IEEE Trans Image Process. 2024 Mar 22;PP. doi: 10.1109/TIP.2024.3378473. Online ahead of print.

ABSTRACT

The goal of moving object segmentation is separating moving objects from stationary backgrounds in videos. One major challenge in this problem is how to develop a universal model for videos from various natural scenes since previous methods are often effective only in specific scenes. In this paper, we propose a method called Learning Temporal Distribution and Spatial Correlation (LTS) that has the potential to be a general solution for universal moving object segmentation. In the proposed approach, the distribution from temporal pixels is first learned by our Defect Iterative Distribution Learning (DIDL) network for a scene-independent segmentation. Notably, the DIDL network incorporates the use of an improved product distribution layer that we have newly derived. Then, the Stochastic Bayesian Refinement (SBR) Network, which learns the spatial correlation, is proposed to improve the binary mask generated by the DIDL network. Benefiting from the scene independence of the temporal distribution and the accuracy improvement resulting from the spatial correlation, the proposed approach performs well for almost all videos from diverse and complex natural scenes with fixed parameters. Comprehensive experiments on standard datasets including LASIESTA, CDNet2014, BMC, SBMI2015 and 128 real world videos demonstrate the superiority of proposed approach compared to state-of-the-art methods with or without the use of deep learning networks. To the best of our knowledge, this work has high potential to be a general solution for moving object segmentation in real world environments. The code and real-world videos can be found on GitHub https://github.com/guanfangdong/LTS-UniverisalMOS.

PMID:38517718 | DOI:10.1109/TIP.2024.3378473

IEEE Trans Image Process. 2024 Mar 22;PP. doi: 10.1109/TIP.2024.3378469. Online ahead of print.

ABSTRACT

Accurate 6-DoF pose estimation of surgical instruments during minimally invasive surgeries can substantially improve treatment strategies and eventual surgical outcome. Existing deep learning methods have achieved accurate results, but they require custom approaches for each object and laborious setup and training environments often stretching to extensive simulations, whilst lacking real-time computation. We propose a general-purpose approach of data acquisition for 6-DoF pose estimation tasks in X-ray systems, a novel and general purpose YOLOv5-6D pose architecture for accurate and fast object pose estimation and a complete method for surgical screw pose estimation under acquisition geometry consideration from a monocular cone-beam X-ray image. The proposed YOLOv5-6D pose model achieves competitive results on public benchmarks whilst being considerably faster at 42 FPS on GPU. In addition, the method generalizes across varying X-ray acquisition geometry and semantic image complexity to enable accurate pose estimation over different domains. Finally, the proposed approach is tested for bone-screw pose estimation for computer-aided guidance during spine surgeries. The model achieves a 92.41% by the 0.1∙d ADD-S metric, demonstrating a promising approach for enhancing surgical precision and patient outcomes. The code for YOLOv5-6D is publicly available at https://github.com/cviviers/YOLOv5-6D-Pose.

PMID:38517715 | DOI:10.1109/TIP.2024.3378469

Insights Imaging. 2024 Mar 22;15(1):81. doi: 10.1186/s13244-024-01657-0.

ABSTRACT

BACKGROUND: Cerebrovascular diseases have emerged as significant threats to human life and health. Effectively segmenting brain blood vessels has become a crucial scientific challenge. We aimed to develop a fully automated deep learning workflow that achieves accurate 3D segmentation of cerebral blood vessels by incorporating classic convolutional neural networks (CNNs) and transformer models.

METHODS: We used a public cerebrovascular segmentation dataset (CSD) containing 45 volumes of 1.5 T time-of-flight magnetic resonance angiography images. We collected data from another private middle cerebral artery (MCA) with lenticulostriate artery (LSA) segmentation dataset (MLD), which encompassed 3.0 T three-dimensional T1-weighted sequences of volumetric isotropic turbo spin echo acquisition MRI images of 107 patients aged 62 ± 11 years (42 females). The workflow includes data analysis, preprocessing, augmentation, model training with validation, and postprocessing techniques. Brain vessels were segmented using the U-Net, V-Net, UNETR, and SwinUNETR models. The model performances were evaluated using the dice similarity coefficient (DSC), average surface distance (ASD), precision (PRE), sensitivity (SEN), and specificity (SPE).

