Tomography. 2024 Nov 28;10(12):1895-1914. doi: 10.3390/tomography10120138.

ABSTRACT

Background: Early diagnosis of depression is crucial for effective treatment and suicide prevention. Traditional methods rely on self-report questionnaires and clinical assessments, lacking objective biomarkers. Combining functional magnetic resonance imaging (fMRI) with artificial intelligence can enhance depression diagnosis using neuroimaging indicators, but depression-specific fMRI datasets are often small and imbalanced, posing challenges for classification models. New Method: We propose the Spatio-Temporal Aggregation Network (STANet) for diagnosing depression by integrating convolutional neural networks (CNN) and recurrent neural networks (RNN) to capture both temporal and spatial features of brain activity. STANet comprises the following steps: (1) Aggregate spatio-temporal information via independent component analysis (ICA). (2) Utilize multi-scale deep convolution to capture detailed features. (3) Balance data using the synthetic minority over-sampling technique (SMOTE) to generate new samples for minority classes. (4) Employ the attention-Fourier gate recurrent unit (AFGRU) classifier to capture long-term dependencies, with an adaptive weight assignment mechanism to enhance model generalization. Results: STANet achieves superior depression diagnostic performance, with 82.38% accuracy and a 90.72% AUC. The Spatio-Temporal Feature Aggregation module enhances classification by capturing deeper features at multiple scales. The AFGRU classifier, with adaptive weights and a stacked Gated Recurrent Unit (GRU), attains higher accuracy and AUC. SMOTE outperforms other oversampling methods. Additionally, spatio-temporal aggregated features achieve better performance compared to using only temporal or spatial features. Comparison with existing methods: STANet significantly outperforms traditional classifiers, deep learning classifiers, and functional connectivity-based classifiers. Conclusions: The successful performance of STANet contributes to enhancing the diagnosis and treatment assessment of depression in clinical settings on imbalanced and small fMRI.

PMID:39728900 | DOI:10.3390/tomography10120138

Nurs Rep. 2024 Dec 20;14(4):4162-4172. doi: 10.3390/nursrep14040303.

ABSTRACT

BACKGROUND: Optimum efficiency and responsiveness to callers of mental health helplines can only be achieved if call priority is accurately identified. Currently, call operators making a triage assessment rely heavily on their clinical judgment and experience. Due to the significant morbidity and mortality associated with mental illness, there is an urgent need to identify callers to helplines who have a high level of distress and need to be seen by a clinician who can offer interventions for treatment. This study delves into the potential of using machine learning (ML) to estimate call priority from the properties of the callers' voices rather than evaluating the spoken words.

METHOD: Phone callers' speech is first isolated using existing APIs, then features or representations are extracted from the raw speech. These are then fed into a series of deep learning neural networks to classify priority level from the audio representation.

RESULTS: Development of a deep learning neural network architecture that instantly determines positive and negative levels in the input speech segments. A total of 459 call records from a mental health helpline were investigated. The final ML model achieved a balanced accuracy of 92% correct identification of both positive and negative instances of call priority.

CONCLUSIONS: The priority level provides an estimate of voice quality in terms of positive or negative demeanor that can be simultaneously displayed using a web interface on a computer or smartphone.

PMID:39728664 | DOI:10.3390/nursrep14040303

J Xenobiot. 2024 Dec 4;14(4):1901-1918. doi: 10.3390/jox14040101.

ABSTRACT

In the field of computational chemistry, computer models are quickly and cheaply constructed to predict toxicology hazards and results, with no need for test material or animals as these computational predictions are often based on physicochemical properties of chemical structures. Multiple methodologies are employed to support in silico assessments based on machine learning (ML) and deep learning (DL). This review introduces the development of computational toxicology, focusing on ML and DL and emphasizing their importance in the field of toxicology. A fine balance between target potency, selectivity, absorption, distribution, metabolism, excretion, toxicity (ADMET) and clinical safety properties should be achieved to discover a potential new drug. It is advantageous to perform virtual predictions as early as possible in drug development processes, even before a molecule is synthesized. Currently, there are numerous commercially available and free web-based programs for toxicity prediction, which can be used to construct various predictive models. The key features of the QSAR method are also outlined, and the selection of appropriate physicochemical descriptors is a prerequisite for robust predictions. In addition, examples of open-source tools applied to toxicity prediction are included, as well as examples of the application of different computational methods for the prediction of toxicity in drug design and environmental toxicology.

