Adv Health Sci Educ Theory Pract. 2024 Sep 9. doi: 10.1007/s10459-024-10369-5. Online ahead of print.

ABSTRACT

Health professional education stands to gain substantially from collective efforts toward building video databases of skill performances in both real and simulated settings. An accessible resource of videos that demonstrate an array of performances - both good and bad-provides an opportunity for interdisciplinary research collaborations that can advance our understanding of movement that reflects technical expertise, support educational tool development, and facilitate assessment practices. In this paper we raise important ethical and legal considerations when building and sharing health professions education data. Collective data sharing may produce new knowledge and tools to support healthcare professional education. We demonstrate the utility of a data-sharing culture by providing and leveraging a database of cardio-pulmonary resuscitation (CPR) performances that vary in quality. The CPR skills performance database (collected for the purpose of this research, hosted at UK Data Service's ReShare Repository) contains videos from 40 participants recorded from 6 different angles, allowing for 3D reconstruction for movement analysis. The video footage is accompanied by quality ratings from 2 experts, participants' self-reported confidence and frequency of performing CPR, and the demographics of the participants. From this data, we present an Automatic Clinical Assessment tool for Basic Life Support that uses pose estimation to determine the spatial location of the participant's movements during CPR and a deep learning network that assesses the performance quality.

PMID:39249618 | DOI:10.1007/s10459-024-10369-5

J Cancer Res Clin Oncol. 2024 Sep 9;150(9):415. doi: 10.1007/s00432-024-05930-z.

ABSTRACT

PURPOSE: To construct an integrative radiopathomics model for predicting progression-free survival (PFS) in nonmetastatic nasopharyngeal carcinoma (NPC) patients.

METHODS: 357 NPC patients who underwent pretreatment MRI and pathological whole-slide imaging (WSI) were included in this study and randomly divided into two groups: a training set (n = 250) and validation set (n = 107). Radiomic features extracted from MRI were selected using the minimum redundancy maximum relevance and least absolute shrinkage and selection operator methods. The pathomics signature based on WSI was constructed using a deep learning architecture, the Swin Transformer. The radiopathomics model was constructed by incorporating three feature sets: the radiomics signature, pathomics signature, and independent clinical factors. The prognostic efficacy of the model was assessed using the concordance index (C-index). Kaplan-Meier curves for the stratified risk groups were tested by the log-rank test.

RESULTS: The radiopathomics model exhibited superior predictive performance with C-indexes of 0.791 (95% confidence interval [CI]: 0.724-0.871) in the training set and 0.785 (95% CI: 0.716-0.875) in the validation set compared to any single-modality model (radiomics: 0.619, 95% CI: 0.553-0.706; pathomics: 0.732, 95% CI: 0.662-0.802; clinical model: 0.655, 95% CI: 0.581-0.728) (all, P < 0.05). The radiopathomics model effectively stratified patients into high- and low-risk groups in both the training and validation sets (P < 0.001).

CONCLUSION: The developed radiopathomics model demonstrated its reliability in predicting PFS for NPC patients. It effectively stratified individual patients into distinct risk groups, providing valuable insights for prognostic assessment.

PMID:39249584 | DOI:10.1007/s00432-024-05930-z

J Neural Transm (Vienna). 2024 Sep 9. doi: 10.1007/s00702-024-02830-x. Online ahead of print.