RESULTS: During 4-fold cross-validation, SwinUNETR obtained the highest DSC in each fold. On the CSD test set, SwinUNETR achieved the best DSC (0.853), PRE (0.848), SEN (0.860), and SPE (0.9996), while V-Net achieved the best ASD (0.99). On the MLD test set, SwinUNETR demonstrated good MCA segmentation performance and had the best DSC, ASD, PRE, and SPE for segmenting the LSA.

CONCLUSIONS: The workflow demonstrated excellent performance on different sequences of MRI images for vessels of varying sizes. This method allows doctors to visualize cerebrovascular structures.

CRITICAL RELEVANCE STATEMENT: A deep learning-based 3D cerebrovascular segmentation workflow is feasible and promising for visualizing cerebrovascular structures and monitoring cerebral small vessels, such as lenticulostriate arteries.

KEY POINTS: • The proposed deep learning-based workflow performs well in cerebrovascular segmentation tasks. • Among comparison models, SwinUNETR achieved the best DSC, ASD, PRE, and SPE values in lenticulostriate artery segmentation. • The proposed workflow can be used for different MR sequences, such as bright and black blood imaging.

PMID:38517610 | DOI:10.1186/s13244-024-01657-0

J Med Eng Technol. 2024 Mar 22:288-297. doi: 10.1080/03091902.2024.2327459. Online ahead of print.

ABSTRACT

Visual inspection is the typical way for evaluating burns, due to the rising occurrence of burns globally, visual inspection may not be sufficient to detect skin burns because the severity of burns can vary and some burns may not be immediately apparent to the naked eye. Burns can have catastrophic and incapacitating effects and if they are not treated on time can cause scarring, organ failure, and even death. Burns are a prominent cause of considerable morbidity, but for a variety of reasons, traditional clinical approaches may struggle to effectively predict the severity of burn wounds at an early stage. Since computer-aided diagnosis is growing in popularity, our proposed study tackles the gap in artificial intelligence research, where machine learning has received a lot of attention but transfer learning has received less attention. In this paper, we describe a method that makes use of transfer learning to improve the performance of ML models, showcasing its usefulness in diverse applications. The transfer learning approach estimates the severity of skin burn damage using the image data of skin burns and uses the results to improve future methods. The DL technique consists of a basic CNN and seven distinct transfer learning model types. The photos are separated into those displaying first, second, and third-degree burns as well as those showing healthy skin using a fully connected feed-forward neural network. The results demonstrate that the accuracy of 93.87% for the basic CNN model which is significantly lower, with the VGG-16 model achieving the greatest accuracy at 97.43% and being followed by the DenseNet121 model at 96.66%. The proposed approach based on CNN and transfer learning techniques are tested on datasets from Kaggle 2022 and Maharashtra Institute of Technology open-school medical repository datasets that are clubbed together. The suggested CNN-based approach can assist healthcare professionals in promptly and precisely assessing burn damage, resulting in appropriate therapies and greatly minimising the detrimental effects of burn injuries.

PMID:38517037 | DOI:10.1080/03091902.2024.2327459

J Chem Inf Model. 2024 Mar 22. doi: 10.1021/acs.jcim.3c02058. Online ahead of print.

ABSTRACT

Molecular property prediction is a fundamental task of drug discovery. With the rapid development of deep learning, computational approaches for predicting molecular properties are experiencing increasing popularity. However, these existing methods often ignore the 3D information on molecules, which is critical in molecular representation learning. In the past few years, several self-supervised learning (SSL) approaches have been proposed to exploit the geometric information by using pre-training on 3D molecular graphs and fine-tuning on 2D molecular graphs. Most of these approaches are based on the global geometry of molecules, and there is still a challenge in capturing the local structure and local interpretability. To this end, we propose local geometry-guided graph attention (LGGA), which integrates local geometry into the attention mechanism and message-passing of graph neural networks (GNNs). LGGA introduces a novel method to model molecules, enhancing the model's ability to capture intricate local structural details. Experiments on various data sets demonstrate that the integration of local geometry has a significant impact on the improved results, and our model outperforms the state-of-the-art methods for molecular property prediction, establishing its potential as a promising tool in drug discovery and related fields.