PMID:39728409 | DOI:10.3390/jox14040101

J Imaging. 2024 Dec 23;10(12):333. doi: 10.3390/jimaging10120333.

ABSTRACT

Alzheimer's disease (AD), a degenerative condition affecting the central nervous system, has witnessed a notable rise in prevalence along with the increasing aging population. In recent years, the integration of cutting-edge medical imaging technologies with forefront theories in artificial intelligence has dramatically enhanced the efficiency of identifying and diagnosing brain diseases such as AD. This paper presents an innovative two-stage automatic auxiliary diagnosis algorithm for AD, based on an improved 3D DenseNet segmentation model and an improved MobileNetV3 classification model applied to brain MR images. In the segmentation network, the backbone network was simplified, the activation function and loss function were replaced, and the 3D GAM attention mechanism was introduced. In the classification network, firstly, the CA attention mechanism was added to enhance the model's ability to capture positional information of disease features; secondly, dilated convolutions were introduced to extract richer features from the input feature maps; and finally, the fully connected layer of MobileNetV3 was modified and the idea of transfer learning was adopted to improve the model's feature extraction capability. The results of the study showed that the proposed approach achieved classification accuracies of 97.85% for AD/NC, 95.31% for MCI/NC, 93.96% for AD/MCI, and 92.63% for AD/MCI/NC, respectively, which were 3.1, 2.8, 2.6, and 2.8 percentage points higher than before the improvement. Comparative and ablation experiments have validated the proposed classification performance of this method, demonstrating its capability to facilitate an accurate and efficient automated auxiliary diagnosis of AD, offering a deep learning-based solution for it.

PMID:39728230 | DOI:10.3390/jimaging10120333

Physiol Plant. 2025 Jan-Feb;177(1):e70011. doi: 10.1111/ppl.70011.

ABSTRACT

As an important source of pollution in the papermaking process, the presence of lignin in poplar can seriously affect the quality and process of pulping. During lignin synthesis, Caffeoyl-CoA-O methyltransferase (CCoAOMT), as a specialized catalytic transferase, can effectively regulate the methylation of caffeoyl-coenzyme A (CCoA) to feruloyl-coenzyme A. Targeting CCoAOMT, this study investigated the substrate recognition mechanism and the possible reaction mechanism, the key residues of lignin binding were mutated and the lignin content was validated by deep convolutional neural-network model based on genome-wide prediction (DCNGP). The molecular mechanics results indicate that the binding of S-adenosyl methionine (SAM) and CCoA is sequential, with SAM first binding and inducing inward constriction of the CCoAOMT; then CCoA binds to the pocket, and this process closes the outer channel, preventing contamination by impurities and ensuring that the reaction proceeds. Next, the key residues in the recognition process of SAM (F69 and D91) and CCoA (I40, N170, Y188 and D218) were analyzed, and we identified that K146 as a base catalyst is important for inducing the methylation reaction. Immediately after that, the possible methylation reaction mechanism was deduced by the combination of Restrained Electrostatic Potential (RESP) and Independent Gradient Model (IGM) analysis, focusing on the catalytic center electron cloud density and RESP charge distribution. Finally, the DCNGP results verified that the designed mutant groups were all able to effectively reduce the lignin content and increase the S-lignin content/ G-lignin content ratio, which was beneficial for the subsequent lignin removal. Multifaceted consideration of factors that reduce lignin content and combined deep learning to screen for favorable mutations in target traits provides new ideas for targeted breeding of low-lignin poplars.

PMID:39727026 | DOI:10.1111/ppl.70011

Brief Bioinform. 2024 Nov 22;26(1):bbae672. doi: 10.1093/bib/bbae672.

