Spectral phase-based automatic calibration scheme for swept source-based optical coherence tomography systems

Distinctive approach in brain tumor detection and feature extraction using biologically inspired DWT method and SVM

Brain tumors result from uncontrolled cell growth, potentially leading to fatal consequences if left untreated. While significant efforts have been made with some promising results, the segmentation and classification of brain tumors remain challenging due to their diverse locations, shapes, and sizes. In this study, we employ a combination of Discrete Wavelet Transform (DWT) and Principal Component Analysis (PCA) to enhance performance and streamline the medical image segmentation process. Proposed method using Otsu's segmentation method followed by PCA to identify the most informative features. Leveraging the grey-level co-occurrence matrix, we extract numerous valuable texture features. Subsequently, we apply a Support Vector Machine (SVM) with various kernels for classification. We evaluate the proposed method's performance using metrics such as accuracy, sensitivity, specificity, and the Dice Similarity Index coefficient. The experimental results validate the effectiveness of our approach, with recall rates of 86.9%, precision of 95.2%, F-measure of 90.9%, and overall accuracy. Simulation of the results shows improvements in both quality and accuracy compared to existing techniques. In results section, experimental Dice Similarity Index coefficient of 0.82 indicates a strong overlap between the machine-extracted tumor region and the manually delineated tumor region.

MRChexNet: Multi-modal bridge and relational learning for thoracic disease recognition in chest X-rays

While diagnosing multiple lesion regions in chest X-ray (CXR) images, radiologists usually apply pathological relationships in medicine before making decisions. Therefore, a comprehensive analysis of labeling relationships in different data modes is essential to improve the recognition performance of the model. However, most automated CXR diagnostic methods that consider pathological relationships treat different data modalities as independent learning objects, ignoring the alignment of pathological relationships among different data modalities. In addition, some methods that use undirected graphs to model pathological relationships ignore the directed information, making it difficult to model all pathological relationships accurately. In this paper, we propose a novel multi-label CXR classification model called MRChexNet that consists of three modules: a representation learning module (RLM), a multi-modal bridge module (MBM) and a pathology graph learning module (PGL). RLM captures specific pathological features at the image level. MBM performs cross-modal alignment of pathology relationships in different data modalities. PGL models directed relationships between disease occurrences as directed graphs. Finally, the designed graph learning block in PGL performs the integrated learning of pathology relationships in different data modalities. We evaluated MRChexNet on two large-scale CXR datasets (ChestX-Ray14 and CheXpert) and achieved state-of-the-art performance. The mean area under the curve (AUC) scores for the 14 pathologies were 0.8503 (ChestX-Ray14) and 0.8649 (CheXpert). MRChexNet effectively aligns pathology relationships in different modalities and learns more detailed correlations between pathologies. It demonstrates high accuracy and generalization compared to competing approaches. MRChexNet can contribute to thoracic disease recognition in CXR.

A multi-band centroid contrastive reconstruction fusion network for motor imagery electroencephalogram signal decoding

Motor imagery (MI) brain-computer interface (BCI) assist users in establishing direct communication between their brain and external devices by decoding the movement intention of human electroencephalogram (EEG) signals. However, cerebral cortical potentials are highly rhythmic and sub-band features, different experimental situations and subjects have different categories of semantic information in specific sample target spaces. Feature fusion can lead to more discriminative features, but simple fusion of features from different embedding spaces leading to the model global loss is not easily convergent and ignores the complementarity of features. Considering the similarity and category contribution of different sub-band features, we propose a multi-band centroid contrastive reconstruction fusion network (MB-CCRF). We obtain multi-band spatio-temporal features by frequency division, preserving the task-related rhythmic features of different EEG signals; use a multi-stream cross-layer connected convolutional network to perform a deep feature representation for each sub-band separately; propose a centroid contrastive reconstruction fusion module, which maps different sub-band and category features into the same shared embedding space by comparing with category prototypes, reconstructing the feature semantic structure to ensure that the global loss of the fused features converges more easily. Finally, we use a learning mechanism to model the similarity between channel features and use it as the weight of fused sub-band features, thus enhancing the more discriminative features, suppressing the useless features. The experimental accuracy is 79.96% in the BCI competition Ⅳ-Ⅱa dataset. Moreover, the classification effect of sub-band features of different subjects is verified by comparison tests, the category propensity of different sub-band features is verified by confusion matrix tests and the distribution in different classes of each sub-band feature and fused feature are showed by visual analysis, revealing the importance of different sub-band features for the EEG-based MI classification task.

