning algorithm and spectrum data processing algorithm, and defines spectrum data matrix as Wherein Representing the point of sampling of the spectrum, Representing the number of wavelength channels, and normalizing the optical data to obtain a normalized matrix Calculating the mean square error weighting matrix The formula is adopted: ; Calculating a spectral feature weighting index Constructing a nonlinear mapping relation between the number of cancer cells and the spectrum intensity; Wherein, the Representing the first of the matrix of raw spectral data The sampling point is at the first The spectral intensity of the individual wavelength channels, Representing the first of the normalized spectral data matrix The sampling point is at the first The spectral intensity of the individual wavelength channels, Represents the first The average of the spectral data is normalized for each wavelength channel, Representing the total number of spectral sample points, Representing the total number of wavelength channels, Represents the normalized post-first The sampling point is at the first Mean square error weighting values for the individual wavelength channels.
2. A Raman spectrum analysis apparatus for quantitative detection of a blood tumor marker according to claim 1, wherein the adaptive load adjustment mechanism employs an algorithm formula of F=f (ρ, C), wherein F represents centrifugal force, ρ represents blood density, C represents blood component concentration, and F represents an adaptive adjustment function.
3. The Raman spectrum analysis device for quantitative detection of blood tumor markers according to claim 2, wherein the quantum entanglement Raman signal enhancement algorithm is represented by the following formula I Raman =g (phi, χ, P), wherein I Raman represents the intensity of Raman scattering signals, phi represents the phase of a quantum entanglement light source, χ represents the conversion efficiency of a nonlinear optical crystal, P represents the laser pulse power, and g represents a signal enhancement function.
4. The Raman spectrum analysis device for quantitative detection of blood tumor markers according to claim 3, wherein the adaptive noise suppression algorithm uses the following formula: (x, y) =i (x, y) -h (N (x, y)), wherein (X, y) represents the denoised image, I (x, y) represents the original image, N (x, y) represents noise, and h represents a noise suppression function.
5. The Raman spectrum analysis device for quantitative detection of blood tumor markers according to claim 4, wherein the multi-scale image fusion algorithm is represented by the following formula I fused = W i ⋅I i (x, y), where I fused represents the fused image, I i (x, y) represents the I-th layer image details, and w i represents the fusion weight.
6. The Raman spectrum analysis device for quantitative detection of blood tumor markers according to claim 5, wherein the STCN network has a loss function formula as follows: ; Where L represents the loss function, Y true represents the true label, Y pred represents the predicted label, θ represents the network parameters, and λ represents the regularization coefficient for preventing overfitting.
7. The Raman spectrum analysis apparatus for quantitative detection of blood tumor markers according to claim 6, further comprising a data storage unit for storing all data generated during the detection process, including raw spectrum data, processed image data, using the formula: ; Calculating corrected spectral intensity The diagnosis result, the user information and the like are used for subsequent analysis and tracing; Wherein, the Representing the intensity of the spectrum after correction, Represents the first The raw spectral intensities of the individual sample points, Representing the average of the raw spectral intensities of all the sample points, Representing the standard deviation of the raw spectral intensities of all the sample points, Represents the first The weighting coefficients of the individual sample points are, Representing the total number of sample points, Representing the index of a single sample point.
8. A Raman spectrum analysis apparatus for quantitative detection of a blood tumor marker according to claim 1, further comprising an automatic calibration unit for calibrating the Raman spectrometer before each detection, ensuring accuracy and consistency of the spectrum data, the automatic calibration unit comprising a calibration light source for providing light of known spectral characteristics and a calibration sample for providing a substance of known Raman spectral characteristics for calibrating the Raman spectrometer before the detection.
9. The Raman spectrum analysis device for quantitative detection of blood tumor markers according to claim 1, further comprising a remote monitoring and diagnosis unit connected to the device via a network for remotely monitoring the operation state of the device, receiving detection data, and performing remote fault diagnosis and update.
