Oman Med J. 2024 May 30;39(3):e632. doi: 10.5001/omj.2024.79. eCollection 2024 May.

ABSTRACT

OBJECTIVES: Chronic liver disease and cirrhosis are persistent global health threats, ranking among the top causes of death. Despite medical advancements, their mortality rates have remained stagnant for decades. Existing scoring systems such as Child-Turcotte-Pugh and Mayo End-Stage Liver Disease have limitations, prompting the exploration of more accurate predictive methods using artificial intelligence and machine learning (ML).

METHODS: We retrospectively reviewed the data of all adult patients with acute decompensated liver cirrhosis admitted to a tertiary hospital during 2015-2021. The dataset underwent preprocessing to handle missing values and standardize continuous features. Traditional ML and deep learning algorithms were applied to build a 28-day mortality prediction model.

RESULTS: The subjects were 173 cirrhosis patients, whose medical records were examined. We developed and evaluated multiple models for 28-day mortality prediction. Among traditional ML algorithms, logistic regression outperformed was achieving an accuracy of 82.9%, precision of 55.6%, recall of 71.4%, and an F1-score of 0.625. Naive Bayes and Random Forest models also performed well, both achieving the same accuracy (82.9%) and precision (54.5%). The deep learning models (multilayer artificial neural network, recurrent neural network, and Long Short-Term Memory) exhibited mixed results, with the multilayer artificial neural network achieving an accuracy of 74.3% but lower precision and recall. The feature importance analysis identified key predictability contributors, including admission in the intensive care unit (importance: 0.112), use of mechanical ventilation (importance: 0.095), and mean arterial pressure (importance: 0.073).

CONCLUSIONS: Our study demonstrates the potential of ML in predicting 28-day mortality following hospitalization with acute decompensation of liver cirrhosis. Logistic Regression, Naive Bayes, and Random Forest models proved effective, while deep learning models exhibited variable performance. These models can serve as useful tools for risk stratification and timely intervention. Implementing these models in clinical practice has the potential to improve patient outcomes and resource allocation.

PMID:39497942 | PMC:PMC11532584 | DOI:10.5001/omj.2024.79

J Cheminform. 2024 Nov 4;16(1):121. doi: 10.1186/s13321-024-00912-2.

ABSTRACT

We introduce an advanced model for predicting protein-ligand interactions. Our approach combines the strengths of graph neural networks with physics-based scoring methods. Existing structure-based machine-learning models for protein-ligand binding prediction often fall short in practical virtual screening scenarios, hindered by the intricacies of binding poses, the chemical diversity of drug-like molecules, and the scarcity of crystallographic data for protein-ligand complexes. To overcome the limitations of existing machine learning-based prediction models, we propose a novel approach that fuses three independent neural network models. One classification model is designed to perform binary prediction of a given protein-ligand complex pose. The other two regression models are trained to predict the binding affinity and root-mean-square deviation of a ligand conformation from an input complex structure. We trained the model to account for both deviations in experimental and predicted binding affinities and pose prediction uncertainties. By effectively integrating the outputs of the triplet neural networks with a physics-based scoring function, our model showed a significantly improved performance in hit identification. The benchmark results with three independent decoy sets demonstrate that our model outperformed existing models in forward screening. Our model achieved top 1% enrichment factors of 32.7 and 23.1 with the CASF2016 and DUD-E benchmark sets, respectively. The benchmark results using the LIT-PCBA set further confirmed its higher average enrichment factors, emphasizing the model's efficiency and generalizability. The model's efficiency was further validated by identifying 23 active compounds from 63 candidates in experimental screening for autotaxin inhibitors, demonstrating its practical applicability in hit discovery.Scientific contributionOur work introduces a novel training strategy for a protein-ligand binding affinity prediction model by integrating the outputs of three independent sub-models and utilizing expertly crafted decoy sets. The model showcases exceptional performance across multiple benchmarks. The high enrichment factors in the LIT-PCBA benchmark demonstrate its potential to accelerate hit discovery.