PMID:38516950 | DOI:10.1021/acs.jcim.3c02058

Diagn Cytopathol. 2024 Mar 22. doi: 10.1002/dc.25295. Online ahead of print.

ABSTRACT

OBJECTIVE: Cervical cancer, a prevalent and deadly disease among women, comes second only to breast cancer, with over 700 daily deaths. The Pap smear test is a widely utilized screening method for detecting cervical cancer in its early stages. However, this manual screening process is prone to a high rate of false-positive outcomes because of human errors. Researchers are using machine learning and deep learning in computer-aided diagnostic tools to address this issue. These tools automatically analyze and sort cervical cytology and colposcopy images, improving the precision of identifying various stages of cervical cancer.

METHODOLOGY: This article uses state-of-the-art deep learning methods, such as ResNet-50 for categorizing cervical cancer cells to assist medical professionals. The method includes three key steps: preprocessing, segmentation using k-means clustering, and classifying cancer cells. The model is assessed based on performance metrics viz; precision, accuracy, kappa score, precision, sensitivity, and specificity. In the end, the high success rate shows that the ResNet50 model is a valuable tool for timely detection of cervical cancer.

OUTPUTS: In conclusion, the infected cervical region is pinpointed using spatial K-means clustering and preprocessing operations. This sequence of actions is followed by a progressive learning technique. The Progressive Learning technique then proceeded through several stages: Stage 1 with 64 × 64 images, Stage 2 with 224 × 224 images, Stage 3 with 512 × 512 images, and the final Stage 4 with 1024 × 1024 images. The outcomes show that the suggested model is effective for analyzing Pap smear tests, achieving 97.4% accuracy and approx. 98% kappa score.

PMID:38516853 | DOI:10.1002/dc.25295

Front Plant Sci. 2024 Mar 7;15:1290845. doi: 10.3389/fpls.2024.1290845. eCollection 2024.

ABSTRACT

Rodents are essential to the balance of the grassland ecosystem, but their population outbreak can cause major economic and ecological damage. Rodent monitoring is crucial for its scientific management, but traditional methods heavily depend on manual labor and are difficult to be carried out on a large scale. In this study, we used UAS to collect high-resolution RGB images of steppes in Inner Mongolia, China in the spring, and used various object detection algorithms to identify the holes of Brandt's vole (Lasiopodomys brandtii). Optimizing the model by adjusting evaluation metrics, specifically, replacing classification strategy metrics such as precision, recall, and F1 score with regression strategy-related metrics FPPI, MR, and MAPE to determine the optimal threshold parameters for IOU and confidence. Then, we mapped the distribution of vole holes in the study area using position data derived from the optimized model. Results showed that the best resolution of UAS acquisition was 0.4 cm pixel-1, and the improved labeling method improved the detection accuracy of the model. The FCOS model had the highest comprehensive evaluation, and an R2 of 0.9106, RMSE of 5.5909, and MAPE of 8.27%. The final accuracy of vole hole counting in the stitched orthophoto was 90.20%. Our work has demonstrated that UAS was able to accurately estimate the population of grassland rodents at an appropriate resolution. Given that the population distribution we focus on is important for a wide variety of species, our work illustrates a general remote sensing approach for mapping and monitoring rodent damage across broad landscapes for studies of grassland ecological balance, vegetation conservation, and land management.

PMID:38516671 | PMC:PMC10955068 | DOI:10.3389/fpls.2024.1290845

Biophys Rep. 2023 Aug 31;9(4):177-187. doi: 10.52601/bpr.2023.230014.

