With the continuous development of spectral analysis technology, on-line spectral analysis technology has been widely applied. It can real-time monitor the key links in the process of liquid transportation, and provide real-time guidance for reliable and efficient liquid delivery. In the process of non-sampling real-time measurement of the solution in a flexible conveying tube, the accuracy of the spectral analysis is reduced due to the differences in optical parameters of the flexible conveying tubes. Therefore, this paper studies the effects of differences in flexible conveying tubes on on-line spectral analysis. A standard solution calibration method was proposed to suppress the interference caused by the differences of flexible conveying tubes based on the modified Lambert Beer's law. The calibrated spectral data is modeled by partial least squares regression to reduce the analysis error introduced by the optical differences of the flexible conveying tubes. An experiment was designed to verify the feasibility of the method by using a mixed solution of Intra-lipid and India-Ink as an analytical sample and using a polyvinyl chloride (PVC) material tube as a flexible conveying tube. The experimental results show that the method of calibrating the differences of the flexible conveying tubes by the standard solution is feasible, and effectively inhibits the impact of the differences of the conveying tubes on the online spectral analysis.
In the spectral analysis for biological fluids compositions, the sensitivity of trace compositions reflected in the spectral is very small, which results in the low measurement accuracy. This paper proposes a “Multi-dimension and Multi-mode spectroscopy method” to improve the measurement accuracy of trace compositions by two aspects of spectral acquisition and data processing: measuring the biological fluids sample at multiple modes so that multiple spectral for each sample can be obtained which will carry more information about trace compositions, then connecting these spectral to increase the spectral data dimension that is equivalent to increase the number of constraint equations in the modeling process, the error will be reduced through more constraint equations. An experiment was designed: taking the cholesterol concentration (2.57-8.1mmol/L) in blood plasma as the tested object, the blood plasma was irradiated with tungsten lamp and ultraviolet light source respectively, ultraviolet light stimulates blood plasma can produces fluorescence, the obtained transmission spectrum and fluorescence spectrum were rearranged to build the model. Experimental results showed that the analysis accuracy of cholesterol had been significantly improved. This research provided a new thinking of the analysis for biological fluids trace compositions.
Multiple diseases such as breast tumor poses a great threat to women's health and life, while the traditional detection method is complex, costly and unsuitable for frequently self-examination, therefore, an inexpensive, convenient and efficient method for tumor self-inspection is needed urgently, and lesion localization is an important step. This paper proposes an self-examination method for positioning of a lesion.
The method adopts transillumination to acquire the hyperspectral images and to assess the spatial information of lesion. Firstly, multi-wavelength sources are modulated with frequency division, which is advantageous to separate images of different wavelength, meanwhile, the source serves as fill light to each other to improve the sensitivity in the low-lightlevel imaging. Secondly, the signal-to-noise ratio of transmitted images after demodulation are improved by frame accumulation technology. Next, gray distributions of transmitted images are analyzed. The gray-level differences is constituted by the actual transmitted images and fitting transmitted images of tissue without lesion, which is to rule out individual differences. Due to scattering effect, there will be transition zones between tissue and lesion, and the zone changes with wavelength change, which will help to identify the structure details of lesion. Finally, image segmentation is adopted to extract the lesion and the transition zones, and the spatial features of lesion are confirmed according to the transition zones and the differences of transmitted light intensity distributions. Experiment using flat-shaped tissue as an example shows that the proposed method can extract the space information of lesion.