Aberrant intrinsic functional brain topology in methamphetamine-dependent individuals after six-months of abstinence

Our aim was to explore the aberrant intrinsic functional topology in methamphetamine-dependent individuals after six months of abstinence using resting-state functional magnetic imaging (rs-fMRI). Eleven methamphetamines (MA) abstainers who have abstained for six months and eleven healthy controls (HC) were recruited for rs-fMRI examination. The graph theory and functional connectivity (FC) analysis were employed to investigate the aberrant intrinsic functional brain topology between the two groups at multiple levels. Compared with the HC group, the characteristic shortest path length ($ {L}_{p} $) showed a significant decrease at the global level, while the global efficiency ($ {E}_{glob} $) and local efficiency ($ {E}_{loc} $) showed an increase considerably. After FDR correction, we found significant group differences in nodal degree and nodal efficiency at the regional level in the ventral attentional network (VAN), dorsal attentional network (DAN), somatosensory network (SMN), visual network (VN) and default mode network (DMN). In addition, the NBS method presented the aberrations in edge-based FC, including frontoparietal network (FPN), subcortical network (SCN), VAN, DAN, SMN, VN and DMN. Moreover, the FC of large-scale functional brain networks revealed a decrease within the VN and SCN and between the networks. These findings suggest that some functions, e.g., visual processing skills, object recognition and memory, may not fully recover after six months of withdrawal. This leads to the possibility of relapse behavior when confronted with MA-related cues, which may contribute to explaining the relapse mechanism. We also provide an imaging basis for revealing the neural mechanism of MA-dependency after six months of abstinence.

An adversarially consensus model of augmented unlabeled data for cardiac image segmentation (CAU+)

High quality medical images play an important role in intelligent medical analyses. However, the difficulty of acquiring medical images with professional annotation makes the required medical image datasets, very expensive and time-consuming. In this paper, we propose a semi-supervised method, $ {\mathrm{C}\mathrm{A}\mathrm{U}}^{+} $, which is a consensus model of augmented unlabeled data for cardiac image segmentation. First, the whole is divided into two parts: the segmentation network and the discriminator network. The segmentation network is based on the teacher student model. A labeled image is sent to the student model, while an unlabeled image is processed by CTAugment. The strongly augmented samples are sent to the student model and the weakly augmented samples are sent to the teacher model. Second, $ {\mathrm{C}\mathrm{A}\mathrm{U}}^{+} $ adopts a hybrid loss function, which mixes the supervised loss for labeled data with the unsupervised loss for unlabeled data. Third, an adversarial learning is introduced to facilitate the semi-supervised learning of unlabeled images by using the confidence map generated by the discriminator as a supervised signal. After evaluating on an automated cardiac diagnosis challenge (ACDC), our proposed method $ {\mathrm{C}\mathrm{A}\mathrm{U}}^{+} $ has good effectiveness and generality and $ {\mathrm{C}\mathrm{A}\mathrm{U}}^{+} $ is confirmed to have a improves dice coefficient (DSC) by up to 18.01, Jaccard coefficient (JC) by up to 16.72, relative absolute volume difference (RAVD) by up to 0.8, average surface distance (ASD) and 95% Hausdorff distance ($ {HD}_{95} $) reduced by over 50% than the latest semi-supervised learning methods.

