Automatic sleep stage classification plays a crucial role in assessing sleep quality and diagnosing sleep disorders. While many automated methods exist for sleep stage classification, most rely solely on single-channel electroencephalogram (EEG) signals. In contrast, multi-channel polysomnography (PSG), which includes EEG, electrooculogram (EOG), electromyogram (EMG), and electrocardiogram (ECG) signals, provides more comprehensive and accurate sleep staging information. These signals not only reveal complex physiological changes during sleep but also aid in understanding the characteristics and variations of different sleep stages. This capability offers valuable support for both sleep research and clinical practice. Therefore, we propose MCTSleepNet, a Transformer-based model for multi-channel sleep staging. It takes signal images transformed via time-frequency features as input. The core architecture uses a Transformer encoder to capture joint features across multiple channels, complemented by residual connections to retain original input information. The experimental results demonstrate that our model outperforms other state-of-the-art techniques on the publicly available ISRUC-S3 dataset. MCTSleepNet is an effective multi-channel sleep staging method that significantly aids in clinical sleep staging tasks.
KEYWORDS: Image fusion, Denoising, Image contrast enhancement, Image filtering, Image enhancement, Digital filtering, Image processing, Gaussian filters, Human vision and color perception, RGB color model
When the scene reached the human eye or other sensor through the underwater medium, the original color property of the object could be basically lost, and the background of most images was blue-green. In this paper, to remove the blue-green color of the underwater image and increase the color contrast of the image, a novel fusion method of two images was proposed. The two images were derived from the filtering denoising result and the contrast enhancement result after whitebalancing version of the original degraded image. Then, based on the two images, the associated weight maps were designed to enhance edge texture and color contrast of the output image. Finally, to avoid artifacts in the reconstructed image, we adopted the multiscale fusion strategy to fuse the processed two images. Experiments showed that our algorithm achieved better image contrast and edge sharpness than other methods and obtained better exposure for darker areas of the image.
In recent years, the incidence of congenital malformations was increasing obviously due to the significant increase in the number of women with advanced maternal age after the implementation of two children policy in China. Traditional prenatal testing methods faced the problems of the high risk of miscarriage, low sensitivity, and time-consuming. Development of a novel method with the features of noninvasive, rapid, cost-effective and high sensitivity will be of vital clinical value for prenatal testing. This review provided an overview of the common birth defects and compared the merits and drawbacks among current most-used prenatal detection methods. The characteristics of spectroscopic technologies, as well as their applications in prenatal testing were summarized. Spectral karyotyping, real-time fluorescence quantitative PCR, SNP allele site analysis, etc. using fluorescence spectrum analysis method and Raman spectroscopy have been reported in the application of prenatal testing. Finally, a new idea by taking the advantages of SERS for Down syndrome detection in pregnant woman blood was proposed, which may be providing a promising approach for realizing rapid, sensitive and noninvasive prenatal testing.
Due to the vagueness of mobile video shooting at night, the blurry low-light images obtained from it hindered humans from acquiring visual information and computer vision algorithms. In this paper, to lower color and lightness distortion when increasing visibility, a novel brightness mapping function based on the camera mapping model was proposed by using the chi-squared distribution. Then, the well-exposed images were obtained by using the brightness evaluation technique and the brightness mapping function. Finally, an existing image deblurring algorithm based on convolution and dark channel was employed to help deblur well-exposed images. Experiments showed that our method could achieve accurate contrast and lightness enhancement than several state-of-the-art methods and obtain decent sharp well-exposed images.
KEYWORDS: Video, Signal processing, Cameras, Video processing, Sensors, Digital electronics, Analog electronics, Motion models, Data modeling, Visualization
Traditional sensors and systems for wrist pulse acquisition, known as inconvenient, contact, depends on the digital and analog circuits. In this paper, a method in video motion processing at one-demensional signal extraction and analysis is proposed. It can be applied for wrist pulse acquisition and based on subtle video motion amplification algorithm. This is the demostraction of a low-cost (only need ordinary camera on the cell phone), accurate method for cantact-free wrist pulse acquiring that is capable of performing measurements in different illumination conditions. As the ultimate recovered single and all periodic pulse signal shown, it can be demonstracted that the proposed method can get accurate pulse signal compared to traditional methods.
Colorectal cancer (CRC) is the third most common type of cancer and forth leading cause of cancer-related death. Early diagnosis is the key to long-term patient survival. Programmatic screening for the general population has shown to be cost-effective in reducing the incidence and mortality from CRC. Current CRC screening strategy relies on a broad range of test techniques such as fecal based tests and endoscopic exams. Occult blood tests like fecal immunochemical test is a cost effective way to detect CRC but have limited diagnostic values in detecting adenomatous polyp, the most treatable precursor to CRC. In the present work, we proposed the use of surface enhanced Raman spectroscopy (SERS) with silver nanoparticles as substrate to analyze blood plasma for detecting both CRC and adenomatous polyps. Blood plasma samples collected from healthy subjects and patients diagnosed with adenomas and CRC were prepared with nanoparticles and measured using a real-time fiber optic probe based Raman system. The collected SERS spectra are analyzed with partial least squares-discriminant analysis. Classification of normal versus CRC plus adenomatous polyps achieved diagnostic sensitivity of 86.4% and specificity of 80%. This exploratory study suggests that blood plasma SERS analysis has potential to become a screening test for detecting both CRC and adenomas.