ABSTRACT

DNA-based point accumulation in nanoscale topography (DNA-PAINT) is a well-established technique for single-molecule localization microscopy (SMLM), enabling resolution of up to a few nanometers. Traditionally, DNA-PAINT involves the utilization of tens of thousands of single-molecule fluorescent images to generate a single super-resolution image. This process can be time-consuming, which makes it unfeasible for many researchers. Here, we propose a simplified DNA-PAINT labeling method and a deep learning-enabled fast DNA-PAINT imaging strategy for subcellular structures, such as microtubules. By employing our method, super-resolution reconstruction can be achieved with only one-tenth of the raw data previously needed, along with the option of acquiring the widefield image. As a result, DNA-PAINT imaging is significantly accelerated, making it more accessible to a wider range of biological researchers.

PMID:38516619 | PMC:PMC10951475 | DOI:10.52601/bpr.2023.230014

Otol Neurotol Open. 2023 Mar 9;3(1):e028. doi: 10.1097/ONO.0000000000000028. eCollection 2023 Mar.

ABSTRACT

OBJECTIVE: To find a more objective method of assessing which patients should be screened for a vestibular schwannoma (VS) with magnetic resonance imaging (MRI) using a deep-learning algorithm to assess clinical and audiometric data.

MATERIALS AND METHODS: Clinical and audiometric data were collected for 592 patients who received an audiogram between January 2015 and 2020 at Duke University Health Center with and without VS confirmed by MRI. These data were analyzed using a deep learning-based analysis to determine if the need for MRI screening could be assessed more objectively with adequate sensitivity and specificity.

RESULTS: Patients with VS showed slightly elevated, but not statistically significant, mean thresholds compared to those without. Tinnitus, gradual hearing loss, and aural fullness were more common in patients with VS. Of these, only the presence of tinnitus was statistically significant. Several machine learning algorithms were used to incorporate and model the collected clinical and audiometric data, but none were able to distinguish ears with and without confirmed VS. When tumor size was taken into account the analysis was still unable to distinguish a difference.

CONCLUSIONS: Using audiometric and clinical data, deep learning-based analyses failed to produce an adequately sensitive and specific model for the detection of patients with VS. This suggests that a specific pattern of audiometric asymmetry and clinical symptoms may not necessarily be predictive of the presence/absence of VS to a level that clinicians would be comfortable forgoing an MRI.

PMID:38516318 | PMC:PMC10950172 | DOI:10.1097/ONO.0000000000000028

Photoacoustics. 2024 Mar 11;37:100601. doi: 10.1016/j.pacs.2024.100601. eCollection 2024 Jun.

ABSTRACT

Photoacoustic tomography (PAT) is a promising imaging technique that can visualize the distribution of chromophores within biological tissue. However, the accuracy of PAT imaging is compromised by light fluence (LF), which hinders the quantification of light absorbers. Currently, model-based iterative methods are used for LF correction, but they require extensive computational resources due to repeated LF estimation based on differential light transport models. To improve LF correction efficiency, we propose to use Fourier neural operator (FNO), a neural network specially designed for estimating partial differential equations, to learn the forward projection of light transport in PAT. Trained using paired finite-element-based LF simulation data, our FNO model replaces the traditional computational heavy LF estimator during iterative correction, such that the correction procedure is considerably accelerated. Simulation and experimental results demonstrate that our method achieves comparable LF correction quality to traditional iterative methods while reducing the correction time by over 30 times.

PMID:38516295 | PMC:PMC10955667 | DOI:10.1016/j.pacs.2024.100601

Photoacoustics. 2024 Mar 11;37:100600. doi: 10.1016/j.pacs.2024.100600. eCollection 2024 Jun.