Dynamic spectrum (DS) method is one of the noninvasive approaches to measure the concentration of components in human blood based on the application of photoplethysmogram (PPG). One of the targets of the DS method is to predict the hemoglobin concentration in human blood noninvasively. In previous works, the usually used wavelength in the spectrum is 600-1100 nm which is regarded as the analysis “window” in human tissues. Optimum wavelengths for measurements of hemoglobin concentration have not been investigated yet. In order to improve the precision and reliability of hemoglobin measurements, a method for wavelength selection based on two-dimension (2D) correlation spectroscopy has been studied in this paper. By analyzing the 2D correlation spectroscopy which is generated by the DS data from subject with different blood hemoglobin concentrations, the wavelength bands which are sensible to hemoglobin concentrations in DS can be found. We developed calibration models between the DS data and hemoglobin concentration based on data from 57 subjects. The correlation coefficient is 0.68 in the test set of the model using the whole wavelength band (600-1100nm), while in the test set of the model using the selected wavelength band (850- 950nm) the correlation coefficient is 0.87. Results show the feasibility of wavelength selection utilizing 2Dcorrelation spectroscopy.
Spatially resolved diffuse reflectance spectroscopy method has been proved to be more effective than single point spectroscopy method in the experiment to predict the concentration of the Intralipid diluted solutions. However, Intralipid diluted solution is simple, cannot be the representative of turbid liquids. Blood is a natural and meaningful turbid liquid, more complicate. Hemoglobin is the major constituent of the whole blood. And hemoglobin concentration is commonly used in clinical medicine to diagnose many diseases. In this paper, near infrared spatially resolved transmission spectra (NIRSRTS) and Partial Least Square Regression (PLSR) were used to predict the hemoglobin concentration of human blood. The results showed the prediction ability for hemoglobin concentration of the proposed method is better than single point transmission spectroscopy method. This paper demonstrated the feasibility of the spatially resolved diffuse reflectance spectroscopy method for practical liquid composition analysis. This research provided a new thinking of practical turbid liquid composition analysis.
Quantization operation in the imaging process rounds the measured values of intensities into integers. It neglects the possible differences between the intensities having the same rounded numbers, and therefore lowers the grayscale resolution of images. Although some methods have been proposed for the reconstruction of high-grayscale resolution images from multiple subpixel-shifted low-spatial-resolution and low-grayscale-resolution images, the problems of nonsmooth transition within regions and insufficient intensity levels still exist. A grayscale superresolution method based on fill light and a photographing apparatus are proposed to deal with the problems. The photographing apparatus can add fill lights with slightly different intensities to the captured images without changing the brightness of scenes. Our reconstruction method is based on the method of estimating a float number from several rounded integers. Then, a high-grayscale-resolution image is reconstructed from multiple low-grayscale-resolution images with slightly different intensity fill lights. Simulated data and real-world data have been used for the evaluation of the method, and the experimental results show that our method effectively improves the grayscale resolution. Besides, our method is convenient for a graphics processing unit implementation.
An appropriate method for spectrum extraction for better signal-to-noise ratio (SNR) and lower computational cost is essential in noninvasive detection. We have first studied two existing extraction methods for dynamic spectrum (DS) comparatively: frequency domain analysis and single trial estimation; after analyzing the advantages and disadvantages theoretically, a new method based on a fast digital lock-in amplifier (FDLIA) was developed to overcome the limitations of these two existing methods. The feasibility of the new method was verified by experiments and the results demonstrated that the FDLIA method based on DS had greatly simplified the computation of frequency domain analysis without the method error; moreover, the continuous signal was cut into several short segments in FDLIA and the gross errors from an episode of pulse wave were eliminated, and thus SNR improved. Therefore, the FDLIA method utilizing the advantages of both existing methods can be effectively realized in a general embedded system in real time for its simple algorithm.
The phase shift error (PSE) for DS detection is analyzed using frequency domain measure theory of DS. Two factors
lead to the difference among different pulse waves, and introduce PSE into DS. Firstly, the period, amplitude and the
base line of pulse wave are unstable. Secondly, the phases of pulse waves under different wavelengths are different. The
PSE of transmitted and reflected pulse waves are both discussed quantitatively. The results showed that the PSE is
correlated with the position of rectangular intercept window that intercepts pulses from the waves. It can be minimized
into about 10% by selecting the start points of the windows. Because of the sensor contact pressure, the shapes of
transmitted and reflected pulse waves are different. Thus the right intercept locations are at the start of ascending limb
and the dicroticpulse respectively according to minimum PSE rule.
