Search Results (730)

This paper describes a multistage framework for face image analysis in computer-aided speech diagnosis and therapy. Multimodal data processing frameworks have become a significant factor in supporting speech disorders’ treatment. Synchronous and asynchronous remote speech therapy approaches can use audio and video analysis of articulation to deliver robust indicators of disordered speech. Accurate segmentation of articulators in video frames is a vital step in this agenda. We use a dedicated data acquisition system to capture the stereovision stream during speech therapy examination in children. Our goal is to detect and accurately segment four objects in the mouth area (lips, teeth, tongue, and whole mouth) during relaxed speech and speech therapy exercises. Our database contains 17,913 frames from 76 preschool children. We apply a sequence of procedures employing artificial intelligence. For detection, we train the YOLOv6 (you only look once) model to catch each of the three objects under consideration. Then, we prepare the DeepLab v3+ segmentation model in a semi-supervised training mode. As preparation of reliable expert annotations is exhausting in video labeling, we first train the network using weak labels produced by initial segmentation based on the distance-regularized level set evolution over fuzzified images. Next, we fine-tune the model using a portion of manual ground-truth delineations. Each stage is thoroughly assessed using the independent test subset. The lips are detected almost perfectly (average precision and F1 score of 0.999), whereas the segmentation Dice index exceeds 0.83 in each articulator, with a top result of 0.95 in the whole mouth. Full article

Introduction: Cervical cancer is the fourth most common cancer in women globally. It is the most common cancer in Tanzania, resulting in about 9772 new cases and 6695 deaths each year. Research has shown an association between low levels of risk perception and knowledge of the prevention, risks, signs, etiology, and treatment of cervical cancer and low screening uptake, as contributing to high rates of cervical cancer-related mortality. However, there is scant literature on the perspectives of a wider group of stakeholders (e.g., policymakers, healthcare providers (HCPs), and women at risk), especially those living in rural and semi-rural settings. The main objective of this study is to understand knowledge and perspectives on cervical cancer risk and screening among these populations. Methods: We adapted Risso-Gill and colleagues’ framework for a Health Systems Appraisal (HSA), to identify HCPs’ perspective of the extent to which health system requirements for effective cervical cancer screening, prevention, and control are in place in Tanzania. We adapted interview topic guides for cervical cancer screening using the HSA framework approach. Study participants (69 in total) were interviewed between 2014 and 2018—participants included key stakeholders, HCPs, and women at risk for cervical cancer. The data were analyzed using reflexive thematic analysis methodology. Results: Seven themes emerged from our analysis of semi-structured interviews and focus groups: (1) knowledge of the role of screening and preventive care/services (e.g., prevention, risks, signs, etiology, and treatment), (2) training and knowledge of HCPs, (3) knowledge of cervical cancer screening among women at risk, (4) beliefs about cervical cancer screening, (5) role of traditional medicine, (6) risk factors, and (7) symptoms and signs. Conclusions: Our results demonstrate that there is a low level of knowledge of the role of screening and preventive services among stakeholders, HCPs, and women living in rural and semi-rural locations in Tanzania. There is a critical need to implement more initiatives and programs to increase the uptake of screening and related services and allow women to make more informed decisions on their health. Full article

Pilot behavior is crucial for aviation safety. This study aims to investigate the EEG characteristics of pilots, refine training assessment methodologies, and bolster flight safety measures. The collected EEG signals underwent initial preprocessing. The EEG characteristic analysis was performed during left and right turns, involving the calculation of the energy ratio of beta waves and Shannon entropy. The psychological workload of pilots during different flight phases was quantified as well. Based on the EEG characteristics, the pilots’ psychological workload was classified through the use of a support vector machine (SVM). The study results showed significant changes in the energy ratio of beta waves and Shannon entropy during left and right turns compared to the cruising phase. Additionally, the pilots’ psychological workload was found to have increased during these turning phases. Using support vector machines to detect the pilots’ psychological workload, the classification accuracy for the training set was 98.92%, while for the test set, it was 93.67%. This research holds significant importance in understanding pilots’ psychological workload. Full article