PMID:39497201 | DOI:10.1186/s13321-024-00912-2

BMC Med Imaging. 2024 Nov 5;24(1):298. doi: 10.1186/s12880-024-01443-w.

ABSTRACT

OBJECTIVE: To conduct a systematic review of the computer vision applications that detect, diagnose, or analyse tuberculosis (TB) pathology or bacilli using digitised human lung tissue images either through automatic or semi-automatic methods. We categorised the computer vision platform into four technologies: image processing, object/pattern recognition, computer graphics, and deep learning. In this paper, the focus is on image processing and deep learning (DL) applications for either 2D or 3D digitised human lung tissue images. This review is useful for establishing a common practice in TB analysis using human lung tissue as well as identifying opportunities for further research in this space. The review brings attention to the state-of-art techniques for detecting TB, with emphasis on the challenges and limitations of the current techniques. The ultimate goal is to promote the development of more efficient and accurate algorithms for the detection or analysis of TB, and raise awareness about the importance of early detection.

DESIGN: We searched five databases and Google Scholar for articles published between January 2017 and December 2022 that focus on Mycobacterium tuberculosis detection, or tuberculosis pathology using digitised human lung tissue images. Details regarding design, image processing and computer-aided techniques, deep learning models, and datasets were collected and summarised. Discussions, analysis, and comparisons of state-of-the-art methods are provided to help guide future research. Further, a brief update on the relevant techniques and their performance is provided.

RESULTS: Several studies have been conducted to develop automated and AI-assisted methods for diagnosing Mtb and TB pathology from digitised human lung tissue images. Some studies presented a completely automated method of diagnosis, while other studies developed AI-assisted diagnostic methods. Low-level focus areas included the development of a novel μ CT scanner for soft tissue image contract, and use of multiresolution computed tomography to analyse the 3D structure of the human lung. High-level focus areas included the investigation the effects of aging on the number and size of small airways in the lungs using CT and whole lung high-resolution μ CT, and the 3D microanatomy characterisation of human tuberculosis lung using μ CT in conjunction with histology and immunohistochemistry. Additionally, a novel method for acquiring high-resolution 3D images of human lung structure and topology is also presented.

CONCLUSION: The literature indicates that post 1950s, TB was predominantly studied using animal models even though no animal model reflects the full spectrum of human pulmonary TB disease and does not reproducibly transmit Mtb infection to other animals (Hunter, 2011). This explains why there are very few studies that used human lung tissue for detection or analysis of Mtb. Nonetheless, we found 10 studies that used human tissues (predominately lung) of which five studies proposed machine learning (ML) models for the detection of bacilli and the other five used CT on human lung tissue scanned ex-vivo.

PMID:39497049 | DOI:10.1186/s12880-024-01443-w

Sci Rep. 2024 Nov 4;14(1):26666. doi: 10.1038/s41598-024-78393-4.

ABSTRACT

Visual scoring of interstitial lung disease in systemic sclerosis (SSc-ILD) from CT scans is laborious, subjective and time-consuming. This study aims to develop a deep learning framework to automate SSc-ILD scoring. The automated framework is a cascade of two neural networks. The first network selects the craniocaudal positions of the five scoring levels. Subsequently, for each level, the second network estimates the ratio of three patterns to the total lung area: the total extent of disease (TOT), ground glass (GG) and reticulation (RET). To overcome the score imbalance in the second network, we propose a method to augment the training dataset with synthetic data. To explain the network's output, a heat map method is introduced to highlight the candidate interstitial lung disease regions. The explainability of heat maps was evaluated by two human experts and a quantitative method that uses the heat map to produce the score. The results show that our framework achieved a κ of 0.66, 0.58, and 0.65, for the TOT, GG and RET scoring, respectively. Both experts agreed with the heat maps in 91%, 90% and 80% of cases, respectively. Therefore, it is feasible to develop a framework for automated SSc-ILD scoring, which performs competitively with human experts and provides high-quality explanations using heat maps. Confirming the model's generalizability is needed in future studies.

