Deep Learning Approach Research Articles

Deep Learning Approach For Suspicious Activity Detection from Surveillance Video in Examination Hall

Deep Learning Approach For Suspicious Activity Detection from Surveillance Video in Examination Hall

From Pixels to Protection: Deep Learning Approaches for Plant Leaf Disease Detection

From Pixels to Protection: Deep Learning Approaches for Plant Leaf Disease Detection

Risk score stratification of cutaneous melanoma patients based on whole slide images analysis by deep learning.

There is a need to improve risk stratification of primary cutaneous melanomas to better guide adjuvant therapy. Taking into account that haematoxylin and eosin (HE)-stained tumour tissue contains a huge amount of clinically unexploited morphological informations, we developed a weakly-supervised deep-learning approach, SmartProg-MEL, to predict survival outcomes in stages I to III melanoma patients from HE-stained whole slide image (WSI). We designed a deep neural network that extracts morphological features from WSI to predict 5-y overall survival (OS), and assign a survival risk score to each patient. The model was trained and validated on a discovery cohort of primary cutaneous melanomas (IHP-MEL-1, n = 342). Performance was tested on two external and independent datasets (IHP-MEL-2, n = 161; and TCGA cohort n = 63). It was compared with well-established prognostic factors. Concordance index (c-index) was used as a metric. On the discovery cohort, the SmartProg-MEL predicts the 5-y OS with a c-index of 0.78 on the cross-validation data and of 0.72 on the cross-testing series. In the external cohorts, the model achieved a c-index of 0.71 and 0.69 for the IHP-MEL-2 and TCGA dataset respectively. Furthermore, SmartProg-MEL was an independent and the most powerful prognostic factor in multivariate analysis (HR = 1.84, p-value < 0.005). Finally, the model was able to dichotomize patients in two groups-a low and a high-risk group-each associated with a significantly different 5-y OS (p-value < 0.001 for IHP-MEL-1 and p-value = 0.01 for IHP-MEL-2). The performance of our fully automated SmartProg-MEL model outperforms the current clinicopathological factors in terms of prediction of 5-y OS and risk stratification of cutaneous melanoma patients. Incorporation of SmartProg-MEL in the clinical workflow could guide the decision-making process by improving the identification of patients that may benefit from adjuvant therapy.

Dynamic cycles between brain states during creative storytelling.

Many theories suggest that creative thinking involves a dynamic transition between different mental states, yet empirical evidence supporting this notion remains scarce. The dual process model proposes that spontaneous thinking and deliberate thinking drive the dwell in and the transitions between different mental states during creative thinking, but there is a debate over whether the two types of thinking operate in parallel or in sequence. To address these gaps, we conducted a functional magnetic resonance imaging (fMRI) study in 41 college students during a creative storytelling task. We then compared the dynamic brain states in creative versus uncreative storytelling to identify key brain states associated with creative thinking. And we further performed correlation analysis between these key brain states with performance of various creative tasks, trying to link the key brain states with different cognitive processes. The results showed that two key brain states are associated with creative thinking, with one involving whole-brain synchronization and the other involving the synchronization of four networks, including the default mode network and the control network. The transition patterns between the key brain states provide tentative evidence for dynamic circulation between different mental states during creative storytelling. Using a deep learning approach, we demonstrate an alternating interaction between spontaneous and deliberate thinking, driving dwelling in and the transitions between different brain states. These findings deepen our understanding of the cognitive and neural mechanisms underlying creative thinking.

High-throughput markerless pose estimation and home-cage activity analysis of tree shrew using deep learning.

Quantifying the rich home-cage activities of tree shrews provides a reliable basis for understanding their daily routines and building disease models. However, due to the lack of effective behavioral methods, most efforts on tree shrew behavior are limited to simple measures, resulting in the loss of much behavioral information. To address this issue, we present a deep learning (DL) approach to achieve markerless pose estimation and recognize multiple spontaneous behaviors of tree shrews, including drinking, eating, resting, and staying in the dark house, etc. RESULTS: This high-throughput approach can monitor the home-cage activities of 16 tree shrews simultaneously over an extended period. Additionally, we demonstrated an innovative system with reliable apparatus, paradigms, and analysis methods for investigating food grasping behavior. The median duration for each bout of grasping was 0.20 s. This study provides an efficient tool for quantifying and understand tree shrews' natural behaviors.