ABSTRACT

Neurodegenerative diseases are group of debilitating and progressive disorders that primarily affect the structure and functions of nervous system, leading to gradual loss of neurons and subsequent decline in cognitive, and behavioral activities. The two frequent diseases affecting the world's significant population falling in the above category are Alzheimer's disease (AD) and Parkinson's disease (PD). These disorders substantially impact the quality of life and burden healthcare systems and society. The demographic characteristics, and machine learning approaches have now been employed to diagnose these illnesses; however, they possess accuracy limitations. Therefore, the authors have developed ranking-based ensemble approach based on the weighted strategy of deep learning classifiers. The whole modeling procedure of the proposed approach incorporates three phases. In phase I, preprocessing techniques are applied to clean the noise in datasets to make it standardized according to deep learning models as it significantly impacts their performance. In phase II, five deep learning models are selected for classification and calculation of prediction results. In phase III, a ranking-based ensemble approach is proposed to ensemble the results of the five models after calculating the ranks and weights of them. In addition, the Magnetic Resonance Imaging (MRI) datasets named Alzheimer's Disease Neuroimaging Initiative (ADNI) for AD classification and Parkinson's Progressive Marker Initiative (PPMI) for PD classification are selected to validate the proposed approach. Furthermore, the proposed method achieved the classification accuracy on AD- Cognitive Normals (CN) at 97.89%, AD- Mild Cognitive Impairment (MCI) at 99.33% and CN-MCI at 99.44% and on PD-CN at 99.22%, PD- Scans Without Evidence of Dopaminergic Effect (SWEDD) at 97.56% and CN-SWEDD at 98.22% respectively. Also, the multi-class classification shows the promising accuracy of 97.18% for AD and 97.85% for PD for the proposed framework. The findings of the study show that the proposed deep learning-based ensemble technique is competitive for AD and PD prediction in both multiclass and binary class classification. Furthermore, the proposed approach enhances generalization performance in diagnosing neurodegenerative diseases and performs better than existing approaches.

PMID:39249515 | DOI:10.1007/s00702-024-02830-x

Skeletal Radiol. 2024 Sep 9. doi: 10.1007/s00256-024-04796-z. Online ahead of print.

ABSTRACT

OBJECTIVE: To develop a deep learning algorithm for diagnosing lumbar central canal stenosis (LCCS) using abdominal CT (ACT) and lumbar spine CT (LCT).

MATERIALS AND METHODS: This retrospective study involved 109 patients undergoing LCTs and ACTs between January 2014 and July 2021. The dural sac on CT images was manually segmented and classified as normal or stenosed (dural sac cross-sectional area ≥ 100 mm2 or < 100 mm2, respectively). A deep learning model based on U-Net architecture was developed to automatically segment the dural sac and classify the central canal stenosis. The classification performance of the model was compared on a testing set (990 images from 9 patients). The accuracy, sensitivity, and specificity of automatic segmentation were quantitatively evaluated by comparing its Dice similarity coefficient (DSC) and intraclass correlation coefficient (ICC) with those of manual segmentation.

RESULTS: In total, 990 CT images from nine patients (mean age ± standard deviation, 77 ± 7 years; six men) were evaluated. The algorithm achieved high segmentation performance with a DSC of 0.85 ± 0.10 and ICC of 0.82 (95% confidence interval [CI]: 0.80,0.85). The ICC between ACTs and LCTs on the deep learning algorithm was 0.89 (95%CI: 0.87,0.91). The accuracy of the algorithm in diagnosing LCCS with dichotomous classification was 84%(95%CI: 0.82,0.86). In dataset analysis, the accuracy of ACTs and LCTs was 85%(95%CI: 0.82,0.88) and 83%(95%CI: 0.79,0.86), respectively. The model showed better accuracy for ACT than LCT.

CONCLUSION: The deep learning algorithm automatically diagnosed LCCS on LCTs and ACTs. ACT had a diagnostic performance for LCCS comparable to that of LCT.

PMID:39249505 | DOI:10.1007/s00256-024-04796-z

J Chem Inf Model. 2024 Sep 9. doi: 10.1021/acs.jcim.4c01451. Online ahead of print.

ABSTRACT

A growing number of deep learning (DL) methodologies have recently been developed to design novel compounds and expand the chemical space within virtual libraries. Most of these neural network approaches design molecules to specifically bind a target based on its structural information and/or knowledge of previously identified binders. Fewer attempts have been made to develop approaches for de novo design of virtual libraries, as synthesizability of generated molecules remains a challenge. In this work, we developed a new Monte Carlo Search (MCS) algorithm, DrugSynthMC (Drug Synthesis using Monte Carlo), in conjunction with DL and statistical-based priors to generate thousands of interpretable chemical structures and novel drug-like molecules per second. DrugSynthMC produces drug-like compounds using an atom-based search model that builds molecules as SMILES, character by character. Designed molecules follow Lipinski's "rule of 5″, show a high proportion of highly water-soluble nontoxic predicted-to-be synthesizable compounds, and efficiently expand the chemical space within the libraries, without reliance on training data sets, synthesizability metrics, or enforcing during SMILES generation. Our approach can function with or without an underlying neural network and is thus easily explainable and versatile. This ease in drug-like molecule generation allows for future integration of score functions aimed at different target- or job-oriented goals. Thus, DrugSynthMC is expected to enable the functional assessment of large compound libraries covering an extensive novel chemical space, overcoming the limitations of existing drug collections. The software is available at https://github.com/RoucairolMilo/DrugSynthMC.