ABSTRACT

Identifying phage-host interactions (PHIs) is a crucial step in developing phage therapy, which is the promising solution to addressing the issue of antibiotic resistance in superbugs. However, the lifestyle of phages, which strongly depends on their host for life activities, limits their cultivability, making the study of predicting PHIs time-consuming and labor-intensive for traditional wet lab experiments. Although many deep learning (DL) approaches have been applied to PHIs prediction, most DL methods are predominantly based on sequence information, failing to comprehensively model the intricate relationships within PHIs. Moreover, most existing approaches are limited for sub-optimal performance, due to the potential risk of overfitting induced by the highly data sparsity in the task of PHIs prediction. In this study, we propose a novel approach called MI-RGC, which introduces mutual information for feature augmentation and employs regional graph convolution to learn meaningful representations. Specifically, MI-RGC treats the presence status of phages in environmental samples as random variables, and derives the mutual information between these random variables as the dependency relationships among phages. Consequently, a mutual information-based heterogeneous network is construted as feature augmentation for sequence information of phages, which is utilized for building a sequence information-based heterogeneous network. By considering the different contributions of neighboring nodes at varying distances, a regional graph convolutional model is designed, in which the neighboring nodes are segmented into different regions and a regional-level attention mechanism is employed to derive node embeddings. Finally, the embeddings learned from these two networks are aggregated through an attention mechanism, on which the prediction of PHIs is condcuted accordingly. Experimental results on three benchmark datasets demonstrate that MI-RGC derives superior performance over other methods on the task of PHIs prediction.

PMID:39727002 | DOI:10.1093/bib/bbae672

Proc (IEEE Int Conf Healthc Inform). 2024 Jun;2024:305-311. doi: 10.1109/ichi61247.2024.00046. Epub 2024 Aug 22.

ABSTRACT

Delirium is an acute decline or fluctuation in attention, awareness, or other cognitive function that can lead to serious adverse outcomes. Despite the severe outcomes, delirium is frequently unrecognized and uncoded in patients' electronic health records (EHRs) due to its transient and diverse nature. Natural language processing (NLP), a key technology that extracts medical concepts from clinical narratives, has shown great potential in studies of delirium outcomes and symptoms. To assist in the diagnosis and phenotyping of delirium, we formed an expert panel to categorize diverse delirium symptoms, composed annotation guidelines, created a delirium corpus with diverse delirium symptoms, and developed NLP methods to extract delirium symptoms from clinical notes. We compared 5 state-of-the-art transformer models including 2 models (BERT and RoBERTa) from the general domain and 3 models (BERT_MIMIC, RoBERTa_MIMIC, and GatorTron) from the clinical domain. GatorTron achieved the best strict and lenient F1 scores of 0.8055 and 0.8759, respectively. We conducted an error analysis to identify challenges in annotating delirium symptoms and developing NLP systems. To the best of our knowledge, this is the first large language model-based delirium symptom extraction system. Our study lays the foundation for the future development of computable phenotypes and diagnosis methods for delirium.

PMID:39726986 | PMC:PMC11670120 | DOI:10.1109/ichi61247.2024.00046

Front Bioeng Biotechnol. 2024 Dec 11;12:1486789. doi: 10.3389/fbioe.2024.1486789. eCollection 2024.

ABSTRACT

This scoping review summarizes two emerging electrical impedance technologies: electrical impedance myography (EIM) and electrical impedance tomography (EIT). These methods involve injecting a current into tissue and recording the response at different frequencies to understand tissue properties. The review discusses basic methods and trends, particularly the use of electrodes: EIM uses electrodes for either injection or recording, while EIT uses them for both. Ag/AgCl electrodes are prevalent, and current injection is preferred over voltage injection due to better resistance to electrode wear and impedance changes. Advances in digital processing and integrated circuits have shifted EIM and EIT toward digital acquisition, using voltage-controlled current sources (VCCSs) that support multiple frequencies. The review details powerful processing algorithms and reconstruction tools for EIT and EIM, examining their strengths and weaknesses. It also summarizes commercial devices and clinical applications: EIT is effective for detecting cancerous tissue and monitoring pulmonary issues, while EIM is used for neuromuscular disease detection and monitoring. The role of machine learning and deep learning in advancing diagnosis, treatment planning, and monitoring is highlighted. This review provides a roadmap for researchers on device evolution, algorithms, reconstruction tools, and datasets, offering clinicians and researchers information on commercial devices and clinical studies for effective use and innovative research.