In the research of non-invasive concentration blood measurement, the scattering behavior of the tissue may leads to
significant differences in the ideal Lambert Beer's law. In this paper, Monte Carlo method is used to analyses the blood
tissue's influence to the Dynamic Spectrum proposed by Professor LI Gang. The Dynamic Spectrum evaluating only the
pulsatile part of the entire optical signal, this approach is rather independent of individual or time changes in scattering
or absorption characteristics of the tissue. In this paper, Monte Carlo method is used to analyses the scattering
behavior of the blood, the influence of the scattering behavior of the skin tissue to the scattering behavior of the blood.
and their influence to the Dynamic Spectrum. The pulsatile part ofthe spectrum was modeled by performing simulations
of photon migration through the tissue for the diastolic and systolic states. With the simulation of the Monte Carlo
method. the diffuse reflectance and transmittance of the model was calculated, analyzed and compared. The scattering
behavior must be considered in the measurement of Dynamic Spectrum to get the high precision measurement. The error
caused by the transmittance is greater than the error caused by the diffuse reflectance. The thickness of the Epidermis
can influence the nonlinearity of the transmittance, and influence the value of the diffuse reflectance. The thickness of
the tissue can influence the scattering behavior of the tissue.
KEYWORDS: Spectroscopy, Blood, Arteries, Near infrared, Signal processing, Near infrared spectroscopy, Infrared spectroscopy, Tissues, Absorption, Process modeling
In order to reduce the interference of the individual discrepancy in the noninvasive measurement of blood composition, a
new MR spectroscopy- dynamic spectroscopy is put forward and a new near infrared spectrometer is developed for the
dynamic spectroscopy. Experiments indicated that the dynamic spectroscopy can reduce the interference of individual
discrepancy well.
Noninvasive determination of tissue optical properties is essential for clinical applications in medical diagnostics and
therapeutics. In recent years, several methods were successfully introduced to deduce the optical properties of
semi-infinite tissue model from spatially resolved (SR) diffuse reflectance. However, biological tissue is in fact not
homogeneous and usually exhibits a complicated layered structure. The previous methods are not always efficient for the
layered-structure tissue model. In this paper, we introduced a new method to determine the optical properties of the
two-layer medium from the steady-state spatially resolved diffuse reflectance, which is based on the theory of support
vector machine (SVM). The method was validated using the Monte Carlo algorithm generated reflectance from a
two-layer model that consists of a 5mm thick top layer and a semi-infinite bottom layer. The training and predicting time
of SVM are 20s and 5s respectively. The predictive errors of the proposed method were less than 2% for the top-layer
optical properties and less than 4% for the bottom-layer optical properties, showing that the SVM method has a higher
accuracy and a shorter training time comparing with other methods. The principle to deal with regression estimation
problems with SVis briefly introduced firstly. Then, the phantom experiment set and the results are described. In the
end, some limitations and strategies are also discussed.
In this paper, with reference to practical applications, we investigate the accuracy of the PCA-NN method in determining the optical properties μa and μs' from the spatially resolved relative reflectance data produced by Monte Carlo simulations. To test prediction performance of PCA-NN from the reflectance data with different lengths and different measurement noises, we constructed six PCA-NNs respectively corresponding to data length = 5, 10, 15, 20, 25 and 30 mm, which were trained by higher precision reflectance produced with photons = 107. Then lower precision reflectance generated with photons = 104, 2 × 104, 5 × 104, 7 × 104, 105, 2 × 105, 5 × 105, 7 × 105 and 106 were inputted to PCA-NNs to extract μa and μs' and the accuracy of μa and μs' was calculated, respectively. The results showed that for the reflectance with the same data length, the prediction errors of μa and μs' increase as the data noise increases; but for the reflectance with the same data precision, the errors decrease as the data length becomes longer. In conclusion, the preliminary results in this paper provide a guideline for choosing appropriate measurement conditions or estimating the prediction errors in reality.