PMID:39249497 | DOI:10.1021/acs.jcim.4c01451

ACS Sens. 2024 Sep 9. doi: 10.1021/acssensors.4c01584. Online ahead of print.

ABSTRACT

This study introduces a novel deep learning framework for lung health evaluation using exhaled gas. The framework synergistically integrates pyramid pooling and a dual-encoder network, leveraging SHapley Additive exPlanations (SHAP) derived feature importance to enhance its predictive capability. The framework is specifically designed to effectively distinguish between smokers, individuals with chronic obstructive pulmonary disease (COPD), and control subjects. The pyramid pooling structure aggregates multilevel global information by pooling features at four scales. SHAP assesses feature importance from the eight sensors. Two encoder architectures handle different feature sets based on their importance, optimizing performance. Besides, the model's robustness is enhanced using the sliding window technique and white noise augmentation on the original data. In 5-fold cross-validation, the model achieved an average accuracy of 96.40%, surpassing that of a single encoder pyramid pooling model by 10.77%. Further optimization of filters in the transformer convolutional layer and pooling size in the pyramid module increased the accuracy to 98.46%. This study offers an efficient tool for identifying the effects of smoking and COPD, as well as a novel approach to utilizing deep learning technology to address complex biomedical issues.

PMID:39248698 | DOI:10.1021/acssensors.4c01584

Adv Sci (Weinh). 2024 Sep 9:e2401379. doi: 10.1002/advs.202401379. Online ahead of print.

ABSTRACT

Focusing on a specific conversation amidst multiple interfering talkers is challenging, especially for those with hearing loss. Brain-controlled assistive hearing devices aim to alleviate this problem by enhancing the attended speech based on the listener's neural signals using auditory attention decoding (AAD). Departing from conventional AAD studies that relied on oversimplified scenarios with stationary talkers, a realistic AAD task that involves multiple talkers taking turns as they continuously move in space in background noise is presented. Invasive electroencephalography (iEEG) data are collected from three neurosurgical patients as they focused on one of the two moving conversations. An enhanced brain-controlled assistive hearing system that combines AAD and a binaural speaker-independent speech separation model is presented. The separation model unmixes talkers while preserving their spatial location and provides talker trajectories to the neural decoder to improve AAD accuracy. Subjective and objective evaluations show that the proposed system enhances speech intelligibility and facilitates conversation tracking while maintaining spatial cues and voice quality in challenging acoustic environments. This research demonstrates the potential of this approach in real-world scenarios and marks a significant step toward developing assistive hearing technologies that adapt to the intricate dynamics of everyday auditory experiences.

PMID:39248654 | DOI:10.1002/advs.202401379

Rev Sci Instrum. 2024 Sep 1;95(9):095107. doi: 10.1063/5.0222123.

ABSTRACT

Ambulatory electrocardiogram (ECG) testing plays a crucial role in the early detection, diagnosis, treatment evaluation, and prevention of cardiovascular diseases. Clear ECG signals are essential for the subsequent analysis of these conditions. However, ECG signals obtained during exercise are susceptible to various noise interferences, including electrode motion artifact, baseline wander, and muscle artifact. These interferences can blur the characteristic ECG waveforms, potentially leading to misjudgment by physicians. To suppress noise in ECG signals more effectively, this paper proposes a novel deep learning-based noise reduction method. This method enhances the diffusion model network by introducing conditional noise, designing a multi-kernel convolutional transformer network structure based on noise prediction, and integrating the diffusion model inverse process to achieve noise reduction. Experiments were conducted on the QT database and MIT-BIH Noise Stress Test Database and compared with the algorithms in other papers to verify the effectiveness of the present method. The results indicate that the proposed method achieves optimal noise reduction performance across both statistical and distance-based evaluation metrics as well as waveform visualization, surpassing eight other state-of-the-art methods. The network proposed in this paper demonstrates stable performance in addressing electrode motion artifact, baseline wander, muscle artifact, and the mixed complex noise of these three types, and it is anticipated to be applied in future noise reduction analysis of clinical dynamic ECG signals.