PMID:39496802 | DOI:10.1038/s41598-024-78393-4

Sci Rep. 2024 Nov 5;14(1):26729. doi: 10.1038/s41598-024-77550-z.

ABSTRACT

Early and precise diagnosis of craniosynostosis (CSO), which involves premature fusion of cranial sutures in infants, is crucial for effective treatment. Although computed topography offers detailed imaging, its high radiation poses risks, especially to children. Therefore, we propose a deep-learning model for CSO and suture-line classification using 2D cranial X-rays that minimises radiation-exposure risks and offers reliable diagnoses. We used data comprising 1,047 normal and 277 CSO cases from 2006 to 2023. Our approach integrates X-ray-marker removal, head-pose standardisation, skull-cropping, and fine-tuning modules for CSO and suture-line classification using convolution neural networks (CNNs). It enhances the diagnostic accuracy and efficiency of identifying CSO from X-ray images, offering a promising alternative to traditional methods. Four CNN backbones exhibited robust performance, with F1-scores exceeding 0.96 and sensitivity and specificity exceeding 0.9, proving the potential for clinical applications. Additionally, preprocessing strategies further enhanced the accuracy, demonstrating the highest F1-scores, precision, and specificity. A qualitative analysis using gradient-weighted class activation mapping illustrated the focal points of the models. Furthermore, the suture-line classification model distinguishes five suture lines with an accuracy of > 0.9. Thus, the proposed approach can significantly reduce the time and labour required for CSO diagnosis, streamlining its management in clinical settings.

PMID:39496759 | DOI:10.1038/s41598-024-77550-z

Sci Rep. 2024 Nov 4;14(1):26705. doi: 10.1038/s41598-024-77494-4.

ABSTRACT

Clinical proteomics analysis is of great significance for analyzing pathological mechanisms and discovering disease-related biomarkers. Using computational methods to accurately predict disease types can effectively improve patient disease diagnosis and prognosis. However, how to eliminate the errors introduced by peptide precursor identification and protein identification for pathological diagnosis remains a major unresolved issue. Here, we develop a powerful end-to-end deep learning model, termed "MS1Former", that is able to classify hepatocellular carcinoma tumors and adjacent non-tumor (normal) tissues directly using raw MS1 spectra without peptide precursor identification. Our model provides accurate discrimination of subtle m/z differences in MS1 between tumor and adjacent non-tumor tissue, as well as more general performance predictions for data-dependent acquisition, data-independent acquisition, and full-scan data. Our model achieves the best performance on multiple external validation datasets. Additionally, we perform a detailed exploration of the model's interpretability. Prospectively, we expect that the advanced end-to-end framework will be more applicable to the classification of other tumors.

PMID:39496730 | DOI:10.1038/s41598-024-77494-4

Sci Rep. 2024 Nov 4;14(1):26668. doi: 10.1038/s41598-024-77829-1.

ABSTRACT

Early detection of Alzheimer's disease (AD) is crucial to maximize clinical outcomes. Most disease progression analyses include people with diagnoses of cognitive impairment, limiting understanding of AD risk among those with normal cognition. The objective was to establish AD progression models through a deep learning approach to analyze heterogeneous, multi-modal datasets, including clustering analyses of population subsets. A multi-head deep-learning architecture was built to process and learn from biomedical and imaging data from the National Alzheimer's Coordinating Center. Shapley additive explanation algorithms for feature importance ranking and pairwise correlation analysis were used to identify predictors of disease progression. Four primary disease progression clusters (slow, moderate and rapid converters or non-converters) were subdivided into groups by race and sex, yielding 16 sub-clusters of participants with distinct progression patterns. A multi-head and early-fusion convolutional neural network achieved the most competitive performance and demonstrated superiority over a single-head deep learning architecture and conventional tree-based machine-learning methods, with 97% test accuracy, 96% F1 score and 0.19 root mean square error. From 447 features, 2 sets of 100 predictors of disease progression were extracted. Feature importance ranking, correlation analysis and descriptive statistics further enriched cluster analysis and validation of the heterogeneity of risk factors.