Classic Hodgkin’s lymphoma (cHL) is a curable cancer with a disease-free survival rate of over 10 years. Over 80% of diagnosed patients respond favorably to first-line chemotherapy, but few biomarkers exist that can predict the 15–20% of patients who experience refractory or early relapsed disease. To date, the identification of patients who will not respond to first-line therapy based on disease staging and traditional clinical risk factor analysis is still not possible. Three-dimensional (3D) telomere analysis using the TeloView^® software platform has been shown to be a reliable tool to quantify genomic instability and to inform on disease progression and patients’ response to therapy in several cancers. It also demonstrated telomere dysfunction in cHL elucidating biological mechanisms related to disease progression. Here, we report 3D telomere analysis on a multicenter cohort of 156 cHL patients. We used the cohort data as a training data set and identified significant 3D telomere parameters suitable to predict individual patient outcomes at the point of diagnosis. Multivariate analysis using logistic regression procedures allowed for developing a predictive scoring model using four 3D telomere parameters as predictors, including the proportion of t-stumps (very short telomeres), which has been a prominent predictor for cHL patient outcome in a previously published study using TeloView^® analysis. The percentage of t-stumps was by far the most prominent predictor to identify refractory/relapsing (RR) cHL prior to initiation of adriamycin, bleomycin, vinblastine, and dacarbazine (ABVD) therapy. The model characteristics include an AUC of 0.83 in ROC analysis and a sensitivity and specificity of 0.82 and 0.78 respectively. Full article

Introduction: Convolutional Neural Network (CNN) systems in healthcare are influenced by unbalanced datasets and varying sizes. This article delves into the impact of dataset size, class imbalance, and their interplay on CNN systems, focusing on the size of the training set versus imbalance—a unique perspective compared to the prevailing literature. Furthermore, it addresses scenarios with more than two classification groups, often overlooked but prevalent in practical settings. Methods: Initially, a CNN was developed to classify lung diseases using X-ray images, distinguishing between healthy individuals and COVID-19 patients. Later, the model was expanded to include pneumonia patients. To evaluate performance, numerous experiments were conducted with varied data sizes and imbalance ratios for both binary and ternary classifications, measuring various indices to validate the model’s efficacy. Results: The study revealed that increasing dataset size positively impacts CNN performance, but this improvement saturates beyond a certain size. A novel finding is that the data balance ratio influences performance more significantly than dataset size. The behavior of three-class classification mirrored that of binary classification, underscoring the importance of balanced datasets for accurate classification. Conclusions: This study emphasizes the fact that achieving balanced representation in datasets is crucial for optimal CNN performance in healthcare, challenging the conventional focus on dataset size. Balanced datasets improve classification accuracy, both in two-class and three-class scenarios, highlighting the need for data-balancing techniques to improve model reliability and effectiveness. Motivation: Our study is motivated by a scenario with 100 patient samples, offering two options: a balanced dataset with 200 samples and an unbalanced dataset with 500 samples (400 healthy individuals). We aim to provide insights into the optimal choice based on the interplay between dataset size and imbalance, enriching the discourse for stakeholders interested in achieving optimal model performance. Limitations: Recognizing a single model’s generalizability limitations, we assert that further studies on diverse datasets are needed. Full article