PMID:39248622 | DOI:10.1063/5.0222123

J Acoust Soc Am. 2024 Sep 1;156(3):1623-1632. doi: 10.1121/10.0028268.

ABSTRACT

Passive acoustic monitoring is a promising tool for monitoring at-risk populations of vocal species, yet, extracting relevant information from large acoustic datasets can be time-consuming, creating a bottleneck at the point of analysis. To address this, an open-source framework for deep learning in bioacoustics to automatically detect Bornean white-bearded gibbon (Hylobates albibarbis) "great call" vocalizations in a long-term acoustic dataset from a rainforest location in Borneo is adapted. The steps involved in developing this solution are described, including collecting audio recordings, developing training and testing datasets, training neural network models, and evaluating model performance. The best model performed at a satisfactory level (F score = 0.87), identifying 98% of the highest-quality calls from 90 h of manually annotated audio recordings and greatly reduced analysis times when compared to a human observer. No significant difference was found in the temporal distribution of great call detections between the manual annotations and the model's output. Future work should seek to apply this model to long-term acoustic datasets to understand spatiotemporal variations in H. albibarbis' calling activity. Overall, a roadmap is presented for applying deep learning to identify the vocalizations of species of interest, which can be adapted for monitoring other endangered vocalizing species.

PMID:39248557 | DOI:10.1121/10.0028268

Int J Surg. 2024 Sep 6. doi: 10.1097/JS9.0000000000002077. Online ahead of print.

ABSTRACT

BACKGROUND: Loss of chromosome 9p is an important biomarker in the malignant transformation of oral leukoplakia (OLK) to head and neck squamous cell carcinoma (HNSCC), and is associated with the prognosis of HNSCC patients. However, various challenges have prevented 9p loss from being assessed in clinical practice. The objective of this study was to develop a pathomics-based artificial intelligence (AI) model for the rapid and cost-effective prediction of 9p loss (9PLP).

MATERIALS AND METHODS: 333 OLK cases were retrospectively collected with hematoxylin and eosin (H&E)-stained whole slide images and genomic alteration data from multicenter cohorts to develop the genomic alteration prediction AI model. They were divided into a training dataset (n=217), a validation dataset (n=93), and an external testing dataset (n=23). The latest Transformer method and XGBoost algorithm were combined to develop the 9PLP model. The AI model was further applied and validated in two multicenter HNSCC datasets (n=42, n=365, respectively). Moreover, the combination of 9PLP with clinicopathological parameters was used to develop a nomogram model for assessing HNSCC patient prognosis.

RESULTS: 9PLP could predict chromosome 9p loss rapidly and effectively using both OLK and HNSCC images, with the area under the curve achieving 0.890 and 0.825, respectively. Furthermore, the predictive model showed high accuracy in HNSCC patient prognosis assessment (the area under the curve was 0.739 for 1-year prediction, 0.705 for 3-year prediction, and 0.691 for 5-year prediction).

CONCLUSION: To the best of our knowledge, this study developed the first genomic alteration prediction deep learning model in OLK and HNSCC. This novel AI model could predict 9p loss and assess patient prognosis by identifying pathomics features in H&E-stained images with good performance. In the future, the 9PLP model may potentially contribute to better clinical management of OLK and HNSCC.

PMID:39248300 | DOI:10.1097/JS9.0000000000002077

Phys Chem Chem Phys. 2024 Sep 9. doi: 10.1039/d4cp02501f. Online ahead of print.

ABSTRACT

A Bayesian optimisation algorithm for deep learning crystal structure prediction software (CBD-GM) is used to predict the structures of Cu(I) and Cu(II) oxides of 2D and 3D materials. Two known 2D structures and two known 3D structures were anticipated, in addition to the prediction of 5 novel structures. All nine structures were optimised and analysed using density-functional theory (DFT). Firstly, DFT calculations using the PBE functional indicate that the structures should be thermodynamically and dynamically stable. Secondly, we calculated the elastic constants using the "stress-strain" method, and the predicted Young's modulus and Poisson's ratios of the materials suggest that they all should have excellent ductile mechanical properties. Calculations of the band structure of the materials performed using the Heyd-Scuseria-Ernzerhof (HSE) hybrid functional indicate that some of the materials should be semiconductors with useful bandgaps. The results therefore provide inspiration for the synthesis of new copper oxides for industrial applications.