Explainable AI for Pancreatic Cancer Prediction and Survival Prognosis: An Interpretable Deep Learning and Machine Learning Approach

Explainable AI for Pancreatic Cancer Prediction and Survival Prognosis: An Interpretable Deep Learning and Machine Learning Approach

Erlang Spectrogram and Residual Network-Based Features for Fake Audio Detection

Advancements in deep learning and other approaches are making synthetic speech more natural-sounding. These advancements enable people to train a speech synthesizer with a target voice, resulting in a model that can imitate someone's voice with excellent accuracy. To overcome these challenges, in this work, we are presenting a fusion of Spectrogram and Deep Convolutional Neural Network (DeepCNN) for detecting fake audio. The work has been implemented in three phases: Spectrogram generation, feature extraction & reduction, and classification. In the first step, Erlang spectrograms have been generated using the proposed Erlang filter bank. These generated spectrograms are then passed to the Residual Network (ResNet11) for feature extraction, and later these extracted features are fed to the Linear Discriminant Analysis (LDA) for dimensionality reduction. In the last phase, LDA-optimized features were used to train the four machine learning classifiers, one by one, eXtreme Gradient Boosting (XGboost), Naïve Bayes (NB), KNN (K-Nearest Neighbor), and Random Forest (RF). The proposed work has been evaluated on the complete Fake or Real (FoR) dataset for detecting fake audio in different scenarios. The combination of Erlang spectrogram-ResNet11-LDA with XGBoost has achieved a minimum 0% and 1.4% EER for FoR original and FoR 2sec.

Non-parametric Bayesian deep learning approach for whole-body low-dose PET reconstruction and uncertainty assessment.

Positron emission tomography (PET) imaging plays a pivotal role in oncology for the early detection of metastatic tumors and response to therapy assessment due to its high sensitivity compared to anatomical imaging modalities. The balance between image quality and radiation exposure is critical, as reducing the administered dose results in a lower signal-to-noise ratio (SNR) and information loss, which may significantly affect clinical diagnosis. Deep learning (DL) algorithms have recently made significant progress in low-dose (LD) PET reconstruction. Nevertheless, a successful clinical application requires a thorough evaluation of uncertainty to ensure informed clinical judgment. We propose NPB-LDPET, a DL-based non-parametric Bayesian framework for LD PET reconstruction and uncertainty assessment. Our framework utilizes an Adam optimizer with stochastic gradient Langevin dynamics (SGLD) to sample from the underlying posterior distribution. We employed the Ultra-low-dose PET Challenge dataset to assess our framework's performance relative to the Monte Carlo dropout benchmark. We evaluated global reconstruction accuracy utilizing SSIM, PSNR, and NRMSE, local lesion conspicuity using mean absolute error (MAE) and local contrast, and the clinical relevance of uncertainty maps employing correlation between the uncertainty measures and the dose reduction factor (DRF). Our NPB-LDPET reconstruction method exhibits a significantly superior global reconstruction accuracy for various DRFs (paired t-test, , N=10,631). Moreover, we demonstrate a 21% reduction in MAE (573.54 vs. 723.70, paired t-test, , N=28) and an 8.3% improvement in local lesion contrast (2.077 vs. 1.916, paired t-test, , N=28). Furthermore, our framework exhibits a stronger correlation between the predicted uncertainty 95th percentile score and the DRF ( vs. , N=10,631). The proposed framework has the potential to improve clinical decision-making for LD PET imaging by providing a more accurate and informative reconstruction while reducing radiation exposure.

A Joint three-plane Physics-constrained Deep learning based Polynomial Fitting Approach for MR Electrical Properties Tomography.

Magnetic resonance electrical properties tomography can extract the electrical properties of in-vivo tissue. To estimate tissue electrical properties, various reconstruction algorithms have been proposed. However, physics-based reconstructions are prone to various artifacts such as noise amplification and boundary artifact. Deep learning-based approaches are robust to these artifacts but need extensive training datasets and suffer from generalization to unseen data. To address these issues, we introduce a joint three-plane physics-constrained deep learning framework for polynomial fitting MR-EPT by merging physics-based weighted polynomial fitting with deep learning. Within this framework, deep learning is used to discern the optimal polynomial fitting weights for a physics based polynomial fitting reconstruction on the complex B1+ data. For the prediction of optimal fitting coefficients, three neural networks were separately trained on simulated heterogeneous brain models to predict optimal polynomial weighting parameters in three orthogonal planes. Then, the network weights were jointly optimized to estimate the polynomial weights in each plane for a combined conductivity reconstruction. Based on this physics-constrained deep learning approach, we achieved an improvement of conductivity estimation accuracy in comparison to a single plane estimation and a reduction of computational load. The results demonstrate that the proposed method based on 3D data exhibits superior performance in comparison to conventional polynomial fitting methods in terms of capturing anatomical detail and homogeneity. Crucially, in-vivo application of the proposed method showed that the method generalizes well to in-vivo data, without introducing significant errors or artifacts. This generalization makes the presented method a promising candidate for use in clinical applications.