This study presents the development and validation of a robust semi-supervised learning framework specifically designed for the automated segmentation and classification of sandstone thin section images from the Yanchang Formation in the Ordos Basin. Traditional geological image analysis methods encounter significant challenges due to the labor-intensive and error-prone nature of manual labeling, compounded by the diversity and complexity of rock thin sections. Our approach addresses these challenges by integrating the GL-SLIC algorithm, which combines Gabor filters and Local Binary Patterns for effective superpixel segmentation, laying the groundwork for advanced component identification. The primary innovation of this research is the semi-supervised learning model that utilizes a limited set of manually labeled samples to generate high-confidence pseudo labels, thereby significantly expanding the training dataset. This methodology effectively tackles the critical challenge of insufficient labeled data in geological image analysis, enhancing the model’s generalization capability from minimal initial input. Our framework improves segmentation accuracy by closely aligning superpixels with the intricate boundaries of mineral grains and pores. Additionally, it achieves substantial improvements in classification accuracy across various rock types, reaching up to 96.3% in testing scenarios. This semi-supervised approach represents a significant advancement in computational geology, providing a scalable and efficient solution for detailed petrographic analysis. It not only enhances the accuracy and efficiency of geological interpretations but also supports broader hydrocarbon exploration efforts. Full article

Human–Machine Interfaces (HMIs) have gained popularity as they allow for an effortless and natural interaction between the user and the machine by processing information gathered from a single or multiple sensing modalities and transcribing user intentions to the desired actions. Their operability depends on frequent periodic re-calibration using newly acquired data due to their adaptation needs in dynamic environments, where test–time data continuously change in unforeseen ways, a cause that significantly contributes to their abandonment and remains unexplored by the Ultrasound-based (US-based) HMI community. In this work, we conduct a thorough investigation of Unsupervised Domain Adaptation (UDA) algorithms for the re-calibration of US-based HMIs during within-day sessions, which utilize unlabeled data for re-calibration. Our experimentation led us to the proposal of a CNN-based architecture for simultaneous wrist rotation angle and finger gesture prediction that achieves comparable performance with the state-of-the-art while featuring $87.92\%$ less trainable parameters. According to our findings, DANN (a Domain-Adversarial training algorithm), with proper initialization, offers an average $24.99\%$ classification accuracy performance enhancement when compared to no re-calibration setting. However, our results suggest that in cases where the experimental setup and the UDA configuration may differ, observed enhancements would be rather small or even unnoticeable. Full article

The purpose of the article is to provide a practical guide for manual and semi-automated image segmentation of common neurosurgical cranial lesions, namely meningioma, glioblastoma multiforme (GBM) and subarachnoid haemorrhage (SAH), for neurosurgical trainees and researchers. Materials and Methods: The medical images used were sourced from the Medical Image Computing and Computer Assisted Interventions Society (MICCAI) Multimodal Brain Tumour Segmentation Challenge (BRATS) image database and from the local Picture Archival and Communication System (PACS) record with consent. Image pre-processing was carried out using MRIcron software (v1.0.20190902). ITK-SNAP (v3.8.0) was used in this guideline due to its availability and powerful built-in segmentation tools, although others (Seg3D, Freesurfer and 3D Slicer) are available. Quality control was achieved by employing expert segmenters to review. Results: A pipeline was developed to demonstrate the pre-processing and manual and semi-automated segmentation of patient images for each cranial lesion, accompanied by image guidance and video recordings. Three sample segmentations were generated to illustrate potential challenges. Advice and solutions were provided within both text and video. Conclusions: Semi-automated segmentation methods enhance efficiency, increase reproducibility, and are suitable to be incorporated into future clinical practise. However, manual segmentation remains a highly effective technique in specific circumstances and provides initial training sets for the development of more advanced semi- and fully automated segmentation algorithms. Full article

HIV point of care testing (POCT) is a common approach to expanding testing into non-specialised settings. Dental services have untapped potential to screen for health conditions including HIV. However, the perspectives of UK dental patients, dental professionals, and people with HIV are unknown. Ten focus groups were undertaken with dental patients, professionals, and people with HIV. The Framework method was used to analyse the qualitative data. Six themes were generated from the focus group data. The themes explored perceptions of HIV, the purpose, appropriateness, and acceptability of HIV testing in dental settings, and new processes that would need to be established in order to successfully implement point of care HIV testing in UK dental settings. Training needs were identified including communication skills and updates to current knowledge about HIV. HIV testing in dental settings is generally acceptable to dental patients, dental professionals, and PWH. However, of concern were logistical challenges and the risk of patients surprised at being offered an HIV test during a visit to the dentist. Nonetheless, the public health benefits of the intervention were well understood, i.e., early detection of HIV and initiation of treatment to improve health outcomes. Dental teams were able to generate novel solutions that could help to overcome contextual and logistical challenges to implementing HIV testing in dental settings. Full article