PMID:39496718 | DOI:10.1038/s41598-024-77829-1

Sci Rep. 2024 Nov 4;14(1):26702. doi: 10.1038/s41598-024-77878-6.

ABSTRACT

The WCAY (weighted channel attention YOLO) model, which is meticulously crafted to identify fracture features across diverse X-ray image sites, is presented herein. This model integrates novel core operators and an innovative attention mechanism to enhance its efficacy. Initially, leveraging the benefits of dynamic snake convolution (DSConv), which is adept at capturing elongated tubular structural features, we introduce the DSC-C2f module to augment the model's fracture detection performance by replacing a portion of C2f. Subsequently, we integrate the newly proposed weighted channel attention (WCA) mechanism into the architecture to bolster feature fusion and improve fracture detection across various sites. Comparative experiments were conducted, to evaluate the performances of several attention mechanisms. These enhancement strategies were validated through experimentation on public X-ray image datasets (FracAtlas and GRAZPEDWRI-DX). Multiple experimental comparisons substantiated the model's efficacy, demonstrating its superior accuracy and real-time detection capabilities. According to the experimental findings, on the FracAtlas dataset, our WCAY model exhibits a notable 8.8% improvement in mean average precision (mAP) over the original model. On the GRAZPEDWRI-DX dataset, the mAP reaches 64.4%, with a detection accuracy of 93.9% for the "fracture" category alone. The proposed model represents a substantial improvement over the original algorithm compared to other state-of-the-art object detection models. The code is publicly available at https://github.com/cccp421/Fracture-Detection-WCAY .

PMID:39496710 | DOI:10.1038/s41598-024-77878-6

Sci Rep. 2024 Nov 4;14(1):26591. doi: 10.1038/s41598-024-76286-0.

ABSTRACT

Social media has emerged as a dominant platform where individuals freely share opinions and communicate globally. Its role in disseminating news worldwide is significant due to its easy accessibility. However, the increase in the use of these platforms presents severe risks for potentially misleading people. Our research aims to investigate different techniques within machine learning, deep learning, and ensemble learning frameworks in Arabic fake news detection. We integrated FastText word embeddings with various machine learning and deep learning methods. We then leveraged advanced transformer-based models, including BERT, XLNet, and RoBERTa, optimizing their performance through careful hyperparameter tuning. The research methodology involves utilizing two Arabic news article datasets, AFND and ARABICFAKETWEETS datasets, categorized into fake and real subsets and applying comprehensive preprocessing techniques to the text data. Four hybrid deep learning models are presented: CNN-LSTM, RNN-CNN, RNN-LSTM, and Bi-GRU-Bi-LSTM. The Bi-GRU-Bi-LSTM model demonstrated superior performance regarding the F1 score, accuracy, and loss metrics. The precision, recall, F1 score, and accuracy of the hybrid Bi-GRU-Bi-LSTM model on the AFND Dataset are 0.97, 0.97, 0.98, and 0.98, and on the ARABICFAKETWEETS dataset are 0.98, 0.98, 0.99, and 0.99 respectively. The study's primary conclusion is that when spotting fake news in Arabic, the Bi-GRU-Bi-LSTM model outperforms other models by a significant margin. It significantly aids the global fight against false information by setting the stage for future research to expand fake news detection to multiple languages.

PMID:39496680 | DOI:10.1038/s41598-024-76286-0

Sci Rep. 2024 Nov 4;14(1):26661. doi: 10.1038/s41598-024-76711-4.

ABSTRACT

The prediction of photovoltaic (PV) system performance has been intensively studied as it plays an important role in the context of sustainability and renewable energy generation. In this paper, a digital twin (DT) model based on a domain-matched transformer is proposed using convolutional neural network (CNN) for domain-invariant feature extraction, transformer for PV performance prediction, and domain adaptation neural network (DANN) for domain adaptation. The effectiveness of the proposed framework is validated using a PV power prediction dataset. The results indicate an accuracy improvement of up to 39.99% in model performance. Additionally, experiments with varying numbers of timestamps demonstrate enhanced PV power prediction performance as parameters are continuously updated within the DT framework, offering a reliable solution for real-time and adaptive PV power forecasting.