A Deep Learning Approach for Air Pollution Classification Using InceptionV3 with Transfer Learning

A Deep Learning Approach for Air Pollution Classification Using InceptionV3 with Transfer Learning

A deep learning approach for early prediction of breast cancer neoadjuvant chemotherapy response on multistage bimodal ultrasound images.

Neoadjuvant chemotherapy (NAC) is a systemic and systematic chemotherapy regimen for breast cancer patients before surgery. However, NAC is not effective for everyone, and the process is excruciating. Therefore, accurate early prediction of the efficacy of NAC is essential for the clinical diagnosis and treatment of patients. In this study, a novel convolutional neural network model with bimodal layer-wise feature fusion module (BLFFM) and temporal hybrid attention module (THAM) is proposed, which uses multistage bimodal ultrasound images as input for early prediction of the efficacy of neoadjuvant chemotherapy in locally advanced breast cancer (LABC) patients. The BLFFM can effectively mine the highly complex correlation and complementary feature information between gray-scale ultrasound (GUS) and color Doppler blood flow imaging (CDFI). The THAM is able to focus on key features of lesion progression before and after one cycle of NAC. The GUS and CDFI videos of 101 patients collected from cooperative medical institutions were preprocessed to obtain 3000 sets of multistage bimodal ultrasound image combinations for experiments. The experimental results show that the proposed model is effective and outperforms the compared models. The code will be published on the https://github.com/jinzhuwei/BLTA-CNN .

A Deep Ensemble Learning Approach for Automatic AD Detection

A Deep Ensemble Learning Approach for Automatic AD Detection

Segmentation of ADPKD Computed Tomography Images with Deep Learning Approach for Predicting Total Kidney Volume

Segmentation of ADPKD Computed Tomography Images with Deep Learning Approach for Predicting Total Kidney Volume

Coati optimization algorithm for brain tumor identification based on MRI with utilizing phase-aware composite deep neural network

ABSTRACT Brain tumors can cause difficulties in normal brain function and are capable of developing in various regions of the brain. Malignant tumours can develop quickly, pass through neighboring tissues, and extend to further brain regions or the central nervous system. In contrast, healthy tumors typically develop slowly and do not invade surrounding tissues. Individuals frequently struggle with sensory abnormalities, motor deficiencies affecting coordination, and cognitive impairments affecting memory and focus. In this research, Utilizing Phase-aware Composite Deep Neural Network Optimized with Coati Optimized Algorithm for Brain Tumor Identification Based on Magnetic resonance imaging (PACDNN-COA-BTI-MRI) is proposed. First, input images are taken from the brain tumour Dataset. To execute this, the input image is pre-processed using Multivariate Fast Iterative Filtering (MFIF) and it reduces the occurrence of over-fitting from the collected dataset; then feature extraction using Self-Supervised Nonlinear Transform (SSNT) to extract essential features like model, shape, and intensity. Then, the proposed PACDNN-COA-BTI-MRI is implemented in Matlab and the performance metrics Recall, Accuracy, F1-Score, Precision Specificity and ROC are analysed. Performance of the PACDNN-COA-BTI-MRI approach attains 16.7%, 20.6% and 30.5% higher accuracy; 19.9%, 22.2% and 30.1% higher recall and 16.7%, 21.9% and 30.8% higher precision when analysed through existing techniques brain tumor identification using MRI-Based Deep Learning Approach for Efficient Classification of Brain Tumor (MRI-DLA-ECBT), MRI-Based Brain Tumor Detection using Convolutional Deep Learning Methods and Chosen Machine Learning Techniques (MRI-BTD-CDMLT) and MRI-Based Brain Tumor Image Detection using CNN-Based Deep Learning Method (MRI-BTID-CNN) methods, respectively.

Deep learning methods for proteome-scale interaction prediction.