PMID:39726983 | PMC:PMC11670078 | DOI:10.3389/fbioe.2024.1486789

Front Artif Intell. 2024 Dec 11;7:1312115. doi: 10.3389/frai.2024.1312115. eCollection 2024.

ABSTRACT

The ability to accurately predict the yields of different crop genotypes in response to weather variability is crucial for developing climate resilient crop cultivars. Genotype-environment interactions introduce large variations in crop-climate responses, and are hard to factor in to breeding programs. Data-driven approaches, particularly those based on machine learning, can help guide breeding efforts by factoring in genotype-environment interactions when making yield predictions. Using a new yield dataset containing 93,028 records of soybean hybrids across 159 locations, 28 states, and 13 years, with 5,838 distinct genotypes and daily weather data over a 214-day growing season, we developed two convolutional neural network (CNN) models: one that integrates CNN and fully-connected neural networks (CNN model), and another that incorporates a long short-term memory (LSTM) layer after the CNN component (CNN-LSTM model). By applying the Generalized Ensemble Method (GEM), we combined the CNN-based models and optimized their weights to improve overall predictive performance. The dataset provided unique genotype information on seeds, enabling an investigation into the potential of planting different genotypes based on weather variables. We employed the proposed GEM model to identify the best-performing genotypes across various locations and weather conditions, making yield predictions for all potential genotypes in each specific setting. To assess the performance of the GEM model, we evaluated it on unseen genotype-location combinations, simulating real-world scenarios where new genotypes are introduced. By combining the base models, the GEM ensemble approach provided much better prediction accuracy compared to using the CNN-LSTM model alone and slightly better accuracy than the CNN model, as measured by both RMSE and MAE on the validation and test sets. The proposed data-driven approach can be valuable for genotype selection in scenarios with limited testing years. In addition, we explored the impact of incorporating state-level soil data alongside the weather, location, genotype and year variables. Due to data constraints, including the absence of latitude and longitude details, we used uniform soil variables for all locations within the same state. This limitation restricted our spatial information to state-level knowledge. Our findings suggested that integrating state-level soil variables did not substantially enhance the predictive capabilities of the models. We also performed a feature importance analysis using RMSE change to identify crucial predictors. Location showed the highest RMSE change, followed by genotype and year. Among weather variables, maximum direct normal irradiance (MDNI) and average precipitation (AP) displayed higher RMSE changes, indicating their importance.

PMID:39726891 | PMC:PMC11670329 | DOI:10.3389/frai.2024.1312115

Front Artif Intell. 2024 Dec 12;7:1474932. doi: 10.3389/frai.2024.1474932. eCollection 2024.

ABSTRACT

The construction industry is rapidly adopting Industry 4.0 technologies, creating new opportunities to address persistent environmental and operational challenges. This review focuses on how Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) are being leveraged to tackle these issues. It specifically explores AI's role in predicting air pollution, improving material quality, monitoring worker health and safety, and enhancing Cyber-Physical Systems (CPS) for construction. This study evaluates various AI and ML models, including Artificial Neural Networks (ANNs) and Support Vector Machines SVMs, as well as optimization techniques like whale and moth flame optimization. These tools are assessed for their ability to predict air pollutant levels, improve concrete quality, and monitor worker safety in real time. Research papers were also reviewed to understand AI's application in predicting the compressive strength of materials like cement mortar, fly ash, and stabilized clay soil. The performance of these models is measured using metrics such as coefficient of determination (R 2), Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE). Furthermore, AI has shown promise in predicting and reducing emissions of air pollutants such as PM2.5, PM10, NO2, CO, SO2, and O3. In addition, it improves construction material quality and ensures worker safety by monitoring health indicators like standing postures, electrocardiogram, and galvanic skin response. It is also concluded that AI technologies, including Explainable AI and Petri Nets, are also making advancements in CPS for the construction industry. The models' performance metrics indicate they are well-suited for real-time construction operations. The study highlights the adaptability and effectiveness of these technologies in meeting current and future construction needs. However, gaps remain in certain areas of research, such as broader AI integration across diverse construction environments and the need for further validation of models in real-world applications. Finally, this research underscores the potential of AI and ML to revolutionize the construction industry by promoting sustainable practices, improving operational efficiency, and addressing safety concerns. It also provides a roadmap for future research, offering valuable insights for industry stakeholders interested in adopting AI technologies.