PMID:39496679 | DOI:10.1038/s41598-024-76711-4

Sci Rep. 2024 Nov 4;14(1):26621. doi: 10.1038/s41598-024-64961-1.

ABSTRACT

Biological communication system for neurological disorder patients is similar to the Brain Computer Interface in a way that it facilitates the connection to the outside world in real time. The interdisciplinary field of Electroencephalogram based message depiction is gaining importance as it assists the paralysed person to communicate. In the proposed method a novel approach of feature extraction is done by Empirical Mode Decomposition on non- stationary & non-linear kind of EEG signal. EMD helps in the effective time frequency analysis by disintegrating the EEG signal in the form of six Intrinsic Mode Functions with help of the frequency components. In all nine features are extracted from the decomposed IMFs so as to predict the states or messages of the patient. The above computed features are then served to the Deep Neural Network to perform the classification. The performance of suggested method is studied through applying it to the acquired database generated by the designed hardware as well as also in real time message depiction. The maximum classification accuracy 97% for the acquired database & 85% in real time are obtained respectively by comparative analysis. The command messages generated from the proposed system helps the person suffering from neurological disorder to establish the communication link with the outside world in an efficient way. Thus, the proposed novel method shows better performance in real time message depiction purpose as related to other existing methods.

PMID:39496663 | DOI:10.1038/s41598-024-64961-1

Interact J Med Res. 2024 Nov 4;13:e55067. doi: 10.2196/55067.

ABSTRACT

BACKGROUND: Depression is a prevalent global mental health disorder with substantial individual and societal impact. Natural language processing (NLP), a branch of artificial intelligence, offers the potential for improving depression screening by extracting meaningful information from textual data, but there are challenges and ethical considerations.

OBJECTIVE: This literature review aims to explore existing NLP methods for detecting depression, discuss successes and limitations, address ethical concerns, and highlight potential biases.

METHODS: A literature search was conducted using Semantic Scholar, PubMed, and Google Scholar to identify studies on depression screening using NLP. Keywords included "depression screening," "depression detection," and "natural language processing." Studies were included if they discussed the application of NLP techniques for depression screening or detection. Studies were screened and selected for relevance, with data extracted and synthesized to identify common themes and gaps in the literature.

RESULTS: NLP techniques, including sentiment analysis, linguistic markers, and deep learning models, offer practical tools for depression screening. Supervised and unsupervised machine learning models and large language models like transformers have demonstrated high accuracy in a variety of application domains. However, ethical concerns related to privacy, bias, interpretability, and lack of regulations to protect individuals arise. Furthermore, cultural and multilingual perspectives highlight the need for culturally sensitive models.

CONCLUSIONS: NLP presents opportunities to enhance depression detection, but considerable challenges persist. Ethical concerns must be addressed, governance guidance is needed to mitigate risks, and cross-cultural perspectives must be integrated. Future directions include improving interpretability, personalization, and increased collaboration with domain experts, such as data scientists and machine learning engineers. NLP's potential to enhance mental health care remains promising, depending on overcoming obstacles and continuing innovation.

PMID:39496145 | DOI:10.2196/55067

Neurology. 2024 Nov 26;103(10):e209976. doi: 10.1212/WNL.0000000000209976. Epub 2024 Nov 4.

ABSTRACT

BACKGROUND AND OBJECTIVES: Disentangling brain aging from disease-related neurodegeneration in patients with multiple sclerosis (PwMS) is increasingly topical. The brain-age paradigm offers a window into this problem but may miss disease-specific effects. In this study, we investigated whether a disease-specific model might complement the brain-age gap (BAG) by capturing aspects unique to MS.