Proteome-scale interaction prediction is essential for understanding protein functions and disease mechanisms. Traditional experimental methods are often limited by scale and complexity, driving the need for computational approaches. Deep learning has emerged as a powerful tool, enabling high-throughput, accurate predictions of protein interactions. This review highlights recent advances in deep learning methods for protein-protein and protein-ligand interaction screening, along with datasets used for model training. Despite the progress with deep learning, challenges such as data quality and validation biases remain. We also discuss the increasing importance of integrating structural information to enhance prediction accuracy and how structure-based deep learning approaches can help overcome current limitations, ultimately advancing biological research and drug discovery.

Advanced fault diagnostics in rotating machinery: A deep learning approach using octave theory of dynamic vibration signals

Advanced fault diagnostics in rotating machinery: A deep learning approach using octave theory of dynamic vibration signals

TSF-MDD: A Deep Learning Approach for Electroencephalography-Based Diagnosis of Major Depressive Disorder with Temporal–Spatial–Frequency Feature Fusion

TSF-MDD: A Deep Learning Approach for Electroencephalography-Based Diagnosis of Major Depressive Disorder with Temporal–Spatial–Frequency Feature Fusion

Deterministic, Stochastic, and Deep Learning Approaches to Understand the Economic Fluctuations in India

Deterministic, Stochastic, and Deep Learning Approaches to Understand the Economic Fluctuations in India

Perfusion estimation from dynamic non-contrast computed tomography using self-supervised learning and a physics-inspired U-net transformer architecture.

Pulmonary perfusion imaging is a key lung health indicator with clinical utility as a diagnostic and treatment planning tool. However, current nuclear medicine modalities face challenges like low spatial resolution and long acquisition times which limit clinical utility to non-emergency settings and often placing extra financial burden on the patient. This study introduces a novel deep learning approach to predict perfusion imaging from non-contrast inhale and exhale computed tomography scans (IE-CT). We developed a U-Net Transformer architecture modified for Siamese IE-CT inputs, integrating insights from physical models and utilizing a self-supervised learning strategy tailored for lung function prediction. We aggregated 523IE-CT images from nine different 4DCT imaging datasets for self-supervised training, aiming to learn a low-dimensional IE-CT feature space by reconstructing image volumes from random data augmentations. Supervised training for perfusion prediction used this feature space and transfer learning on a cohort of 44 patients who had both IE-CT and single-photon emission CT (SPECT/CT) perfusion scans. Testing with random bootstrapping, we estimated the mean and standard deviation of the spatial Spearman correlation between our predictions and the ground truth (SPECT perfusion) to be 0.742 ± 0.037, with a mean median correlation of 0.792 ± 0.036. These results represent a new state-of-the-art accuracy for predicting perfusion imaging from non-contrast CT. Our approach combines low-dimensional feature representations of both inhale and exhale images into a deep learning model, aligning with previous physical modeling methods for characterizing perfusion from IE-CT. This likely contributes to the high spatial correlation with ground truth. With further development, our method could provide faster and more accurate lung function imaging, potentially expanding its clinical applications beyond what is currently possible with nuclear medicine.

Deep learning-based detection of incisal translucency patterns.

The evaluation of incisal translucency in anterior teeth greatly influences esthetic treatment outcomes. This evaluation is mostly subjective and often overlooked among dental professionals. The application of artificial intelligence-based models to detect the incisal translucency of anterior teeth may be of value to dentists in their restorative dental practice, but studies are lacking. The purpose of this study was to assess the accuracy of deep learning models in predicting the translucency patterns of anterior teeth. Approximately 240 Joint Photographic Experts Group (JPEG) images of anterior teeth from participants over 18 years were collected using a smartphone. These images were resized to 224×224 pixels and classified by the presence or absence of translucency. Augmentation techniques enhanced the training dataset, and a 3-model deep learning approach was used: YOLOv5 detected central incisors, Vision Transformers (ViT) identified translucency, and U-Net segmented the translucent areas. The images were split 80 to 20 for training and testing, with performance evaluated using accuracy, precision, recall, F1 score, confusion matrix, and dice scores. YOLOv5 achieved a precision of 1.00 at a confidence threshold of 0.910. The ViT system showed an accuracy of 91.66%, with 58 of 64 images predicting correctly with an F1 score of 94.83%. U-Net segmentation after training with annotated images achieved an accuracy of 91% with a dice score of 0.948. The integration of YOLOv5 for detection, ViT for classification, and U-Net for segmentation demonstrates a comprehensive approach to addressing the classification of incisal translucencies. By leveraging the strengths of deep learning models, high accuracy and precision can be achieved in detecting the incisal translucency patterns of anterior teeth.