ABSTRACT

The unique advantage of optical-resolution photoacoustic microscopy (OR-PAM) is its ability to achieve high-resolution microvascular imaging without exogenous agents. This ability has excellent potential in the study of tissue microcirculation. However, tracing and monitoring microvascular morphology and hemodynamics in tissues is challenging because the segmentation of microvascular in OR-PAM images is complex due to the high density, structure complexity, and low contrast of vascular structures. Various microvasculature extraction techniques have been developed over the years but have many limitations: they cannot consider both thick and thin blood vessel segmentation simultaneously, they cannot address incompleteness and discontinuity in microvasculature, there is a lack of open-access datasets for DL-based algorithms. We have developed a novel segmentation approach to extract vascularity in OR-PAM images using a deep learning network incorporating a weak signal attention mechanism and multi-scale perception (WSA-MP-Net) model. The proposed WSA network focuses on weak and tiny vessels, while the MP module extracts features from different vessel sizes. In addition, Hessian-matrix enhancement is incorporated into the pre-and post-processing of the input and output data of the network to enhance vessel continuity. We constructed normal vessel (NV-ORPAM, 660 data pairs) and tumor vessel (TV-ORPAM, 1168 data pairs) datasets to verify the performance of the proposed method. We developed a semi-automatic annotation algorithm to obtain the ground truth for our network optimization. We applied our optimized model successfully to monitor glioma angiogenesis in mouse brains, thus demonstrating the feasibility and excellent generalization ability of our model. Compared to previous works, our proposed WSA-MP-Net extracts a significant number of microvascular while maintaining vessel continuity and signal fidelity. In quantitative analysis, the indicator values of our method improved by about 1.3% to 25.9%. We believe our proposed approach provides a promising way to extract a complete and continuous microvascular network of OR-PAM and enables its use in many microvascular-related biological studies and medical diagnoses.

PMID:38516294 | PMC:PMC10955652 | DOI:10.1016/j.pacs.2024.100600

Data Brief. 2024 Mar 2;54:110264. doi: 10.1016/j.dib.2024.110264. eCollection 2024 Jun.

ABSTRACT

The Intelligent Transportation System (ITS) seeks to improve traffic flow to guarantee transportation safety. One of the ITS's fundamental tenets is identifying and classifying vehicles into various classes. Although the issues related to small size, variety of forms, and similarity in visual appearance of the vehicles, as well as the influence of the weather on the video and image quality, make it challenging to categorize vehicles using unmanned aerial vehicles (UAV); they are becoming more popular in computer vision-related applications. Traffic accidents are now a serious public health concern that must be addressed in the Kurdistan Region of Iraq. An automatic vehicle detection and classification system can be considered one of the remedies to solve this issue. This paper presents a dataset of 2,160 images of vehicles on the roads in the Iraqi Kurdistan Region to address the issue of the absence of such a dataset. The images in the proposed collection were taken with a Mavic Air 2 drone in the Iraqi cities of Sulaymaniyah and Erbil. The images are categorized into five classes: bus, truck, taxi, personal car, and motorcycle. Data gathering considered diverse circumstances, multiple vehicle sizes, weather and lighting conditions, and massive camera movements. Pre-processing and data augmentation methods were applied to the images in our proposed dataset, including auto-orient, brightness, hue, and noise algorithm, which can be used to build an efficient deep learning (DL) model. After applying these augmentation techniques for the car, taxi, truck, motorcycle, and bus classes, the number of images was increased to 5,353, 1,500, 1,192, 282, and 176, respectively.

PMID:38516279 | PMC:PMC10950728 | DOI:10.1016/j.dib.2024.110264

Plant Pathol. 2023 Apr;72(3):536-547. doi: 10.1111/ppa.13684. Epub 2023 Jan 10.

ABSTRACT

Crop diseases can cause major yield losses, so the ability to detect and identify them in their early stages is important for disease control. Deep learning methods have shown promise in classifying multiple diseases; however, many studies do not use datasets that represent real field conditions, necessitating either further image processing or reducing their applicability. In this paper, we present a dataset of wheat images taken in real growth situations, including both field and glasshouse conditions, with five categories: healthy plants and four foliar diseases, yellow rust, brown rust, powdery mildew and Septoria leaf blotch. This dataset was used to train a deep learning model. The resulting model, named CerealConv, reached a 97.05% classification accuracy. When tested against trained pathologists on a subset of images from the larger dataset, the model delivered an accuracy score 2% higher than the best-performing pathologist. Image masks were used to show that the model was using the correct information to drive its classifications. These results show that deep learning networks are a viable tool for disease detection and classification in the field, and disease quantification is a logical next step.