A novel method combining the PCA-NN algorithm established on the single-layer tissue model and the genetic algorithm based on the two-layer diffusion model has been presented to determine the optical properties of the two-layer medium from the steady-state spatially resolved diffuse reflectance. In detail, we firstly employ the PCA-NN algorithm established on the semi-infinite tissue model to extract the optical properties of the top layer from the spatially resolved reflectance that results from the photons migrating mainly within the top layer. With the knowledge of the optical properties of the top layer, the optical properties of the bottom layer are then obtained by use of the genetic algorithm for fitting the two-layer diffusion model to the reflectance data far from the source. The method was validated using the Monte Carlo generated reflectance for the two-layer medium of skin overlying fat or skin overlying muscle. And, the skin thickness was assumed to be known a priori and fixed at 5 mm. The results showed that all the optical properties of two layers can be determined by the method with the accuracy of better than 10%.
Near-IR spectroscopy holds great promise for non-invasive concentration measurements of blood on the basis of its potential for reagent-less, nondestructive, and noninvasive measurements. The main difficulty for determining absolute or even exact relative concentrations is the scattering behavior of the tissue. This leads to significant differences in the ideal Lambert Beer's law. In this paper, the approach of the Dynamic Spectrum in the frequency domain was proposed by Professor LI Gang etc. is shown, it is based on Photo-plethysmography (PPG) with fast Fourier transforms. The magnitude of fundamental wave of the pulse wave at each wavelength divided by the peak value of the pulse wave, get the natural logarithm of quotient at each wavelength and then the Dynamic Spectrum in the frequency domain is got. Evaluating only the pulsatile part of the entire optical signal, this approach is rather independent of individual or time changes in scattering or absorption characteristics of the tissue. Because of the noise and the resolution of the spectrometer, the Dynamic Spectrum is very difficult to get. In this paper, a series of measures is taken, and high-precision Dynamic Spectrum in the frequency domain is got with the experiment. The approach is verified. The advantage of getting Dynamic Spectrum in the frequency domain is analyzed, and compared with the Dynamic Spectrum in the time domain. The paper shows that the technique enables high precision measurement of changes in tissue absorbance caused by blood pulsation. It is very important in the non-invasive in vivo concentration measurement of blood.
As a numerical experiment, Monte Carlo simulation (MCS) has been proven to be a credible and flexible method for predicting the distribution of light in random media. It has full control of many parameters of optical system, which may be cumbersome to obtain in a real experiment. In standard OCT system, confocal microscopy structure with different Numerical Aperture (NA) is selected to acquire superior transverse resolution and unique property of optical sectioning. But the effects of numerical aperture on the probing depth of OCT system are difficult to estimate. In this paper, a new Monte Carlo simulation model of OCT system based on confocal mode is put forward to simulate the confocal microscopy structure and focused gaussian beam. It makes up the deficiency of traditional MCS model, which can only be applied to infinity narrow beam. By applying this new model, the effects of NA on probing depth of OCT system are analyzed, and the estimation of critical probing depth of OCT system is discussed. Study indicates that a smaller numerical aperture has more advantage on the probing depth when the transverse resolution is ensured.