Abnormal static and dynamic amplitude of low-frequency fluctuations in multiple brain regions of methamphetamine abstainers

Methamphetamine (meth) addiction is a significant social and public health problem worldwide. The relapse rate of meth abstainers is significantly high, but the underlying physiological mechanisms are unclear. Therefore, in this study, we performed resting-state functional magnetic resonance imaging (rs-fMRI) analysis to detect differences in the spontaneous neural activity between the meth abstainers and the healthy controls, and identify the physiological mechanisms underlying the high relapse rate among the meth abstainers. The fluctuations and time variations in the blood oxygenation level-dependent (BOLD) signal of the local brain activity was analyzed from the pre-processed rs-fMRI data of 11 meth abstainers and 11 healthy controls and estimated the amplitude of low-frequency fluctuations (ALFF) and the dynamic ALFF (dALFF). In comparison with the healthy controls, meth abstainers showed higher ALFF in the anterior central gyrus, posterior central gyrus, trigonal-inferior frontal gyrus, middle temporal gyrus, dorsolateral superior frontal gyrus, and the insula, and reduced ALFF in the paracentral lobule and middle occipital gyrus. Furthermore, the meth abstainers showed significantly reduced dALFF in the supplementary motor area, orbital inferior frontal gyrus, middle frontal gyrus, medial superior frontal gyrus, middle occipital gyrus, insula, middle temporal gyrus, anterior central gyrus, and the cerebellum compared to the healthy controls ($ P < 0.05 $). These data showed abnormal spontaneous neural activity in several brain regions related to the cognitive, executive, and other social functions in the meth abstainers and potentially represent the underlying physiological mechanisms that are responsible for the high relapse rate. In conclusion, a combination of ALFF and dALFF analytical methods can be used to estimate abnormal spontaneous brain activity in the meth abstainers and make a more reasonable explanation for the high relapse rate of meth abstainers.

Fall detection based on dynamic key points incorporating preposed attention

Accidental falls pose a significant threat to the elderly population, and accurate fall detection from surveillance videos can significantly reduce the negative impact of falls. Although most fall detection algorithms based on video deep learning focus on training and detecting human posture or key points in pictures or videos, we have found that the human pose-based model and key points-based model can complement each other to improve fall detection accuracy. In this paper, we propose a preposed attention capture mechanism for images that will be fed into the training network, and a fall detection model based on this mechanism. We accomplish this by fusing the human dynamic key point information with the original human posture image. We first propose the concept of dynamic key points to account for incomplete pose key point information in the fall state. We then introduce an attention expectation that predicates the original attention mechanism of the depth model by automatically labeling dynamic key points. Finally, the depth model trained with human dynamic key points is used to correct the detection errors of the depth model with raw human pose images. Our experiments on the Fall Detection Dataset and the UP-Fall Detection Dataset demonstrate that our proposed fall detection algorithm can effectively improve the accuracy of fall detection and provide better support for elderly care.

Facial expression recognition using lightweight deep learning modeling

Facial expression is a type of communication and is useful in many areas of computer vision, including intelligent visual surveillance, human-robot interaction and human behavior analysis. A deep learning approach is presented to classify happy, sad, angry, fearful, contemptuous, surprised and disgusted expressions. Accurate detection and classification of human facial expression is a critical task in image processing due to the inconsistencies amid the complexity, including change in illumination, occlusion, noise and the over-fitting problem. A stacked sparse auto-encoder for facial expression recognition (SSAE-FER) is used for unsupervised pre-training and supervised fine-tuning. SSAE-FER automatically extracts features from input images, and the softmax classifier is used to classify the expressions. Our method achieved an accuracy of 92.50% on the JAFFE dataset and 99.30% on the CK+ dataset. SSAE-FER performs well compared to the other comparative methods in the same domain.