PMID:38516179 | PMC:PMC10953319 | DOI:10.1111/ppa.13684

Ther Adv Ophthalmol. 2024 Mar 20;16:25158414241232258. doi: 10.1177/25158414241232258. eCollection 2024 Jan-Dec.

ABSTRACT

BACKGROUND: New developments in artificial intelligence, particularly with promising results in early detection and management of keratoconus, have favorably altered the natural history of the disease over the last few decades. Features of artificial intelligence in different machine such as anterior segment optical coherence tomography, and femtosecond laser technique have improved safety, precision, effectiveness, and predictability of treatment modalities of keratoconus (from contact lenses to keratoplasty techniques). These options ingrained in artificial intelligence are already underway and allow ophthalmologist to approach disease in the most non-invasive way.

OBJECTIVES: This study comprehensively describes all of the treatment modalities of keratoconus considering machine learning strategies.

DESIGN: A multidimensional comprehensive systematic narrative review.

DATA SOURCES AND METHODS: A comprehensive search was done in the five main electronic databases (PubMed, Scopus, Web of Science, Embase, and Cochrane), without language and time or type of study restrictions. Afterward, eligible articles were selected by screening the titles and abstracts based on main mesh keywords. For potentially eligible articles, the full text was also reviewed.

RESULTS: Artificial intelligence demonstrates promise in keratoconus diagnosis and clinical management, spanning early detection (especially in subclinical cases), preoperative screening, postoperative ectasia prediction after keratorefractive surgery, and guiding surgical decisions. The majority of studies employed a solitary machine learning algorithm, whereas minor studies assessed multiple algorithms that evaluated the association of various keratoconus staging and management strategies. Last but not least, AI has proven effective in guiding the implantation of intracorneal ring segments in keratoconus corneas and predicting surgical outcomes.

CONCLUSION: The efficient and widespread clinical translation of machine learning models in keratoconus management is a crucial goal of potential future approaches to better visual performance in keratoconus patients.

TRIAL REGISTRATION: The article has been registered through PROSPERO, an international database of prospectively registered systematic reviews, with the ID: CRD42022319338.

PMID:38516169 | PMC:PMC10956165 | DOI:10.1177/25158414241232258

IEEE Int Conf Healthc Inform. 2023 Jun;2023:49-57. doi: 10.1109/ichi57859.2023.00018. Epub 2023 Dec 11.

ABSTRACT

Alzheimer's disease (AD) is one of the leading causes of death in the United States, especially among the elderly. Recent studies have shown how hypertension is related to cognitive decline in elderly patients, which in turn leads to increased mortality as well as morbidity. There have been various studies that have looked at the effect of antihypertensive drugs in reducing cognitive decline, and their results have proved inconclusive. However, most of these studies assume the treatment effect is similar for all patients, thus considering only the average treatment effects of antihypertensive drugs. In this paper, we assume that the effect of antihypertensives on the onset of AD depends on patient characteristics. We develop a deep learning method called LASSO-Dragonnet to estimate the individualized treatment effects of each patient. We considered six antihypertensive drugs, and each of the six models considered one of the drugs as the treatment and the remaining as control. Our studies showed that although many antihypertensives have a positive impact in delaying AD onset on average, the impact varies from individual to individual, depending on their various characteristics. We also analyzed the importance of various covariates in such an estimation. Our results showed that the individualized treatment effects of each patient could be estimated accurately using a deep learning method, and that the importance of various covariates could be determined.

PMID:38516035 | PMC:PMC10956734 | DOI:10.1109/ichi57859.2023.00018

Bioinformatics. 2024 Mar 21:btae160. doi: 10.1093/bioinformatics/btae160. Online ahead of print.