METHODS: In this retrospective study, we collected 3D T1-weighted brain MRI scans of PwMS to build (1) a cross-sectional multicentric cohort for age and disease duration (DD) modeling and (2) a longitudinal single-center cohort of patients with early MS as a clinical use case. We trained and evaluated a 3D DenseNet architecture to predict DD from minimally preprocessed images while age predictions were obtained with the DeepBrainNet model. The brain-predicted DD gap (the difference between predicted and actual duration) was proposed as a DD-adjusted global measure of MS-specific brain damage. Model predictions were scrutinized to assess the influence of lesions and brain volumes while the DD gap was biologically and clinically validated within a linear model framework assessing its relationship with BAG and physical disability measured with the Expanded Disability Status Scale (EDSS).

RESULTS: We gathered MRI scans of 4,392 PwMS (69.7% female, age: 42.8 ± 10.6 years, DD: 11.4 ± 9.3 years) from 15 centers while the early MS cohort included 749 sessions from 252 patients (64.7% female, age: 34.5 ± 8.3 years, DD: 0.7 ± 1.2 years). Our model predicted DD better than chance (mean absolute error = 5.63 years, R2 = 0.34) and was nearly orthogonal to the brain-age model (correlation between DD and BAGs: r = 0.06 [0.00-0.13], p = 0.07). Predictions were influenced by distributed variations in brain volume and, unlike brain-predicted age, were sensitive to MS lesions (difference between unfilled and filled scans: 0.55 years [0.51-0.59], p < 0.001). DD gap significantly explained EDSS changes (B = 0.060 [0.038-0.082], p < 0.001), adding to BAG (ΔR2 = 0.012, p < 0.001). Longitudinally, increasing DD gap was associated with greater annualized EDSS change (r = 0.50 [0.39-0.60], p < 0.001), with an incremental contribution in explaining disability worsening compared with changes in BAG alone (ΔR2 = 0.064, p < 0.001).

DISCUSSION: The brain-predicted DD gap is sensitive to MS-related lesions and brain atrophy, adds to the brain-age paradigm in explaining physical disability both cross-sectionally and longitudinally, and may be used as an MS-specific biomarker of disease severity and progression.

PMID:39496109 | DOI:10.1212/WNL.0000000000209976

PLoS One. 2024 Nov 4;19(11):e0308944. doi: 10.1371/journal.pone.0308944. eCollection 2024.

ABSTRACT

In the era of digital intelligence empowerment, the data-driven approach to the mining and organization of humanistic knowledge has ushered in new development opportunities. However, current research on allusions, an important type of humanities data, mainly focuses on the adoption of a traditional paradigm of humanities research. Conversely, little attention is paid to the application of auto-computing techniques to allusive resources. In light of this research gap, this work proposes a model of allusive word sentiment recognition and application based on text semantic enhancement. First, explanatory texts of 36,080 allusive words are introduced for text semantic enhancement. Subsequently, the performances of different deep learning-based approaches are compared, including three baselines and two optimized models. The best model, ERNIE-RCNN, which exhibits a 6.35% improvement in accuracy, is chosen for the sentiment prediction of allusive words based on text semantic enhancement. Next, according to the binary relationships between allusive words and their source text, explanatory text, and sentiments, the overall and time-based distribution regularities of allusive word sentiments are explored. In addition, the sentiments of the source text are inferred according to the allusive word sentiments. Finally, the LDA model is utilized for the topic extraction of allusive words, and the sentiments and topics are fused to construct an allusive word-sentiment theme relationship database, which provides two modes for the semantic association and organization of allusive resources. The empirical results show that the proposed model can achieve the discovery and association of allusion-related humanities knowledge.

PMID:39495816 | DOI:10.1371/journal.pone.0308944

PLoS Comput Biol. 2024 Nov 4;20(11):e1012576. doi: 10.1371/journal.pcbi.1012576. Online ahead of print.