In this paper, we introduce WindFormer, an innovative transformer-based model engineered for short-term wind speed forecasting, leveraging multivariate time series data. Unlike traditional approaches, WindFormer excels in processing diverse meteorological features—temperature, humidity, and power—to intricately map their spatiotemporal interdependencies with wind speeds. Utilizing a novel unsupervised pre-training strategy, WindFormer initially learns from vast volumes of unlabeled data to capture generalized feature representations. This foundation enhances the subsequent fine-tuning phase on labeled wind speed data, in which our model demonstrates exceptional predictive accuracy. Empirical evaluations across various public datasets illustrate that WindFormer markedly surpasses both conventional statistical models and contemporary deep learning techniques. The model not only achieves superior accuracy in forecasting wind speeds but also reveals a significant enhancement in handling complex spatiotemporal data dynamics. These advancements facilitate more effective wind farm management and power grid scheduling, making a substantial impact on operational efficiencies and renewable energy utilization. Our findings confirm the robustness of WindFormer in a real-world setting, underscoring its potential as a pivotal tool in meteorological and energy sectors. The integration of unsupervised pre-training with multi-task fine-tuning establishes a new benchmark for short-term wind speed prediction. Full article

A machine learning (ML) model, optimized by the Particle Swarm Optimization (PSO) algorithm, was developed in this study to predict the shear slip load of adhesive/bolt-reinforced corroded steel plates. An extensive database comprising 490 experimental or numerical specimens was initially employed to train the ML models. Eight ML algorithms (RF, AdaBoost, XGBoost, GBT, SVR, kNN, LightGBM, and CatBoost) were utilized for shear slip load prediction, with their hyperparameters set to default values. Subsequently, the PSO algorithm was employed to optimize the hyperparameters of the above ML algorithms. Finally, performance metrics, error analysis, and score analysis were employed to evaluate the prediction capabilities of the optimized ML models, identifying PSO-GBT as the optimal predictive model. A user-friendly graphical user interface (GUI) was also developed to facilitate engineers using the PSO-GBT model developed in this study to predict the shear slip load of adhesive/bolt-reinforced corroded steel plates. Full article

Objective: Trauma is a leading cause of death worldwide, with many incidents resulting in hemorrhage before the patient reaches the hospital. Despite advances in trauma care, the majority of deaths occur within the first three hours of hospital admission, offering a very limited window for effective intervention. Unfortunately, a significant increase in mortality from hemorrhagic trauma is primarily due to delays in hemorrhage control. Therefore, we propose a machine learning model to predict the need for urgent hemorrhage intervention. Methods: This study developed and validated an XGBoost-based machine learning model using data from the National Trauma Data Bank (NTDB) from 2017 to 2019. It focuses on demographic and clinical data from the initial hours following trauma for model training and validation, aiming to predict whether trauma patients require urgent hemorrhage intervention. Results: The XGBoost model demonstrated superior performance across multiple datasets, achieving an AUROC of 0.872 on the training set, 0.869 on the internal validation set, and 0.875 on the external validation set. The model also showed high sensitivity (77.8% on the external validation set) and specificity (82.1% on the external validation set), with an accuracy exceeding 81% across all datasets, highlighting its high reliability for clinical applications. Conclusions: Our study shows that the XGBoost model effectively predicts urgent hemorrhage interventions using data from the National Trauma Data Bank (NTDB). It outperforms other machine learning algorithms in accuracy and robustness across various datasets. These results highlight machine learning’s potential to improve emergency responses and decision-making in trauma care. Full article