PMID:39726889 | PMC:PMC11669660 | DOI:10.3389/frai.2024.1474932

J Biomed Phys Eng. 2024 Dec 1;14(6):579-592. doi: 10.31661/jbpe.v0i0.2207-1521. eCollection 2024 Dec.

ABSTRACT

BACKGROUND: The P300 signal, an endogenous component of event-related potentials, is extracted from an electroencephalography signal and employed in Brain-computer Interface (BCI) devices.

OBJECTIVE: The current study aimed to address challenges in extracting useful features from P300 components and detecting P300 through a hybrid unsupervised manner based on Convolutional Neural Network (CNN) and Long Short-term Memory (LSTM).

MATERIAL AND METHODS: In this cross-sectional study, CNN as a useful method for the P300 classification task emphasizes spatial characteristics of data. However, CNN and LSTM networks are combined to modify the classification system by extracting both spatial and temporal features. Then, the CNN-LSTM network was trained in an unsupervised learning method based on an autoencoder to improve Signal-to-noise Ratio (SNR) by extracting main components from latent space. To deal with imbalanced data, an Adaptive Synthetic Sampling Approach (ADASYN) is used and augmented without any duplication.

RESULTS: The trained model, tested on the BCI competition III dataset, including two normal subjects, with an accuracy of 95% and 94% for subjects A and B in P300 detection, respectively.

CONCLUSION: CNN-LSTM, was embedded into an autoencoder and introduced to simultaneously extract spatial and temporal features and manage the computational complexity of the method. Further, ADASYN as an augmentation method was proposed to deal with the imbalanced nature of data, which not only maintained feature space as before but also preserved anatomical features of P300. High-quality results highlight the suitable efficiency of the proposed method.

PMID:39726882 | PMC:PMC11668936 | DOI:10.31661/jbpe.v0i0.2207-1521

Digit Discov. 2024 Dec 16. doi: 10.1039/d4dd00311j. Online ahead of print.

ABSTRACT

Deep learning has significantly accelerated drug discovery, with 'chemical language' processing (CLP) emerging as a prominent approach. CLP approaches learn from molecular string representations (e.g., Simplified Molecular Input Line Entry Systems [SMILES] and Self-Referencing Embedded Strings [SELFIES]) with methods akin to natural language processing. Despite their growing importance, training predictive CLP models is far from trivial, as it involves many 'bells and whistles'. Here, we analyze the key elements of CLP and provide guidelines for newcomers and experts. Our study spans three neural network architectures, two string representations, three embedding strategies, across ten bioactivity datasets, for both classification and regression purposes. This 'hitchhiker's guide' not only underscores the importance of certain methodological decisions, but it also equips researchers with practical recommendations on ideal choices, e.g., in terms of neural network architectures, molecular representations, and hyperparameter optimization.

PMID:39726698 | PMC:PMC11667676 | DOI:10.1039/d4dd00311j

Front Med (Lausanne). 2024 Dec 12;11:1520878. doi: 10.3389/fmed.2024.1520878. eCollection 2024.

ABSTRACT

OBJECTIVES: To implement state-of-the-art deep learning architectures such as Deep-Residual-U-Net and DeepLabV3+ for precise segmentation of hippocampus and ventricles, in functional magnetic resonance imaging (fMRI). Integrate VGG-16 with Random Forest (VGG-16-RF) and VGG-16 with Support Vector Machine (VGG-16-SVM) to enhance the binary classification accuracy of Alzheimer's disease, comparing their performance against traditional classifiers.

METHOD: OpenNeuro and Harvard's Data verse provides Alzheimer's coronal functional MRI data. Ventricles and hippocampus are segmented using a Deep-Residual-UNet and Deep labV3+ system. The functional features were extracted from each segmented component and classified using SVM, Adaboost, Logistic regression, and VGG 16, DenseNet-169, VGG-16-RF, and VGG-16-SVM classifier.