PMID:39248038 | DOI:10.1039/d4cp02501f

Front Neuroinform. 2024 Aug 23;18:1451529. doi: 10.3389/fninf.2024.1451529. eCollection 2024.

ABSTRACT

INTRODUCTION: Gait analysis, an expanding research area, employs non-invasive sensors and machine learning techniques for a range of applications. In this study, we investigate the impact of cognitive decline conditions on gait performance, drawing connections between gait deterioration in Parkinson's Disease (PD) and healthy individuals dual tasking.

METHODS: We employ Explainable Artificial Intelligence (XAI) specifically Layer-Wise Relevance Propagation (LRP), in conjunction with Convolutional Neural Networks (CNN) to interpret the intricate patterns in gait dynamics influenced by cognitive loads.

RESULTS: We achieved classification accuracies of 98% F1 scores for PD dataset and 95.5% F1 scores for the combined PD dataset. Furthermore, we explore the significance of cognitive load in healthy gait analysis, resulting in robust classification accuracies of 90% ± 10% F1 scores for subject cognitive load verification. Our findings reveal significant alterations in gait parameters under cognitive decline conditions, highlighting the distinctive patterns associated with PD-related gait impairment and those induced by multitasking in healthy subjects. Through advanced XAI techniques (LRP), we decipher the underlying features contributing to gait changes, providing insights into specific aspects affected by cognitive decline.

DISCUSSION: Our study establishes a novel perspective on gait analysis, demonstrating the applicability of XAI in elucidating the shared characteristics of gait disturbances in PD and dual-task scenarios in healthy individuals. The interpretability offered by XAI enhances our ability to discern subtle variations in gait patterns, contributing to a more nuanced comprehension of the factors influencing gait dynamics in PD and dual-task conditions, emphasizing the role of XAI in unraveling the intricacies of gait control.

PMID:39247901 | PMC:PMC11377268 | DOI:10.3389/fninf.2024.1451529

Front Artif Intell. 2024 Aug 23;7:1258086. doi: 10.3389/frai.2024.1258086. eCollection 2024.

ABSTRACT

Estimating travel time (ETA) for arbitrary paths is crucial in urban intelligent transportation systems. Previous studies primarily focus on constructing complex feature systems for individual road segments or sub-segments, which fail to effectively model the influence of each road segment on others. To address this issue, we propose an end-to-end model, AttentionTTE. It utilizes a self-attention mechanism to capture global spatial correlations and a recurrent neural network to capture temporal dependencies from local spatial correlations. Additionally, a multi-task learning module integrates global spatial correlations and temporal dependencies to estimate the travel time for both the entire path and each local path. We evaluate our model on a large trajectory dataset, and extensive experimental results demonstrate that AttentionTTE achieves state-of-the-art performance compared to other methods.

PMID:39247849 | PMC:PMC11378341 | DOI:10.3389/frai.2024.1258086

Front Artif Intell. 2024 Aug 23;7:1423535. doi: 10.3389/frai.2024.1423535. eCollection 2024.

ABSTRACT

Lung cancer is a predominant cause of cancer-related mortality worldwide, necessitating precise tumor segmentation of medical images for accurate diagnosis and treatment. However, the intrinsic complexity and variability of tumor morphology pose substantial challenges to segmentation tasks. To address this issue, we propose a multitask connected U-Net model with a teacher-student framework to enhance the effectiveness of lung tumor segmentation. The proposed model and framework integrate PET knowledge into the segmentation process, leveraging complementary information from both CT and PET modalities to improve segmentation performance. Additionally, we implemented a tumor area detection method to enhance tumor segmentation performance. In extensive experiments on four datasets, the average Dice coefficient of 0.56, obtained using our model, surpassed those of existing methods such as Segformer (0.51), Transformer (0.50), and UctransNet (0.43). These findings validate the efficacy of the proposed method in lung tumor segmentation tasks.

PMID:39247847 | PMC:PMC11377414 | DOI:10.3389/frai.2024.1423535

IEEE J Transl Eng Health Med. 2024 Aug 22;12:589-599. doi: 10.1109/JTEHM.2024.3448392. eCollection 2024.