Noninvasive determination of μs' and μa is essential for clinical applications in medical diagnostics and therapeutics. Spatially resolved diffuse reflectance method is more advantageous than other techniques because of its simplicity and low-cost. The methods for solving the nonlinear inverse problem of estimates of μs' and μa from spatially resolved diffuse reflectance Rd(r) can be classified into the algorithms based on absolute or relative reflectance measurements in nature. Since absolute reflectance measurements are technically more difficult to perform than the relative one, study on the methods based on the relative reflectance has a more important meaning for real applications. Considering that there were several normalizations of Rd(r), in this paper we discussed the varieties of prediction rms errors of μs' and μa extracted from relative reflectance data of different normalization forms including Rd(r)/Rd(r)max, r2(Rd(r)/Rd(r)max), 1n(Rd(r)/Rd(r)max) and 1n(r2(Rd(r)/Rd(r)max)). Additionally, we compared the accuracies of μs' and μa determined from absolute reflectance data Rd(r) and 1n(Rd(r)) with that from relative reflectance data to study the loss of accuracy due to normalization. Rather than the traditional neural network methods, we used a new method -- PCA-NN trained with diffuse reflectance data from Monte Carlo simulations to derive μs' and μa. All the PCA-NNs were trained and tested on the space with μs' between 0.1 and 2.0 mm-1 and μa between 0.01 and 0.1 mm-1. The test results indicate that the rms errors in μs' and μa are 0.72% and 2.57% for Rd(r), 0.28% and 0.55% for 1n(Rd(r), 2.98% and 5.44% for Rd(r)/Rd(r)max, 2.22% and 3.21% for 1n(Rd(r)/Rd(r)max), 6.52% and 20.7% for r2(Rd(r)/Rd(r)max), and 2.22% and 3.21% for 1n(r2(Rd(r)/Rd(r)max)), suggesting that the normalization form 1n(Rd(r)/Rd(r)max) would be the first choice for the estimates of μs' and μa from relative reflectance data by PCA-NN. Although the loss of accuracy due to normalization is considerable, the preliminary results provide a guideline for relative reflectance measurements.
Increases in intracranial pressure (ICP) may occur in patients of cerebral edema, brain tumors, encephalitis, brain injury etc. The care of these patients has been improved with continuous ICP monitoring. However, all the present clinical ICP monitors are invasive. Non-invasive ICP monitor is greatly expected. In this paper, a new method for ICP non-invasive monitoring using near-infrared light is proposed. Both theoretical and experimental studies have shown that correlation analysis applied to ICP, cerebrospinal fluid (CSF) and near-infrared diffuse reflection light from the brain tissue provides a possibility for non-invasive ICP detection. First, the correlation between the reflected light and the thickness of CSF is studied with Monte-Carlo simulations. It is concluded that the intensity of the reflected light changes significantly with the thickness of CSF, suggesting the feasibility to detect ICP by measuring the diffuse reflection near-infrared light from the brain tissue because the thickness of CSF changes with ICP. Secondly, the correlation of the diffuse reflection light and ICP is studied based on the experimental data acquired with helps of volunteers. It has been proved very promising to determine ICP non-invasively using near-infrared light.
KEYWORDS: Digital signal processing, Optical coherence tomography, Data acquisition, Signal processing, Imaging systems, Tissues, Data processing, Demodulation, Modulators, Image processing
Optical Coherence Tomography (OCT) is a novel optical imaging technology that provides high-resolution cross-sectional views of subsurface microstructure of biological tissues. It features high sensitivity, noninvasiveness, high resolution of micron scale, probing depth of 2cm for transparent tissue and 1~2mm for highly scattering tissue, and etc. OCT has shown a promising future to become a complement to the conventional imaging techniques in the fields of medicine and biology. But there are still a number of problems should be solved before OCT technology can be applied to practical usage. One of those is the limited imaging speed. In this paper, a high-speed data acquisition and processing (DAP) system for OCT is presented. Built around the high-powered Digital signal Processor (DSP) TMS32OVC5410A, the system is designed to cooperate with a PC to realize the image-scanning control, signal acquisition, data processing, transmission, image reconstruction and display. Superior to conventional data acquisition systems (DAQs), this system implants pre-processing of raw data into DSP, thus reduces the image acquisition time by carrying out the large amount of computation in the DSP, rather than in the PC. In addition, it can present two images with different information at per 2-D scans. And the system can be extended to future diverse applications by loading flexible digital signal processing schemes.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.