RESULTS: This research proposes a precise and efficient deep-learning architecture like DeepLab V3+ and Deep Residual U-NET for hippocampus and ventricle segmentation in detection of AD. DeepLab V3+ has produced a good segmentation accuracy of 94.62% with Jaccard co-efficient of 85.5% and dice co-efficient of 84.75%. Among the three ML classifiers used, SVM has provided a good accuracy of 93%. Among some DL techniques, VGG-16-RF classifier has given better accuracy of 96.87%.

CONCLUSION: The novelty of this work lies in the seamless integration of advanced segmentation techniques with hybrid classifiers, offering a robust and scalable framework for early AD detection. The proposed study demonstrates a significant advancement in the early detection of Alzheimer's disease by integrating state-of-the-art deep learning models and comprehensive functional connectivity analysis. This early detection capability is crucial for timely intervention and better management of the disease in neurodegenerative disorder diagnostics.

PMID:39726682 | PMC:PMC11669652 | DOI:10.3389/fmed.2024.1520878

Interspeech. 2024 Sep;2024:927-931. doi: 10.21437/interspeech.2024-2236.

ABSTRACT

Magnetic Resonance Imaging (MRI) allows analyzing speech production by capturing high-resolution images of the dynamic processes in the vocal tract. In clinical applications, combining MRI with synchronized speech recordings leads to improved patient outcomes, especially if a phonological-based approach is used for assessment. However, when audio signals are unavailable, the recognition accuracy of sounds is decreased when using only MRI data. We propose a contrastive learning approach to improve the detection of phonological classes from MRI data when acoustic signals are not available at inference time. We demonstrate that frame-wise recognition of phonological classes improves from an f1 of 0.74 to 0.85 when the contrastive loss approach is implemented. Furthermore, we show the utility of our approach in the clinical application of using such phonological classes to assess speech disorders in patients with tongue cancer, yielding promising results in the recognition task.

PMID:39726638 | PMC:PMC11671147 | DOI:10.21437/interspeech.2024-2236

BMC Bioinformatics. 2024 Dec 26;25(1):391. doi: 10.1186/s12859-024-06016-w.

ABSTRACT

As a heterogeneous disease, prostate cancer (PCa) exhibits diverse clinical and biological features, which pose significant challenges for early diagnosis and treatment. Metabolomics offers promising new approaches for early diagnosis, treatment, and prognosis of PCa. However, metabolomics data are characterized by high dimensionality, noise, variability, and small sample sizes, presenting substantial challenges for classification. Despite the wide range of applications of deep learning methods, the use of deep learning in metabolomics research has not been extensively explored. In this study, we propose a hybrid model, TransConvNet, which combines transformer and convolutional neural networks for the classification of prostate cancer metabolomics data. We introduce a 1D convolution layer for the inputs to the dot-product attention mechanism, enabling the interaction of both local and global information. Additionally, a gating mechanism is incorporated to dynamically adjust the attention weights. The features extracted by multi-head attention are further refined through 1D convolution, and a residual network is introduced to alleviate the gradient vanishing problem in the convolutional layers. We conducted comparative experiments with seven other machine learning algorithms. Through five-fold cross-validation, TransConvNet achieved an accuracy of 81.03% and an AUC of 0.89, significantly outperforming the other algorithms. Additionally, we validated TransConvNet's generalization ability through experiments on the lung cancer dataset, with the results demonstrating its robustness and adaptability to different metabolomics datasets. We also proposed the MI-RF (Mutual Information-based random forest) model, which effectively identified key biomarkers associated with prostate cancer by leveraging comprehensive feature weight coefficients. In contrast, traditional methods identified only a limited number of biomarkers. In summary, these results highlight the potential of TransConvNet and MI-RF in both classification tasks and biomarker discovery, providing valuable insights for the clinical application of prostate cancer diagnosis.

PMID:39725937 | DOI:10.1186/s12859-024-06016-w

Nan Fang Yi Ke Da Xue Xue Bao. 2024 Dec 20;44(12):2412-2420. doi: 10.12122/j.issn.1673-4254.2024.12.18.