ABSTRACT

OBJECTIVE: Optimized deep brain stimulation (DBS) is fast becoming a therapy of choice for the treatment of Parkinson's disease (PD). However, the post-operative optimization (aimed at maximizing patient clinical benefits and minimizing adverse effects) of all possible DBS parameter settings using the standard-of-care clinical protocol requires numerous clinical visits, which substantially increases the time to optimization per patient (TPP), patient cost burden and limit the number of patients who can undergo DBS treatment. The TPP is further elongated in electrodes with stimulation directionality or in diseases with latency in clinical feedback. In this work, we proposed a deep learning and fMRI-based pipeline for DBS optimization that can potentially reduce the TPP from ~1 year to a few hours during a single clinical visit.

METHODS AND PROCEDURES: We developed an unsupervised autoencoder (AE)-based model to extract meaningful features from 122 previously acquired blood oxygenated level dependent (BOLD) fMRI datasets from 39 a priori clinically optimized PD patients undergoing DBS therapy. The extracted features are then fed into multilayer perceptron (MLP)-based parameter classification and prediction models for rapid DBS parameter optimization.

RESULTS: The AE-extracted features of optimal and non-optimal DBS were disentangled. The AE-MLP classification model yielded accuracy, precision, recall, F1 score, and combined AUC of 0.96 ± 0.04, 0.95 ± 0.07, 0.92 ± 0.07, 0.93 ± 0.06, and 0.98 respectively. Accuracies of 0.79 ± 0.04, 0.85 ± 0.04, 0.82 ± 0.05, 0.83 ± 0.05, and 0.70 ± 0.07 were obtained in the prediction of voltage, frequency, and x-y-z contact locations, respectively.

CONCLUSION: The proposed AE-MLP models yielded promising results for fMRI-based DBS parameter classification and prediction, potentially facilitating rapid semi-automated DBS parameter optimization. Clinical and Translational Impact Statement-A deep learning-based pipeline for semi-automated DBS parameter optimization is presented, with the potential to significantly decrease the optimization duration per patient and patients' financial burden while increasing patient throughput.

PMID:39247846 | PMC:PMC11379443 | DOI:10.1109/JTEHM.2024.3448392

IEEE J Transl Eng Health Med. 2024 Aug 23;12:600-612. doi: 10.1109/JTEHM.2024.3448457. eCollection 2024.

ABSTRACT

The integration of electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) can facilitate the advancement of brain-computer interfaces (BCIs). However, existing research in this domain has grappled with the challenge of the efficient selection of features, resulting in the underutilization of the temporal richness of EEG and the spatial specificity of fNIRS data.To effectively address this challenge, this study proposed a deep learning architecture called the multimodal DenseNet fusion (MDNF) model that was trained on two-dimensional (2D) EEG data images, leveraging advanced feature extraction techniques. The model transformed EEG data into 2D images using a short-time Fourier transform, applied transfer learning to extract discriminative features, and consequently integrated them with fNIRS-derived spectral entropy features. This approach aimed to bridge existing gaps in EEG-fNIRS-based BCI research by enhancing classification accuracy and versatility across various cognitive and motor imagery tasks.Experimental results on two public datasets demonstrated the superiority of our model over existing state-of-the-art methods.Thus, the high accuracy and precise feature utilization of the MDNF model demonstrates the potential in clinical applications for neurodiagnostics and rehabilitation, thereby paving the method for patient-specific therapeutic strategies.

PMID:39247844 | PMC:PMC11379445 | DOI:10.1109/JTEHM.2024.3448457

Proc SPIE Int Soc Opt Eng. 2024 Feb;12925:1292516. doi: 10.1117/12.3008622. Epub 2024 Apr 1.

ABSTRACT

Sparse CT reconstruction continues to be an area of interest in a number of novel imaging systems. Many different approaches have been tried including model-based methods, compressed sensing approaches, and most recently deep-learning-based processing. Diffusion models, in particular, have become extremely popular due to their ability to effectively encode rich information about images and to allow for posterior sampling to generate many possible outputs. One drawback of diffusion models is that their recurrent structure tends to be computationally expensive. In this work we apply a new Fourier diffusion approach that permits processing with many fewer time steps than the standard scalar diffusion model. We present an extension of the Fourier diffusion technique and evaluate it in a simulated breast cone-beam CT system with a sparse view acquisition.

PMID:39247536 | PMC:PMC11378968 | DOI:10.1117/12.3008622

Med Image Comput Comput Assist Interv. 2019 Oct;11765:420-429. doi: 10.1007/978-3-030-32245-8_47. Epub 2019 Oct 10.