Quantitative methods for noninvasive diagnosis of scars are a challenging issue in medicine. This work aims to implement a texture analysis method for quantitatively discriminating abnormal scars from normal scars based on second-harmonic generation (SHG) images. A local difference local binary pattern (LD-LBP) operator combined with a wavelet transform was explored to extract diagnosis features from scar SHG images that were related to the alteration in collagen morphology. Based on the quantitative parameters including the homogeneity, directional and coarse features in SHG images, the scar collagen SHG images were classified into normal or abnormal scars by a support vector machine classifier in a leave-one-out cross-validation procedure. Our experiments and data analyses demonstrated apparent differences between normal and abnormal scars in terms of their morphological structure of collagen. By comparing with gray level co-occurrence matrix, wavelet transform, and combined basic local binary pattern and wavelet transform with respect to the accuracy and receiver operating characteristic analysis, the method proposed herein was demonstrated to achieve higher accuracy and more reliable classification of SHG images. This result indicated that the extracted texture features with the proposed method were effective in the classification of scars. It could provide assistance for physicians in the diagnostic process.
Texture analysis plays a crucial role in identifying objects or regions of interest in an image. It has been applied to a variety of medical image processing, ranging from the detection of disease and the segmentation of specific anatomical structures, to differentiation between healthy and pathological tissues. Second harmonic generation (SHG) microscopy as a potential noninvasive tool for imaging biological tissues has been widely used in medicine, with reduced phototoxicity and photobleaching. In this paper, we clarified the principles of texture analysis including statistical, transform, structural and model-based methods and gave examples of its applications, reviewing studies of the technique. Moreover, we tried to apply texture analysis to the SHG images for the differentiation of human skin scar tissues. Texture analysis method based on local binary pattern (LBP) and wavelet transform was used to extract texture features of SHG images from collagen in normal and abnormal scars, and then the scar SHG images were classified into normal or abnormal ones. Compared with other texture analysis methods with respect to the receiver operating characteristic analysis, LBP combined with wavelet transform was demonstrated to achieve higher accuracy. It can provide a new way for clinical diagnosis of scar types. At last, future development of texture analysis in SHG images were discussed.
The capabilities of micro-Raman spectroscopy for differentiating normal and malignant nasopharyngeal tissues were evaluated. Raman scattering signals were acquired from 22 normal and 52 malignant nasopharyngeal tissue samples. Distinctive spectral differences in Raman spectra between normal and malignant nasopharyngeal tissues were found, particularly in the spectral ranges of 853, 937, 1094, 1209, 1268, 1290 to 1340, 1579, and 1660 cm −1 , which primarily contain signals related to proteins, DNA, and lipids. Compared to normal tissues, the band intensity located at 853, and 937 cm −1 were significantly lower for cancerous tissues (p<0.05 ), while the band intensity located at 1094, 1209, 1268, and 1579 cm −1 were significantly higher (p<0.05 ). The band intensity located at 1290 to 1340, and 1660 cm −1 were also higher for cancerous tissues; but the differences were not statistically significant (p>0.05 ). Principal component analysis (PCA) and linear discriminate analysis (LDA) were employed to generate diagnostic algorithms for classification of Raman spectra of the two nasopharyngeal tissue types. The PCA-LDA algorithms together with leave-one-out, cross-validation technique yielded diagnostic sensitivity of 92% and specificity of 82%. This work demonstrated that the Raman spectroscopy technique associated with PCA-LDA diagnostic algorithms has potential for improving the diagnosis of nasopharyngeal cancers.
In this paper, a novel algorithm is proposed for unsupervised segmentation that incorporates a local trinary pattern (LTP)
operator representation of textures under a geometric active contour framework. First, by combining the gray levels of
pixels with texture information of an image, this method can be used for segmentation of a texture image or a nontexture
image. And then, the method is modified to avoid the additional computation problem without re-initialization
repeatedly. The simulation experiments show that the proposed segmentation method is more efficient, accurate and fast.
Nasopharyngeal carcinoma (NPC) is one of the most common malignancies in china, with a deep and hidden
localization. Recently, methods for early diagnosis of NPC has become one of the most important research topics in
medical field. Early monitoring of morphological change of NPC cells during the carcinogenesis is of great importance,
and early information extracted from the NPC cells during the initial stage of NPC is critical for diagnosis and treatment.
In this paper, image processing methods for two-photon microscopic image of NPC cells was investigated with the
purpose of providing useful information for early diagnosis and treatment of NPC.
There is abundant information in a two-photon microscopic image of NPC cells, which can be analyzed and processed
by means of computer and image pattern processing algorithm. In this paper, firstly, a mathematical method of transform
of Bottom-hat based on Matlab platform was employed to enhance the image of NPC cells, making the image easier to
distinguish; Then, several classical edge detection algorithms were compared and discussed, for example, Roberts
operator, Prewitt operator, and Canny operator etc. According to the inherent characteristics of two-photon microscopic
image of NPC cells, corrosion algorithm was used to define the edge of NPC cells. Furthermore, the article gets the
iterative threshold segmentation after noise denoising, on the other hand, improved discriminant analysis was adopted for
threshold segmentation of NPC cells, better results were obtained.
