The latest advancement in large foundational model, SAM2, has demonstrated significant potential in 3D medical image segmentation due to their capability to effectively segment video streams. However, its application in medical image segmentation presents challenges, requiring extensive training on medical images or high-quality prompts provided by experts to achieve optimal performance. To address the aforementioned limitations, we propose SAM2-SP, which adopts Low-Rank Adaption for parameter-efficient fine-tuning and introduces a novel dynamic self-prompting strategy that generates most confident prompt templates from voxel features, enabling SAM2 to achieve domain adaptation in medical image segmentation without reliance on expert-level prompts. Extensive experiments show that SAM2-SP achieves state-of-the-art performance on the public Synapse dataset and the private EDC dataset, and even outperforms the compared task-specific segmentation approaches, the vanilla SAM and other SAM-based approaches.

Persistent inward currents (PICs) play a crucial role in regulating neuronal excitability. These currents are composed of calcium (CaL) and sodium (NaP) components in vertebrate spinal neurons. Recent studies have reported that PICs are expressed in serotonergic neurons (5-HT) in medulla of mice. Multiple patterns of PICs were identified in 5-HT neurons, corresponding to a range of distinct repetitive firing types. The mechanisms underlying formation of these PIC patterns and firing types remain unknown. Using combined modeling and experimental approaches we explored the ionic mechanisms responsible for the PIC patterns and firing types. The whole cell patch clamp recordings were performed on the medullary 5-HT neurons of postnatal day 3-6 mice. A 5-HT neuron model was built based on the membrane properties of the 5-HT neurons and kinetics of voltage-gated channels. Results from physiological experiments and modeling simulations included: (1) PICs in 5-HT neurons were classified into six patterns based on their current trajectory induced by bi-ramp voltage, while repetitive firings were categorized into three types according to their response to bi-ramp current. Modulation of PICs conductance and kinetics altered the PIC patterns and firing types. (2) NaP conductance contributed to amplitude of PICs, whereas the slow inactivation gate (Sgate) of NaP regulated the PIC patterns and firing types. Increasing Sgate changed trajectory of PICs from counterclockwise to clockwise and firing types from asymmetrical to symmetric types induced by bi-ramp current. (3) CaL conductance dominated the amplitude of PICs, while CaL kinetics (half-activation voltage and slope) determined inactivation of PICs and prolongation of repetitive firing. (4) The novel finding was that distribution of CaL in distal dendrites modulated the PIC patterns and firing types. This study provides insights into the ionic mechanisms underlying generation of multiple PIC patterns and firing types in 5-HT neurons.

Ultra-high field magnetic resonance imaging (MRI), such as 7-Tesla (7T) MRI, provides significantly enhanced tissue contrast and anatomical details compared to 3T MRI. However, 7T MRI scanners are more costly and less accessible in clinical settings than 3T scanners. In this paper, we propose a wavelet-based frequency attention network (WFANet) and a semi-supervised method named dual domain consistency learning (DDCL), and combine them to form a WFANet-DDCL framework for 7T MRI synthesis. WFANet leverages the frequency sensitivity of the proposed wavelet-based frequency attention encoder (WFAE) along with the large receptive field of dilated convolution. WFAE is proposed as an independent module to capture multi-scale frequency attention via the proposed wavelet-based frequency attention (WFA) mechanism. WFAE can be integrated into any backbone network as a plug-and-play component and improve network performance. To tackle the challenge of limited paired data for network training, DDCL is proposed to take advantage of both paired and unpaired data. Frequency domain perturbation is proposed and combined with Gaussian noise to regularize the supervised learning process in dual domains, better avoiding overfitting. Extensive experimental results demonstrate that WFANet-DDCL can achieve comparable performance to state-of-the-art supervised methods even using 66% of all paired data.