ABSTRACT

The increasing availability of enzyme turnover number measurements from experiments and of turnover number predictions from deep learning models prompts the use of these enzyme parameters in precise metabolic engineering. Yet, there is no computational approach that allows the prediction of metabolic engineering strategies that rely on the modification of turnover numbers. It is also unclear if modifications of turnover numbers without alterations in the host's transcriptional regulatory machinery suffice to increase the production of chemicals of interest. Here, we present a constraint-based modeling approach, termed Overcoming Kinetic rate Obstacles (OKO), that uses enzyme-constrained metabolic models to predict in silico strategies to increase the production of a given chemical, while ensuring specified cell growth. We demonstrate that the application of OKO to enzyme-constrained metabolic models of Escherichia coli and Saccharomyces cerevisiae results in strategies that can at least double the production of over 40 compounds with little penalty to growth. Interestingly, we show that the overproduction of compounds of interest does not entail only an increase in the values of turnover numbers. Lastly, we demonstrate that a refinement of OKO, allowing also for manipulation of enzyme abundance, facilitates the usage of the available compendia and deep learning models of turnover numbers in the design of precise metabolic engineering strategies. Our results expand the usage of genome-scale metabolic models toward the identification of targets for protein engineering, allowing their direct usage in the generation of innovative metabolic engineering designs for various biotechnological applications.

PMID:39495797 | DOI:10.1371/journal.pcbi.1012576

IEEE J Biomed Health Inform. 2024 Nov 4;PP. doi: 10.1109/JBHI.2024.3484951. Online ahead of print.

ABSTRACT

The rapid integration of deep learningpowered artificial intelligence systems in diverse applications such as healthcare, credit assessment, employment, and criminal justice has raised concerns about their fairness, particularly in how they handle various demographic groups. This study delves into the existing biases and their ethical implications in deep learning models. It introduces an UnBias approach for assessing bias in different deep neural network architectures and detects instances where bias seeps into the learning process, shifting the model's focus away from the main features. This contributes to the advancement of equitable and trustworthy AI applications in diverse social settings, especially in healthcare. A case study on COVID-19 detection is carried out, involving chest X-ray scan datasets from various publicly accessible repositories and five well-represented and underrepresented gender-based models across four deep-learning architectures: ResNet50V2, DenseNet121, InceptionV3, and Xception.

PMID:39495690 | DOI:10.1109/JBHI.2024.3484951

IEEE J Biomed Health Inform. 2024 Nov 4;PP. doi: 10.1109/JBHI.2024.3491732. Online ahead of print.

ABSTRACT

The subcellular localization of circular RNAs (circRNAs) is crucial for understanding their functional relevance and regulatory mechanisms. CircRNA subcellular localization exhibits variations across different cell lines, demonstrating the diversity and complexity of circRNA regulation within distinct cellular contexts. However, existing computational methods for predicting circRNA subcellular localization often ignore the importance of cell line specificity and instead train a general model on aggregated data from all cell lines. Considering the diversity and context-dependent behavior of circRNAs across different cell lines, it is imperative to develop cell line-specific models to accurately predict circRNA subcellular localization. In the study, we proposed CellCircLoc, a sequence-based deep learning model for circRNA subcellular localization prediction, which is trained for different cell lines. CellCircLoc utilizes a combination of convolutional neural networks, Transformer blocks, and bidirectional long short-term memory to capture both sequence local features and long-range dependencies within the sequences. In the Transformer blocks, CellCircLoc uses an attentive convolution mechanism to capture the importance of individual nucleotides. Extensive experiments demonstrate the effectiveness of CellCircLoc in accurately predicting circRNA subcellular localization across different cell lines, outperforming other computational models that do not consider cell line specificity. Moreover, the interpretability of CellCircLoc facilitates the discovery of important motifs associated with circRNA subcellular localization. The CellCircLoc web server is available at http://csuligroup.com:8000/cellcircloc. The source code can be obtained from https://github.com/CSUBioGroup/CellCircLoc.

PMID:39495689 | DOI:10.1109/JBHI.2024.3491732

ACS Synth Biol. 2024 Nov 4. doi: 10.1021/acssynbio.4c00542. Online ahead of print.