ABSTRACT

MOTIVATION: Deep learning algorithms applied to structural biology often struggle to converge to meaningful solutions when limited data is available, since they are required to learn complex physical rules from examples. State-of-the-art force-fields, however, cannot interface with deep learning algorithms due to their implementation.

RESULTS: We present MadraX, a forcefield implemented as a differentiable PyTorch module, able to interact with deep learning algorithms in an end-to-end fashion.

AVAILABILITY AND IMPLEMENTATION: MadraX documentation, together with tutorials and installation guide, is available at madrax.readthedocs.io.

PMID:38514422 | DOI:10.1093/bioinformatics/btae160

Epidemiol Health. 2024 Mar 13:e2024039. doi: 10.4178/epih.e2024039. Online ahead of print.

ABSTRACT

OBJECTIVES: To achieve the ambitious goal of eliminating schistosome infections, the Chinese government has implemented diverse control strategies. This study explored the progress of the 2 most recent national schistosomiasis control programs in an endemic area along the Yangtze River in China.

METHODS: We obtained village-level parasitological data from cross-sectional surveys combined with environmental data in Anhui Province, China from 1997 to 2015. A convolutional neural network (CNN) based on a hierarchical integro-difference equation (IDE) framework (i.e., CNN-IDE) was used to model spatio-temporal variations in schistosomiasis. Two traditional models were also constructed for comparison with 2 evaluation indicators: the mean-squared prediction error (MSPE) and continuous ranked probability score (CRPS).

RESULTS: The CNN-IDE model was the optimal model, with the lowest overall average MSPE of 0.04 and the CRPS of 0.19. From 1997 to 2011, the prevalence exhibited a notable trend: it increased steadily until peaking at 1.6 per 1000 in 2005, then gradually declined, stabilizing at a lower rate of approximately 0.6 per 1000 in 2006, and approaching zero by 2011. During this period, noticeable geographic disparities in schistosomiasis prevalence were observed; high-risk areas were initially dispersed, followed by contraction. Predictions for the period 2012 to 2015 demonstrated a consistent and uniform decrease.

CONCLUSION: The proposed CNN-IDE model captured the intricate and evolving dynamics of schistosomiasis prevalence, offering a promising alternative for future risk modeling of the disease. The comprehensive strategy is expected to help diminish schistosomiasis infection, emphasizing the necessity to continue implementing this strategy.

PMID:38514196 | DOI:10.4178/epih.e2024039

J Neurointerv Surg. 2024 Mar 21:jnis-2024-021536. doi: 10.1136/jnis-2024-021536. Online ahead of print.

ABSTRACT

BACKGROUND: Early detection of large vessel occlusion (LVO) facilitates triage to an appropriate stroke center to reduce treatment times and improve outcomes. Prehospital stroke scales are not sufficiently sensitive, so we investigated the ability of the portable Openwater optical blood flow monitor to detect LVO.

METHODS: Patients were prospectively enrolled at two comprehensive stroke centers during stroke alert evaluation within 24 hours of onset with National Institutes of Health Stroke Scale (NIHSS) score ≥2. A 70 s bedside optical blood flow scan generated cerebral blood flow waveforms based on relative changes in speckle contrast. Anterior circulation LVO was determined by CT angiography. A deep learning model trained on all patient data using fivefold cross-validation and learned discriminative representations from the raw speckle contrast waveform data. Receiver operating characteristic (ROC) analysis compared the Openwater diagnostic performance (ie, LVO detection) with prehospital stroke scales.

RESULTS: Among 135 patients, 52 (39%) had an anterior circulation LVO. The median NIHSS score was 8 (IQR 4-14). The Openwater instrument had 79% sensitivity and 84% specificity for the detection of LVO. The rapid arterial occlusion evaluation (RACE) scale had 60% sensitivity and 81% specificity and the Los Angeles motor scale (LAMS) had 50% sensitivity and 81% specificity. The binary Openwater classification (high-likelihood vs low-likelihood) had an area under the ROC (AUROC) of 0.82 (95% CI 0.75 to 0.88), which outperformed RACE (AUC 0.70; 95% CI 0.62 to 0.78; P=0.04) and LAMS (AUC 0.65; 95% CI 0.57 to 0.73; P=0.002).