In non-contact respiratory monitoring, reducing motion artifact and selecting the appropriate Region of Interest (ROI) pose significant challenges. Most motion artifact removal methods rely on signal periodicity assumptions, while respiratory signals usually are non-periodic in real-world scenarios. Existing automated ROI selection approaches are mostly primarily impacted by the texture of clothing, absence of chest landmarks, and obstruction of face. To improve the quality of respiratory signals, in this study, we propose a framework for automatic respiratory ROI selection based on video, namely, Optimizing Video-based Respiration Monitoring (OVRM), which consists of peak-trough adaptive motion artifact removal and characteristic-driven adaptive ROI selection. This motion artifact removal strategy removes motion artifacts by using a dynamic ratio-based judgment mechanism, and reconstructs signals using sinusoidal interpolation. The adaptive ROI method scores signals based on periodicity, similarity, smoothness, and energy, selecting the highest-scoring blocks as the ROIs to match respiratory signals efficiently. Experimental results, validated across four datasets, demonstrate that OVRM effectively reduces signal noise caused by subject movement and outperforms state-of-the-art non-contact respiratory monitoring algorithms. The dataset and code are publicly available at:https://github.com/zxx5058/OVRM.

Ground-penetrating radar (GPR) is a pivotal noninvasive tool that yields subsurface images critical to archeology, near-surface characterization, geotechnical studies, and disaster response. The antenna central frequency of the GPR system has a significant impact on penetration depth and resolution. Lower antenna frequencies penetrate deeper but at lower resolutions, while higher frequencies offer detailed images at reduced depths. Therefore, improving the resolution of low-frequency radar with increased detection depth is an essential research focus. Inspired by image super-resolution advancements, supervised deep learning methods that rely on strictly paired training data have achieved remarkable success. However, acquiring such paired samples in practical scenarios is often a formidable challenge. To tackle this, we propose a novel resolution enhancement technique through weakly supervised learning, effectively addressing the scarcity of strictly paired samples in real-world situations. We utilize two sets of antennas with different central frequencies to construct our training data, with a low-frequency antenna as input and a high-frequency antenna as the learning target. A cycle-consistent generative adversarial network (Cycle-GAN) is trained to discern the mapping between low-resolution inputs and unpaired high-resolution data. The refined network is then employed to improve low-frequency GPR data resolution. Our work is validated on synthetic and real-world datasets. The proposed method effectively strengthens critical high-frequency details for finer imaging and broadens the frequency bandwidth. Significantly, it enhances resolution without compromising the detection depth of low-resolution GPR data, marking a substantial advancement in subsurface imaging technology.

Microscopic hyperspectral imaging technology combined with deep learning method emerges medical field recently as a multiplexed imaging technology. With the semantic segmentation of hyperspectral histopathological image of pathological tissue, doctors can quickly locate suspicious areas, diagnose and arrange treatment accurately and rapidly, reducing the workload of them. Cholangiocarcinoma is a rare and devastating disease with few hyperspectral histopathological data. Moreover, achieving high-quality annotations of hyperspectral histopathological image is challenging and costs time for pathologists, so generally, rough labels are annotated, but directly using the low-quality labels will reduce the performance of segmentation networks. So how to fully utilize few high-quality annotations and dozens of low-quality labels to enhance the segmentation performance of cholangiocarcinoma hyperspectral image remains to be resolved. In this paper, we proposed a two-stage hyperspectral segmentation deep learning framework based on Labels-to-Photo translation and Swin-Spec Transformer(L2P-SST). In stage-I, the OASIS generative network and the Swin-Spec Transformer discriminative network are used for adversarial training, and a spectral perceptual loss function is proposed to generate highquality hyperspectral images; in stage-II, parameters of the generative network is fixed and the generated hyperspectral images are used as data augmentation in the training of Swin-Spec Transformer segmentation network. The proposed framework achieved 76.16% mIoU(mean Intersection over Union), 85.80% mDice(mean Dice), 90.96% Accuracy and 71.65% Kappa coefficient in the semantic segmentation task of the Multidimensional Choledoch Database. Compared with other methods, the results demonstrate our framework provides a competitive segmentation performance.