ABSTRACT

METHODS: We retrospectively collected CT scan data from 276 patients with pathologically confirmed primary bone tumors from 4 medical centers in Guangdong Province between January, 2010 and August, 2021. A convolutional neural network (CNN) was employed as the deep learning architecture. The optimal baseline deep learning model (R-Net) was determined through transfer learning, and an optimized model (S-Net) was obtained through algorithmic improvements. Multivariate logistic regression analysis was used to screen the clinical features such as sex, age, mineralization location, and pathological fractures, which were then connected with the imaging features to construct the deep learning fusion model (SC-Net). The diagnostic performance of the SC-Net model and machine learning models were compared with radiologists' diagnoses, and their classification performance was evaluated using the area under the receiver operating characteristic curve (AUC) and F1 score.

RESULTS: In the external test set, the fusion model (SC-Net) achieved the best performance with an AUC of 0.901 (95% CI: 0.803-1.00), an accuracy of 83.7% (95% CI: 69.3%-93.2%) and an F1 score of 0.857, and outperformed the S-Net model with an AUC of 0.818 (95% CI: 0.694-0.942), an accuracy of 76.7% (95% CI: 61.4%-88.2%), and an F1 score of 0.828. The overall classification performance of the fusion model (SC-Net) exceeded that of radiologists' diagnoses.

CONCLUSIONS: The deep learning fusion model based on multi-center CT images and clinical features is capable of accurate classification of osseous and chondroid matrix mineralization and may potentially improve the accuracy of clinical diagnoses of osteogenic versus chondrogenic primary bone tumors.

PMID:39725631 | DOI:10.12122/j.issn.1673-4254.2024.12.18

Oral Surg Oral Med Oral Pathol Oral Radiol. 2024 Nov 12:S2212-4403(24)00794-6. doi: 10.1016/j.oooo.2024.11.002. Online ahead of print.

ABSTRACT

OBJECTIVE: This study aimed to develop 3 models based on computed tomography (CT) images of patients with craniofacial fibrous dysplasia (CFD): a radiomics model (Model Rad), a deep learning (DL) model (Model DL), and a hybrid radiomics and DL model (Model Rad+DL), and evaluate the ability of these models to distinguish between adolescents with active lesion progression and adults with stable lesion progression.

METHODS: We retrospectively analyzed preoperative CT scans from 148 CFD patients treated at Shanghai Ninth People's Hospital. The images were processed using 3D-Slicer software to segment and extract regions of interest for radiomics and DL analysis. Feature selection was performed using t-tests, mutual information, correlation tests, and the least absolute shrinkage and selection operator algorithm to develop the 3 models. Model accuracy was evaluated using measurements including the area under the curve (AUC) derived from receiver operating characteristic analysis, sensitivity, specificity, and F1 score. Decision curve analysis (DCA) was conducted to evaluate clinical benefits.

RESULTS: In total, 1,130 radiomics features and 512 DL features were successfully extracted. Model Rad+DL demonstrated superior AUC values compared to Model Rad and Model DL in the training and validation sets. DCA revealed that Model Rad+DL offered excellent clinical benefits when the threshold probability exceeded 20%.

CONCLUSIONS: Model Rad+DL exhibits superior potential in evaluating CFD progression, determining the optimal surgical timing for adult CFD patients.

PMID:39725588 | DOI:10.1016/j.oooo.2024.11.002

Arthroscopy. 2024 Dec 24:S0749-8063(24)01088-0. doi: 10.1016/j.arthro.2024.12.024. Online ahead of print.

ABSTRACT

PURPOSE: Improve the accuracy of one-stage object detection by modifying the YOLOv7 with Convolutional Block Attention Module (CBAM), known as YOLOv7-CBAM, which can automatically identify torn or intact rotator cuff tendon to assist physicians in diagnosing rotator cuff lesions through ultrasound.

METHODS: Between 2020 and 2021, patients who experienced shoulder pain for over 3 months and had both ultrasound and MRI examinations were categorized into torn and intact group. To ensure balanced training, we included the same number of patients on both groups. Transfer learning was conducted using a pre-trained model of Yolov7 and an improved model with CBAM. The mean average precision (mAP), sensitivity and F1-score were calculated to evaluate the models. Gradient-weighted Class Activation Mapping (Grad-CAM) method was employed to visualize important regions using a heatmap. Simulation dataset was recruited to evaluate the diagnostic performance of clinical physicians using our AI-assisted model.