SAMS-Net: Fusion of attention mechanism and multi-scale features network for tumor infiltrating lymphocytes segmentation

Automatic segmentation of tumor-infiltrating lymphocytes (TILs) from pathological images is essential for the prognosis and treatment of cancer. Deep learning technology has achieved great success in the segmentation task. It is still a challenge to realize accurate segmentation of TILs due to the phenomenon of blurred edges and adhesion of cells. To alleviate these problems, a squeeze-and-attention and multi-scale feature fusion network (SAMS-Net) based on codec structure, namely SAMS-Net, is proposed for the segmentation of TILs. Specifically, SAMS-Net utilizes the squeeze-and-attention module with the residual structure to fuse local and global context features and boost the spatial relevance of TILs images. Besides, a multi-scale feature fusion module is designed to capture TILs with large size differences by combining context information. The residual structure module integrates feature maps from different resolutions to strengthen the spatial resolution and offset the loss of spatial details. SAMS-Net is evaluated on the public TILs dataset and achieved dice similarity coefficient (DSC) of 87.2% and Intersection of Union (IoU) of 77.5%, which improved by 2.5% and 3.8% compared with UNet. These results demonstrate the great potential of SAMS-Net in TILs analysis and can further provide important evidence for the prognosis and treatment of cancer.

A weakly supervised learning-based segmentation network for dental diseases

With the development of deep learning, medical image segmentation has become a promising technique for computer-aided medical diagnosis. However, the supervised training of the algorithm relies on a large amount of labeled data, and the private dataset bias generally exists in previous research, which seriously affects the algorithm's performance. In order to alleviate this problem and improve the robustness and generalization of the model, this paper proposes an end-to-end weakly supervised semantic segmentation network to learn and infer mappings. Firstly, an attention compensation mechanism (ACM) aggregating the class activation map (CAM) is designed to learn complementarily. Then the conditional random field (CRF) is introduced to prune the foreground and background regions. Finally, the obtained high-confidence regions are used as pseudo labels for the segmentation branch to train and optimize using a joint loss function. Our model achieves a Mean Intersection over Union (MIoU) score of 62.84% in the segmentation task, which is an effective improvement of 11.18% compared to the previous network for segmenting dental diseases. Moreover, we further verify that our model has higher robustness to dataset bias by improved localization mechanism (CAM). The research shows that our proposed approach improves the accuracy and robustness of dental disease identification.

Research on chest radiography recognition model based on deep learning

With the development of medical informatization and against the background of the spread of global epidemic, the demand for automated chest X-ray detection by medical personnel and patients continues to increase. Although the rapid development of deep learning technology has made it possible to automatically generate a single conclusive sentence, the results produced by existing methods are not reliable enough due to the complexity of medical images. To solve this problem, this paper proposes an improved RCLN (Recurrent Learning Network) model as a solution. The model can generate high-level conclusive impressions and detailed descriptive findings sentence-by-sentence and realize the imitation of the doctoros standard tone by combining a convolutional neural network (CNN) with a long short-term memory (LSTM) network through a recurrent structure, and adding a multi-head attention mechanism. The proposed algorithm has been experimentally verified on publicly available chest X-ray images from the Open-i image set. The results show that it can effectively solve the problem of automatic generation of colloquial medical reports.

Pressure Injury Link to Entropy of Abdominal Temperature

This study examined the association between pressure injuries and complexity of abdominal temperature measured in residents of a nursing facility. The temperature served as a proxy measure for skin thermoregulation. Refined multiscale sample entropy and bubble entropy were used to measure the irregularity of the temperature time series measured over two days at 1-min intervals. Robust summary measures were derived for the multiscale entropies and used in predictive models for pressure injuries that were built with adaptive lasso regression and neural networks. Both types of entropies were lower in the group of participants with pressure injuries (n=11) relative to the group of non-injured participants (n=15). This was generally true at the longer temporal scales, with the effect peaking at scale τ=22 min for sample entropy and τ=23 min for bubble entropy. Predictive models for pressure injury on the basis of refined multiscale sample entropy and bubble entropy yielded 96% accuracy, outperforming predictions based on any single measure of entropy. Combining entropy measures with a widely used risk assessment score led to the best prediction accuracy. Complexity of the abdominal temperature series could therefore serve as an indicator of risk of pressure injury.