Topic modeling extracts useful potential topics that reflect market information from massive financial news and is widely used in data mining and economic research. Traditional topic modeling approaches such as Latent Dirichlet Allocation (LDA) and Non-Negative Matrix Factorization (NMF) lack semantic information, and short texts have feature sparse problems. We develop a topic clustering model based on BERT-LDA joint embedding that takes both contextual semantics and thematic narrative into account. We cluster document embeddings with the HDBSCAN algorithm and utilize a class-based TF-IDF (c-TF-IDF) method to create topic representations. Empirical results show that the BERT-LDA model is competitive compared with traditional and single topic models. It generates coherent topic words that are dissimilar to each other.

Skin biopsy histopathological analysis is one of the primary methods used for pathologists to assess the presence and deterioration of melanoma in clinical. A comprehensive and reliable pathological analysis is the result of correctly segmented melanoma and its interaction with benign tissues, and therefore providing accurate therapy. In this study, we applied the deep convolution network on the hyperspectral pathology images to perform the segmentation of melanoma. To make the best use of spectral properties of three dimensional hyperspectral data, we proposed a 3D fully convolutional network named Hyper-net to segment melanoma from hyperspectral pathology images. In order to enhance the sensitivity of the model, we made a specific modification to the loss function with caution of false negative in diagnosis. The performance of Hyper-net surpassed the 2D model with the accuracy over 92%. The false negative rate decreased by nearly 66% using Hyper-net with the modified loss function. These findings demonstrated the ability of the Hyper-net for assisting pathologists in diagnosis of melanoma based on hyperspectral pathology images.

Hyperspectral LiDAR (HSL) possesses the advantages of the LiDAR and the hyperspectral detection, and detects ranging and spectrum information synchronously, by one HSL system. The data fusion is also avoided. At present, the spectrum range of reported HSLs usually covers only 500 nm-1000 nm (from visual (VIS) to near infrared (NIR) band). However, there is requirement to extend the spectrum range to short wave infrared (SWIR) band, which often contains more useful spectral information. In this paper, a HSL covering the spectrum from VIS to SWIR is reported. In the HSL, the echoes are divided into two sections and are detected by the different optoelectronic devices, of which the spectral response ranges are respectively compatible to the corresponding echoes. The HSL detection experiment in the laboratory was carried out. The waveforms of the echoes were analyzed, and the spectra of different targets were measured by the HSL. The experiment results demonstrate the capability of the prototyped HSL that obtaining the ranging information and the spectrum information of the targets in VIS-SWIR bands synchronously.

Recent success in large-scale image recognition challenge (i.e., ImageNet) fully demonstrates the capability of deep neural network (DNN) in learning complex and semantic representation, and this also motivates the generation of transfer learning model, which fine-tunes state-of-the-art DNN models with other small-scale databases for better performance. Driven by such an idea, a task-specific DNN model fine-tuned from VGG-face is constructed for both gender and identity recognition with hand vein information. Unlike the traditional transfer learning models, which fine-tune directly from source to target, we leverage the coarse-to-fine scheme to train the task-specific models in a step-aware way, such that the inherent correlation between the neighboring databases could serve as initialization base to relieve the problem of over-fitting, which is inevitable with the small-scaled hand vein database, and also speed up the convergence. Besides, the task-driven network training idea, which involves joint optimization of linear regression classifier and network parameters, is also adopted during training of each model to obtain more discriminative representation for specified tasks. Instead of adopting the trained linear regression classifier for gender and identity classification, the large margin distribution machine (LDM) is introduced to ensure the discriminative and generalization performance of the model simultaneously, and it should be noted that before feeding the gender feature vector into the LDM, a supervised feature selection step is incorporated to improve the classification performance by discarding the redundant feature and highlighting the important ones for gender classification. Rigorous experiments using the lab-made database are conducted to demonstrate the effectiveness and feasibility of the proposed model. What is more, additional experiment with a subset of the PolyU database illustrates its generalization ability and robustness.

The linear array 3D imaging sensor uses a common optical path system to acquire image and point cloud data at the same time. The registration of these two types of data is realized by the common optical axis alignment and time synchronous control system. Since the image and point cloud data have different sampling frequencies; the process of generating colored point cloud needs to interpolate the point cloud. According to the principle of IDW method and the characteristics of the data acquired by the linear array 3D imaging sensor, this article proposes an improved IDW method by selecting the power parameter adaptive to the topography and designing a topographic factor in the weight calculating. The experiment results show that the improved IDW method improves the interpolation accuracy.