Regular detection of land-use and land-cover (LULC) changes with high accuracy is necessary for natural resources management and sustainable urban planning. The produced LULC maps from Google Earth Engine (GEE) also illustrate the transformation of the LULC for the respective landscape over time. The selected study area, Cottbus City and the Spree-Neisse district in northeastern Germany, has undergone significant development over the past decades due to various factors, including urbanization and industrialization; also, the landscape has been converted in some areas for post-mining activities. Detection of LULC changes that have taken place over the last few decades thus plays a vital role in quantifying the impact of these factors while improving the knowledge of these developments and supporting the city planners or urban management officials before implementing further long-term development initiatives for the future. Therefore, the study aims to (i) detect LULC changes for the time slices 2002 and 2022, testing machine learning (ML) algorithms in supervised and unsupervised classification for Landsat satellite imageries, and (ii) validate the newly produced LULC maps with the available regional database (RDB) from the federal and state statistical offices, Germany, and the Dynamic World (DW) near real-time 10 m global LULC data set powered by artificial intelligence (AI). The results of the Random Forest (RF) and the Smilecart classifiers of supervised classification using Landsat 9 OLI-2/TIRS-2 in 2022 demonstrated a validation accuracy of 88% for both, with Kappa Index (KI) of 83% and 84%, respectively. Moreover, the Training Overall Accuracy (TOA) was 100% for both years. The wekaKMeans cluster of the unsupervised classification also illustrated a similar transformation pattern in the LULC maps. Overall, the produced LULC maps offered an improved representation of the selected region’s various land-cover classes (i.e., vegetation, waterbodies, built areas, and bare ground) in the last two decades (20022 to 2022). Full article

Within the dynamic realm of variable-geometry shock wave/boundary layer interactions, the wall parameters of the flow field undergo real-time fluctuations. The conventional approach to sensing these changes in wall pressure through sensor measurements is encumbered by a cumbersome process, leading to diminished efficiency and an inability to provide swift predictions of wall parameters. This paper introduces a data-driven methodology that leverages non-contact schlieren imaging to predict wall pressure within the flow field, a technique that holds promise for informing the optimized design of variable-geometry systems. A sophisticated deep learning framework, predicated on Convolutional Neural Networks (CNN), has been engineered to anticipate alterations in wall pressure stemming from high-speed shock wave/boundary layer interactions. Utilizing an impulsive wind tunnel with a Mach number of 6, we have procured a sequence of schlieren images and corresponding wall pressure measurements, capturing the continuous variations induced by an attack angle from a shock wave generator. These data have been instrumental in compiling a comprehensive dataset for the training and evaluation of the CNN. The CNN model, once trained, has adeptly deduced the distribution of wall pressure from the schlieren imagery. Notwithstanding, it was observed that the CNN’s predictive prowess is marginally diminished in regions where pressure variations are most pronounced. To assess the model’s generalization capabilities, we have segmented the dataset according to different temporal intervals for network training. Our findings indicate that while the generalization of all models crafted was less than optimal, Model 4 demonstrated superior generalization. It is thus suggested that augmenting the training set with additional samples and refining the network architecture will be a worthwhile endeavor in subsequent research initiatives. Full article

Kidney segmentation from abdominal computed tomography (CT) images is essential for computer-aided kidney diagnosis, pathology detection, and surgical planning. This paper introduces a kidney segmentation method for clinical contrast-enhanced CT images. First, it begins with shape-based preprocessing to remove the spine and ribs. Second, a novel clustering algorithm and an initial kidney selection strategy are utilized to locate the initial slices and contours. Finally, an adaptive narrow-band approach based on active contours is developed, followed by a clustering postprocessing to address issues with concave parts. Experimental results demonstrate the high segmentation performance of the proposed method, achieving a Dice Similarity Coefficient of 97.4 ± 1.0% and an Average Symmetric Surface Distance of 0.5 ± 0.2 mm across twenty sequences. Notably, this method eliminates the need for manually setting initial contours and can handle intensity inhomogeneity and varying kidney shapes without extensive training or statistical modeling. Full article