ABSTRACT

CRISPR-Cas systems have transformed the field of synthetic biology by providing a versatile method for genome editing. The efficiency of CRISPR systems is largely dependent on the sequence of the constituent sgRNA, necessitating the development of computational methods for designing active sgRNAs. While deep learning-based models have shown promise in predicting sgRNA activity, the accuracy of prediction is primarily governed by the data set used in model training. Here, we trained a convolutional neural network (CNN) model and a large language model (LLM) on balanced and imbalanced data sets generated from CRISPR-Cas12a screening data for the yeast Yarrowia lipolytica and evaluated their ability to predict high- and low-activity sgRNAs. We further tested whether prediction performance can be improved by training on imbalanced data sets augmented with synthetic sgRNAs. Lastly, we demonstrated that adding synthetic sgRNAs to inherently imbalanced CRISPR-Cas9 data sets from Y. lipolytica and Komagataella phaffii leads to improved performance in predicting sgRNA activity, thus underscoring the importance of employing balanced training sets for accurate sgRNA activity prediction.

PMID:39495623 | DOI:10.1021/acssynbio.4c00542

Compr Rev Food Sci Food Saf. 2024 Nov;23(6):e70039. doi: 10.1111/1541-4337.70039.

ABSTRACT

Fish-cutting products are widely loved by consumers due to the unique nutrient composition and flavor in different cuts. However, fish-cutting faces the issue of labor shortage due to the harsh working environment, huge workload, and seasonal work. Hence, some automatic, efficient, and large-scale cutting technologies are needed to overcome these challenges. Accompanied by the development of Industry 4.0, the Internet of Things (IoT), artificial intelligence, big data, and blockchain technologies are progressively applied in the cutting process, which plays pivotal roles in digital production monitoring and product safety enhancement. This review focuses on the main fish-cutting schemes and delves into advanced automatic cutting techniques, showing the latest technological advancements and how they are revolutionizing fish cutting. Additionally, the production monitoring architecture based on IoT in the fish-cutting process is discussed. Fish cutting involves a variety of schemes tailored to the specific characteristics of each fish cut. The cutting process includes deheading and tail removal, filleting, boning, skinning, trimming, and bone inspection. By incorporating sensors, machine vision, deep learning, and advanced cutting tools, these technologies are transforming fish cutting from a manual to an automated process. This transformation has significant practical implications for the industry, offering improved efficiency, consistent product quality, and enhanced safety, ultimately providing a modernized manufacturing approach to fish-cutting automation within the context of Industry 4.0.

PMID:39495567 | DOI:10.1111/1541-4337.70039

Phys Eng Sci Med. 2024 Nov 4. doi: 10.1007/s13246-024-01489-8. Online ahead of print.

ABSTRACT

Since its introduction in 2015, the U-Net architecture used in Deep Learning has played a crucial role in medical imaging. Recognized for its ability to accurately discriminate small structures, the U-Net has received more than 2600 citations in academic literature, which motivated continuous enhancements to its architecture. In hospitals, chest radiography is the primary diagnostic method for pulmonary disorders, however, accurate lung segmentation in chest X-ray images remains a challenging task, primarily due to the significant variations in lung shapes and the presence of intense opacities caused by various diseases. This article introduces a new approach for the segmentation of lung X-ray images. Traditional max-pooling operations, commonly employed in conventional U-Net++ models, were replaced with the discrete wavelet transform (DWT), offering a more accurate down-sampling technique that potentially captures detailed features of lung structures. Additionally, we used attention gate (AG) mechanisms that enable the model to focus on specific regions in the input image, which improves the accuracy of the segmentation process. When compared with current techniques like Atrous Convolutions, Improved FCN, Improved SegNet, U-Net, and U-Net++, our method (U-Net++-DWT) showed remarkable efficacy, particularly on the Japanese Society of Radiological Technology dataset, achieving an accuracy of 99.1%, specificity of 98.9%, sensitivity of 97.8%, Dice Coefficient of 97.2%, and Jaccard Index of 96.3%. Its performance on the Montgomery County dataset further demonstrated its consistent effectiveness. Moreover, when applied to additional datasets of Chest X-ray Masks and Labels and COVID-19, our method maintained high performance levels, achieving up to 99.3% accuracy, thereby underscoring its adaptability and potential for broad applications in medical imaging diagnostics.

PMID:39495449 | DOI:10.1007/s13246-024-01489-8