This paper proposes an age estimation algorithm in Self-Paced Learning (SPL) framework. Facial samples in the training set inherently include both easy and complex images, which is caused by both the characteristic of age and the variation of pose or expression. Furthermore, by randomly hiding patches in face region, data with different difficulty levels can be gradually used by SPL, in which Convolution Neural Network (CNN) is trained to give the estimation. Alternative Optimization Strategy (AVO), for the weight of CNN and the latent weight in SPL regularizer, is adopted in SPL framework. Experiments show that the proposed algorithm is able to give an accurate results especially under pose and expression variation.

The traditional LDA models to solve the problem of scene classification lack the spatial relationship between the fragments of images or the parts of targets and linkages between the global and local information, so their performance is usually poor in stability for the images with clutter background. In this paper, a novel method for the simultaneous classification and annotation of remote sensing images with complex scenes is proposed. The Spatially Consistent Topic Model is defined by making full use of the correlation between image classification and annotation. We choose SIFT features, hue features and texture features as the visual words, which help to endow pixels of similar appearance region with the same hidden topic. Competitive results on remote sensing images demonstrate the precision and robustness of the proposed method.

In high resolution remote sensing images, urban buildings always have characteristics of complex structures and are vulnerable to background interference. For the purpose of detecting and mapping urban buildings automatically in that circumstance, a novel method is proposed in this paper. Firstly, the Conditional Random Field (CRF) is introduced to fuse multiple kinds of features to get the areas objects existing, then we propose a Hierarchical Object Process Model (HOPM), which is used to access to the location of objects as well as accurate depictions of their outline, and finally the corner detection method is utilized to delineate the vector shapes of objects. Competitive results for multiform and complicated urban buildings demonstrate the precision and robustness of the proposed method.

Deciphering gene regulatory mechanisms through the analysis of high-throughput expression data is a challenging computational problem. Previous computational studies have used large expression datasets in order to resolve fine patterns of coexpression, producing clusters or modules of potentially coregulated genes. These methods typically examine promoter sequence information, such as DNA motifs or transcription factor occupancy data, in a separate step after clustering. We needed an alternative and more integrative approach to study the oxygen regulatory network in Saccharomyces cerevisiae using a small dataset of perturbation experiments. Mechanisms of oxygen sensing and regulation underlie many physiological and pathological processes, and only a handful of oxygen regulators have been identified in previous studies. We used a new machine learning algorithm called MEDUSA to uncover detailed information about the oxygen regulatory network using genome-wide expression changes in response to perturbations in the levels of oxygen, heme, Hap1, and Co2+. MEDUSA integrates mRNA expression, promoter sequence, and ChIP-chip occupancy data to learn a model that accurately predicts the differential expression of target genes in held-out data. We used a novel margin-based score to extract significant condition-specific regulators and assemble a global map of the oxygen sensing and regulatory network. This network includes both known oxygen and heme regulators, such as Hap1, Mga2, Hap4, and Upc2, as well as many new candidate regulators. MEDUSA also identified many DNA motifs that are consistent with previous experimentally identified transcription factor binding sites. Because MEDUSA's regulatory program associates regulators to target genes through their promoter sequences, we directly tested the predicted regulators for OLE1, a gene specifically induced under hypoxia, by experimental analysis of the activity of its promoter. In each case, deletion of the candidate regulator resulted in the predicted effect on promoter activity, confirming that several novel regulators identified by MEDUSA are indeed involved in oxygen regulation. MEDUSA can reveal important information from a small dataset and generate testable hypotheses for further experimental analysis. Supplemental data are included.

A neural network method was developed with the Moderate-resolution Imaging Spectroradiometer (MODIS) land surface temperature product as training and validation datasets, and it was used to retrieve land surface temperatures (LSTs) from direct-broadcast MODIS data in Northeast China in April and May 2003 before fire events. The result shows that LST increases as the day gets closer to the fire day, and this trend can be observed about three days before the fire day. This is similar to the result of fire potential index, so the LST can also be used to evaluate forest fire risk.