RESULTS: A total of 280 patients were included in this study, with 80% of 840 ultrasound images randomly allocated for model training. The accuracy for test set was 0.96 for Yolov7 and 0.98 for Yolov7-CBAM, the precision and sensitivity were 0.94 and 0.98 for Yolov7, 0.98 and 0.98 for Yolov7-CBAM. F1-score and mAP@0.5 were higher for Yolov7-CBAM (0.980 and 0.993) than Yolov7 (0.961 and 0.965). Furthermore, the Grad-CAM method elucidated that the deep learning model primarily emphasized hypoechoic anechoic defect within the tendon. Following adopting an AI-assisted model (YOLOv7-CBAM model), diagnostic accuracy improved from 80.86% to 88.86% (p=0.01) and interobserver reliability improved from 0.49 to 0.71 among physicians.

CONCLUSION: The YOLOv7-CBAM model demonstrate high accuracy in detecting torn or intact rotator cuff tendon from ultrasound images. Integrating this model into the diagnostic process can assist physicians in improving diagnostic accuracy and interobserver reliability across different physicians.

CLINICAL RELEVANCE: The attentional deep learning model aids physicians in improving the accuracy and consistency of ultrasound diagnosis of torn or intact rotator cuff tendons.

PMID:39725049 | DOI:10.1016/j.arthro.2024.12.024

J Neuropathol Exp Neurol. 2024 Dec 26:nlae120. doi: 10.1093/jnen/nlae120. Online ahead of print.

ABSTRACT

Microinfarcts and microhemorrhages are characteristic lesions of cerebrovascular disease. Although multiple studies have been published, there is no one universal standard criteria for the neuropathological assessment of cerebrovascular disease. In this study, we propose a novel application of machine learning in the automated screening of microinfarcts and microhemorrhages. Utilizing whole slide images (WSIs) from postmortem human brain samples, we adapted a patch-based pipeline with convolutional neural networks. Our cohort consisted of 22 cases from the University of California Davis Alzheimer's Disease Research Center brain bank with hematoxylin and eosin-stained formalin-fixed, paraffin-embedded sections across 3 anatomical areas: frontal, parietal, and occipital lobes (40 WSIs with microinfarcts and/or microhemorrhages, 26 without). We propose a multiple field-of-view prediction step to mitigate false positives. We report screening performance (ie, the ability to distinguish microinfarct/microhemorrhage-positive from microinfarct/microhemorrhage-negative WSIs), and detection performance (ie, the ability to localize the affected regions within a WSI). Our proposed approach improved detection precision and screening accuracy by reducing false positives thereby achieving 100% screening accuracy. Although this sample size is small, this pipeline provides a proof-of-concept for high efficacy in screening for characteristic brain changes of cerebrovascular disease to aid in screening of microinfarcts/microhemorrhages at the WSI level.

PMID:39724914 | DOI:10.1093/jnen/nlae120

Georgian Med News. 2024 Oct;(355):184-187.

ABSTRACT

The corona virus disease-19 (COVID-19) epidemic, the whole globe is suffering from a medical condition catastrophe that is unprecedented in scale. As the coronavirus spreads, scientists are worried about offering or helping in the supply of remedies to preserve lives and end the epidemic. Artificial intelligence (AI), for example, has occurred altered to deal with the difficulties raised by pandemics. We provide an in-depth learning approach for locating and extracting attributes of COVID-19 from Chest X-rays. Hierarchical multilevel ResNet50 (HMResNet50) was adjusted to better COVID-19 data, which was collected to build this dataset with images of a typical chest X-ray from numerous public sources. We employed information enhancement methods such as randomized rotations with a ten-ten-degree slant, random noise, and horizontal flips to generate numerous images of chest X-ray. Outcome of the research is encouraging: the suggested models correctly identified COVID-19 chest X-rays or standard with an accuracy of 99.10 % for Resnet50 and 97.20 % for hierarchal Multilevel Resnet50. The findings suggest that the proposed is effective, with high performance and simple COVID-19 recognition methods.

PMID:39724901