10. The Raman spectrum analysis apparatus for quantitative detection of blood tumor markers according to claim 9, further comprising a user interface for displaying the detection result, providing an operation guide, and receiving user input.

Description

Raman spectrum analysis device for quantitative detection of blood tumor markers Technical Field The invention belongs to the technical field of biomedical detection, and particularly relates to a Raman spectrum analysis device for quantitative detection of a blood tumor marker. Background In the field of medical diagnostics, detection of hematological tumor markers is of vital importance for early detection, diagnosis and treatment of cancer. However, existing blood tumor marker detection techniques face a number of challenges. First, since the concentration of tumor markers in blood tends to be very low, especially in early stages of cancer, this makes the detection process extremely susceptible to various interference factors such as biological background noise, degradation during sample processing, and the like, resulting in insufficient sensitivity of detection. Second, the blood composition varies greatly from individual to individual, which makes conventional detection methods limited in accuracy and reliability. In addition, the existing detection technology can only provide qualitative detection results, and cannot accurately quantify tumor markers, which limits the application of the detection technology in clinical diagnosis and treatment to a certain extent, so that a Raman spectrum analysis device for quantitatively detecting blood tumor markers is provided to solve the problems. Disclosure of Invention The invention aims to provide a Raman spectrum analysis device for quantitative detection of blood tumor markers, which is used for solving the problems in the background technology. In order to achieve the above purpose, the invention provides a Raman spectrum analysis device for quantitative detection of blood tumor markers, comprising: The blood collection centrifugal unit is provided with a self-adaptive load regulation mechanism and is used for monitoring the density and the components of blood in real time and automatically regulating centrifugal force and time parameters according to the density and the components of the blood; The optical fiber Raman spectrometer is provided with an automatic wavelength calibration system, an ultrafast laser pulse emitter and a high-sensitivity photon counting detector are integrated on the optical fiber Raman spectrometer, and the optical fiber Raman spectrometer is based on a quantum entanglement light source and a nonlinear optical crystal to enhance Raman scattering signals and is used for synchronously capturing and analyzing spectrum data of a plurality of wavelengths; The image optimization unit is used for preprocessing the Raman spectrum image based on the self-adaptive noise suppression algorithm and the multi-scale image fusion algorithm, separating a target signal from complex background noise and enhancing the image through the high-resolution image reconstruction algorithm; The machine learning and diagnosis unit is based on an advanced deep learning framework combining a deep reinforcement learning algorithm and a multi-mode learning mechanism and comprises STCN networks for feature learning and mode recognition, and the machine learning and diagnosis unit further comprises an adaptive decision support module for updating and optimizing a detection model in real time according to clinical data and scientific research progress; cancer cell identification and quantification unit, based on multidimensional Raman spectrum characteristic analysis calculation model, combines machine learning algorithm and spectrum data processing algorithm, and defines spectrum data matrix as WhereinRepresenting the point of sampling of the spectrum,Representing the number of wavelength channels, and normalizing the optical data to obtain a normalized matrixCalculating the mean square error weighting matrixThe formula is adopted: ; Calculating a spectral feature weighting index Constructing a nonlinear mapping relation between the number of cancer cells and the spectrum intensity; Wherein, the Representing the first of the matrix of raw spectral dataThe sampling point is at the firstThe spectral intensity of the individual wavelength channels,Representing the first of the normalized spectral data matrixThe sampling point is at the firstThe spectral intensity of the individual wavelength channels,Represents the firstThe average of the spectral data is normalized for each wavelength channel,Representing the total number of spectral sample points,Representing the total number of wavelength channels,Represents the normalized post-firstThe sampling point is at the firstMean square error weighting values for the individual wavelength channels. Preferably, the adaptive load regulation mechanism employs an algorithm formula f=f (ρ, C), where F represents centrifugal force, ρ represents blood density, C represents blood constituent concentration, and F represents an adaptive regulation function. Preferably, the quantum entangled Raman signal enhancement algorithm adopts