ABSTRACT

Deep convolutional neural networks (CNNs) are state-of-the-art for semantic image segmentation, but typically require many labeled training samples. Obtaining 3D segmentations of medical images for supervised training is difficult and labor intensive. Motivated by classical approaches for joint segmentation and registration we therefore propose a deep learning framework that jointly learns networks for image registration and image segmentation. In contrast to previous work on deep unsupervised image registration, which showed the benefit of weak supervision via image segmentations, our approach can use existing segmentations when available and computes them via the segmentation network otherwise, thereby providing the same registration benefit. Conversely, segmentation network training benefits from the registration, which essentially provides a realistic form of data augmentation. Experiments on knee and brain 3D magnetic resonance (MR) images show that our approach achieves large simultaneous improvements of segmentation and registration accuracy (over independently trained networks) and allows training high-quality models with very limited training data. Specifically, in a one-shot-scenario (with only one manually labeled image) our approach increases Dice scores (%) over an unsupervised registration network by 2.7 and 1.8 on the knee and brain images respectively.

PMID:39247524 | PMC:PMC11378322 | DOI:10.1007/978-3-030-32245-8_47

Heliyon. 2024 Jul 30;10(16):e35273. doi: 10.1016/j.heliyon.2024.e35273. eCollection 2024 Aug 30.

ABSTRACT

With the widespread application of deep learning technology in various fields, power load forecasting, as an important link in power system operation and planning, has also ushered in new opportunities and challenges. Traditional forecasting methods perform poorly when faced with the high uncertainty and complexity of power loads. In view of this, this paper proposes a power load forecasting model PSO-BiTC based on deep learning and particle swarm optimization. This model combines a temporal convolutional network (TCN) and a bidirectional long short-term memory network (BiLSTM), using TCN to process long sequence data and capture features and patterns in time series, while using BiLSTM to capture long-term and short-term dependencies. In addition, the particle swarm optimization algorithm (PSO) is used to optimize model parameters to improve the model's predictive performance and generalization ability. Experimental results show that the PSO-BiTC model performs well in power load forecasting. Compared with traditional methods, this model reduces the MAE (Mean Absolute Error) to 20.18, 17.57, 18.61 and 16.7 on four extensive data sets, respectively. It has been proven that it achieves the best performance in various indicators, with a low number of parameters and training time. This research is of great significance for improving the operating efficiency of the power system, optimizing resource allocation, and promoting carbon emission reduction goals in the urban building sector.

PMID:39247372 | PMC:PMC11379997 | DOI:10.1016/j.heliyon.2024.e35273

Heliyon. 2024 Aug 6;10(16):e35854. doi: 10.1016/j.heliyon.2024.e35854. eCollection 2024 Aug 30.

ABSTRACT

This paper presents an innovative framework for the automated diagnosis of gastric cancer using artificial intelligence. The proposed approach utilizes a customized deep learning model called MobileNetV2, which is optimized using a Dynamic variant of the Pelican Optimization Algorithm (DPOA). By combining these advanced techniques, it is feasible to achieve highly accurate results when applied to a dataset of endoscopic gastric images. To evaluate the performance of the model based on the benchmark, its data is divided into training (80 %) and testing (20 %) sets. The MobileNetV2/DPOA model demonstrated an impressive accuracy of 97.73 %, precision of 97.88 %, specificity of 97.72 %, sensitivity of 96.35 %, Matthews Correlation Coefficient (MCC) of 96.58 %, and F1-score of 98.41 %. These results surpassed those obtained by other well-known models, such as Convolutional Neural Networks (CNN), Mask Region-Based Convolutional Neural Networks (Mask R-CNN), U-Net, Deep Stacked Sparse Autoencoder Neural Networks (SANNs), and DeepLab v3+, in terms of most quantitative metrics. Despite the promising outcomes, it is important to note that further research is needed. Specifically, larger and more diverse datasets as well as exhaustive clinical validation are necessary to validate the effectiveness of the proposed method. By implementing this innovative approach in the detection of gastric cancer, it is possible to enhance the speed and accuracy of diagnosis, leading to improved patient care and better allocation of healthcare resources.

PMID:39247334 | PMC:PMC11380007 | DOI:10.1016/j.heliyon.2024.e35854