Biomedical images denosing based on Partial Differential Equation are well-known for their good
processing results. General denosing methods based on PDE can remove the noises of images with gentle
changes and preserve more structure details of edges, but have a poor effectiveness on the denosing of
biomedical images with many texture details. This paper attempts to make an overview of biomedical images
texture detail denosing based on PDE. Subsequently, Three kinds of important image denosing schemes are
introduced in this paper: one is image denosing based on the adaptive parameter estimation total variation
model, which denosing the images based on region energy distribution; the second is using G norm on the
perception scale, which provides a more intuitive understanding of this norm; the final is multi-scale denosing
decomposition. The above methods involved can preserve more structures of biomedical images texture detail.
Furthermore, this paper demonstrates the applications of those three methods. In the end, the future trend of
biomedical images texture detail denosing Based on PDE is pointed out.
This hybrid segmentation algorithm presented is a combination of three traditional methods. It has some advantages that
traditional methods don't have: first, the edge detected is continuous; second, segmentation result is accurate; third,
overgrowing doesn't exist. The hybrid algorithm has been implemented on some images. Through the segmentation result, it is proved to be effective.
Image edge recognition is a crucial aspect of biomedical image processing. In this paper, a rapid gradient segmentation
method based on the depth-first traverse of images is presented. This method defines the data structure for the pixel
firstly, estimate and catch gradient from four pixels around the arbitrary point coming from an arbitrary pixel of image. If
the pixel satisfies the feature of edge, the edge perpendicular to the directions of gradient is processed by depth-first
traverse, the pixels are marked at the same time. It will withdraw when no pixels satisfy the feature in the directions, then
depth-first traverse from the next direction with gradient, mark the pixel as corner, and traverse the image completely.
The segmentation method has been applied to edge recognition of color biomedical image and other images. The
experimental results showed that edges and corners of biomedical image can be segmented obviously, and be easy to
identify.
Based on the analysis of skin morphological structure, the mechanism of cutaneous fluorescence emission and diffuse
reflectance formation, a six-layer skin optical model was developed, allowing the variation of blood content in both the
upper blood plexus and the deep blood plexus. Monte Carlo simulation was performed to examine the effect of varying
tissue blood contents on skin fluorescence and diffuse reflectance spectra. The results demonstrated that (1) Both
fluorescence and reflectance spectrum can reflect changes of blood content in skin tissue; (2) The impact of blood
content in the upper blood plexus on skin spectra intensity is far larger than that in the deep blood plexus, though there
is far less blood in the upper blood plexus; (3) Fluorescence and reflectance spectrum could be used to detect or analyze
changes of blood content in skin tissue, especially for treatment monitoring and for evaluating the severity of skin
diseases that involving the blood plexus or blood pathological changes in the upper dermis.
By the analysis of the DXF file format of three-dimension image, the scheme that uses the technology of spatial and triangular transformation is formulated. The arithmetic for this scheme was given out in this paper. The control software for this arithmetic has been used in CO2 three-dimension laser processing system. Compared with some different schemes, the advantage of using this scheme in laser processing system was also pointed out. This scheme can be popularized in the other laser processing system.
Wavelet transform was widely used in communication, data compression and image manipulation. Base on Wavelet transform, the model that is designed for enhancing the result of binary image in laser processing system was built. Software for this model to control CO2 laser processing system was wrote out. The result which was given in this paper indicated that this model can effectively improve the output quality of binary image in CO2 laser processing system.
Using the Y type fiber, high efficiency fiber coupler and A1GaInP diode laser has developed a new diode laser therapy apparatus with the wavelength of 650nm. The output laser power from the fiber of the apparatus can reach more than 100mW. The fiber localizer of the apparatus is easy for orientation during the therapy and the output laser power is very stable and reliable. The apparatus has been widely used in about 20 hospitals, and has successfully cured more than 50,000 patients who suffered from various diseases, which mainly including cervical spondylosis, traumatic infection, skin bruise, sudden deafness, tympanitis, tinnitus, infantile rhinallergosis, rheumatism, rheumatoism, scapulohumeral periarthritis, asthma, temporomandibular joint disturbance, ulcer, eczema ani, postoperative crissal edema, and so on. The apparatus has also well performed in the application of haematococcus pluvialis's laser mutagenesis breeding.
By studying the basic principle of C02.laser processing system, energy-adjust method, concentration parameter method and dither method which was used to process computer image are introduced to laser processing system to process gray image. Some new soft modules for these methods are developed which have been used in laser processing system for processing different material such as wood, marble, plastic. The data of the three processing time and the image quality are also given. By comparing the laboratory data, the conclusion is drawn out that the image quality can be improved with concentration parameter method and the processing time is shorted greatly at the same quality of image for different material by using dither method. This result can be popularized in the other laser processing system.
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.