CONCLUSIONS: The Openwater optical blood flow monitor outperformed prehospital stroke scales for the detection of LVO in patients undergoing acute stroke evaluation in the emergency department. These encouraging findings need to be validated in an independent test set and the prehospital environment.

PMID:38514189 | DOI:10.1136/jnis-2024-021536

BMJ Open. 2024 Mar 21;14(3):e079311. doi: 10.1136/bmjopen-2023-079311.

ABSTRACT

BACKGROUND: Cardiovascular disease is a leading cause of global death. Prospective population-based studies have found that changes in retinal microvasculature are associated with the development of coronary artery disease. Recently, artificial intelligence deep learning (DL) algorithms have been developed for the fully automated assessment of retinal vessel calibres.

METHODS: In this study, we validate the association between retinal vessel calibres measured by a DL system (Singapore I Vessel Assessment) and incident myocardial infarction (MI) and assess its incremental performance in discriminating patients with and without MI when added to risk prediction models, using a large UK Biobank cohort.

RESULTS: Retinal arteriolar narrowing was significantly associated with incident MI in both the age, gender and fellow calibre-adjusted (HR=1.67 (95% CI: 1.19 to 2.36)) and multivariable models (HR=1.64 (95% CI: 1.16 to 2.32)) adjusted for age, gender and other cardiovascular risk factors such as blood pressure, diabetes mellitus (DM) and cholesterol status. The area under the receiver operating characteristic curve increased from 0.738 to 0.745 (p=0.018) in the age-gender-adjusted model and from 0.782 to 0.787 (p=0.010) in the multivariable model. The continuous net reclassification improvements (NRIs) were significant in the age and gender-adjusted (NRI=21.56 (95% CI: 3.33 to 33.42)) and the multivariable models (NRI=18.35 (95% CI: 6.27 to 32.61)). In the subgroup analysis, similar associations between retinal arteriolar narrowing and incident MI were observed, particularly for men (HR=1.62 (95% CI: 1.07 to 2.46)), non-smokers (HR=1.65 (95% CI: 1.13 to 2.42)), patients without DM (HR=1.73 (95% CI: 1.19 to 2.51)) and hypertensive patients (HR=1.95 (95% CI: 1.30 to 2.93)) in the multivariable models.

CONCLUSION: Our results support DL-based retinal vessel measurements as markers of incident MI in a predominantly Caucasian population.

PMID:38514140 | DOI:10.1136/bmjopen-2023-079311

Brain Res. 2024 Mar 19:148876. doi: 10.1016/j.brainres.2024.148876. Online ahead of print.

ABSTRACT

Schizophrenia is a profound and enduring mental disorder that imposes significant negative impacts on individuals, their families, and society at large. The development of more accurate and objective diagnostic tools for schizophrenia can be expedited through the employment of deep learning (DL), that excels at deciphering complex hierarchical non-linear patterns. However, the limited interpretability of deep learning has eroded confidence in the model and restricted its clinical utility. At the same time, if the data source is only derived from a single center, the model's generalizability is difficult to test. To enhance the model's reliability and applicability, leave-one-center-out validation with a large and diverse sample from multiple centers is crucial. In this study, we utilized Nine different global centers to train and test the 3D Resnet model's generalizability, resulting in an 82% classification performance (area under the curve) on all datasets sourced from different countries, employing a leave-one-center-out-validation approach. Per our approximation of the feature significance of each region on the atlas, we identified marked differences in the thalamus, pallidum, and inferior frontal gyrus between individuals with schizophrenia and healthy controls, lending credence to prior research findings. At the same time, in order to translate the model's output into clinically applicable insights, the SHapley Additive exPlanations (SHAP) permutation explainer method with an anatomical atlas have been refined, thereby offering precise neuroanatomical and functional interpretations of different brain regions.

PMID:38513996 | DOI:10.1016/j.brainres.2024.148876