An Evaluation of Choroidal and Retinal Nerve Fiber Layer Thicknesses Using SD-OCT in Children with Childhood IgA Vasculitis

Background: We aimed to evaluate choroidal and retinal nerve fiber layer (RNFL) thicknesses in children undergoing the childhood IgA vasculitis (IgAV). Methods: Fifty-two patients with IgAV aged 1−6 years and 54 healthy children were included. Cases’ age, sex, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), RNFL thicknesses, and choroidal thickness values were recorded. Results: Median foveal center choroidal thickness was 374.0 µm (315.0 to 452.0 µm) in the IgAV group and 349.5 µm (285.0 to 442.0 µm) in the control group (p = 0.001). Median average RNFL thickness was 110.0 µm (91.0 to 134.0 µm) in the IgAV group and 104.0 µm (89.0 to 117.0 µm) in the control group (p < 0.001). Choroidal and RNFL thicknesses were significantly greater in all quadrants in the IgAV group than in the control group. No correlation was determined between ESR or CRP and foveal center choroidal and average RNFL thicknesses. Conclusions: Our findings show that choroidal and RNFL thicknesses increased significantly in children undergoing childhood IgA vasculitis compared to the healthy control group. These findings show that the choroid and RNFL are also affected by the inflammatory process in IgAV, which is a systemic vasculitis. We think that the choroidal and RNFL thicknesses can be used as a biomarker for childhood IgAV.

Residual based attention-Unet combing DAC and RMP modules for automatic liver tumor segmentation in CT

Purpose: Due to the complex distribution of liver tumors in the abdomen, the accuracy of liver tumor segmentation cannot meet the needs of clinical assistance yet. This paper aims to propose a new end-to-end network to improve the segmentation accuracy of liver tumors from CT. Method: We proposed a hybrid network, leveraging the residual block, the context encoder (CE), and the Attention-Unet, called ResCEAttUnet. The CE comprises a dense atrous convolution (DAC) module and a residual multi-kernel pooling (RMP) module. The DAC module ensures the network derives high-level semantic information and minimizes detailed information loss. The RMP module improves the ability of the network to extract multi-scale features. Moreover, a hybrid loss function based on cross-entropy and Tversky loss function is employed to distribute the weights of the two-loss parts through training iterations. Results: We evaluated the proposed method in LiTS17 and 3DIRCADb databases. It significantly improved the segmentation accuracy compared to state-of-the-art methods. Conclusions: Experimental results demonstrate the satisfying effects of the proposed method through both quantitative and qualitative analyses, thus proving a promising tool in liver tumor segmentation.

Comparison Study of the Two Biometers Based on Swept-Source Optical Coherence Tomography Technology

This research aimed to investigate the potential differences in the parameters, including axial length (AL), central corneal thickness (CCT), anterior chamber depth (ACD), lens thickness (LT), flat keratometry (Kf), steep keratometry (Ks), mean keratometry (Km), astigmatism, white-to-white (WTW) distance, acquired rate, and intraocular lens (IOL) power, between the two swept-source optical coherence tomography (SS-OCT) biometers, the ANTERION (biometer A) and IOLMaster 700 (biometer B). In a prospective observational comparative case series study, we enrolled 198 eyes undergoing cataract surgery. The AL, CCT, ACD, LT, Kf, Ks, Km, astigmatism, WTW, acquired rate, and IOL power were assessed. McNemar tests compared the acquired rate, and the paired sample t-test compared the quantitative measurement results between the groups. Nineteen eyes were excluded owing to missing AL data for either biometer. Finally, data from 179 eyes were analyzed. Between the two devices, no significant difference was found in AL, astigmatism magnitude, J0, and J45, while significant differences existed in CCT, ACD, LT, Kf, Ks, Km, WTW, astigmatism axis, and IOL power; no statistical significance was found in the AL acquired rate (biometer A, 90.9% and biometer B, 93.9%). Approximately 65.4% of eyes demonstrated ≥0.5-D difference in IOL power between the two biometers. In conclusion, the two biometers showed significant differences in all measurements (CCT, ACD, LT, K, WTW, astigmatism axis, and IOL power), except for AL.

Development and Verification of an Adjustment Factor for Determining the Axial Length Using Optical Biometry in Silicone Oil-Filled Eyes

The aim of this prospective clinical study was to establish and verify an adaptation for axial length (AL) measurement in silicone oil (SO)-filled pseudophakic eyes with a Scheimpflug and partial coherence interferometry (PCI)-based biometer. The AL was measured with a Pentacam AXL (OCULUS Optikgeräte GmbH, Wetzler, Germany) and IOLMaster 700 (Carl Zeiss Meditec, Jena, Germany). The coefficients of variation (CoV) and the mean systematic difference (95% confidence interval (CI)) between the devices were calculated. After implementing a setting for measuring AL in tamponaded eyes with a Pentacam based on data of 29 eyes, another 12 eyes were examined for verification. The mean AL obtained with the Pentacam was 25.53 ± 1.94 mm (range: 21.70 to 30.76 mm), and with IOLMaster, 24.73 ± 1.97 mm (ranged 20.84 to 29.92 mm), resulting in a mean offset of 0.80 ± 0.08 mm (95% CI: 0.77, 0.83 mm), p < 0.001. The AL values of both devices showed a strong linear correlation (r = 0.999). Verification data confirmed good agreement, with a statistically and clinically non-significant mean difference of 0.02 ± 0.04 (95% CI: −0.01, 0.05) mm, p = 0.134. We implemented a specific adaptation for obtaining reliable AL values in SO-filled eyes with the Pentacam AXL.

Data reconciliation using MA-PCA and EWMA-PCA for large dimensional data

In process industries, measurements usually contain errors due to the improper instrumental variation, physical leakages in process streams and nodes, and inaccurate recording/reporting. Thus, these measurements violate the laws of conservation, and do not conform to process constraints. Data reconciliation (DR) is used to resolve the difference between measurements and constraints. DR is also used in reducing the effect of random errors and more accurately estimating the true values. A multivariate technique that is used to obtain estimates of true values while preserving the most significant inherent variation is Principal Component Analysis (PCA). PCA is used to reduce the dimensionality of the data with minimum information loss. In this paper, two new DR techniques are proposed moving-average PCA (MA-PCA) and exponentially weighted moving average PCA (EWMA-PCA) to improve the performance of DR and obtain more accurate and consistent data. These DR techniques are compared based on RMSE. Further, these techniques are analyzed for different values of sample size, weighting factor, and variances.

Directional optical coherence tomography reveals melanin concentration-dependent scattering properties of retinal pigment epithelium

Optical coherence tomography (OCT) is a powerful tool in ophthalmology that provides in vivo morphology of the retinal layers and their light scattering properties. The directional (angular) reflectivity of the retinal layers was investigated with focus on the scattering from retinal pigment epithelium (RPE). The directional scattering of the RPE was studied in three mice strains with three distinct melanin concentrations: albino (BALB/c), agouti (129S1/SvlmJ), and strongly pigmented (C57BL/6J). The backscattering signal strength was measured with a directional OCT system in which the pupil entry position of the narrow OCT beam can be varied across the dilated pupil of the eyes of the mice. The directional reflectivity of other retinal melanin-free layers, including the internal and external limiting membranes, and Bruch's membrane (albinos) were also measured and compared between the strains. The intensity of light backscattered from these layers was found highly sensitive to the angle of illumination, whereas the inner/outer segment (IS/OS) junctions showed a reduced sensitivity. The reflections from the RPE are largely insensitive in highly pigmented mice. The differences in directional scattering between strains shows that directionality decreases with an increase in melanin concentrations in RPE, suggesting increasing contribution of Mie scattering by melanosomes.

Quantitative optical coherence microscopy for the in situ investigation of the biofilm

This paper explores the potential of optical coherence microscopy (OCM) for the <italic<in situ</italic< monitoring of biofilm growth. The quantitative imaging of the early developmental biology of a representative biofilm, <italic<Klebsiella pneumonia</italic< (KP-1), was performed using a swept source-based Fourier domain OCM system. The growth dynamics of the KP-1 biofilms and their transient response under perturbation was investigated using the enface visualization of microcolonies and their spatial localization. Furthermore, the optical density (OD) and planar density of the biofilms are calculated using an OCM technique and compared with OD and colony forming units measured using standard procedures via the sampling of the flow-cell effluent.