A review of deep learning in dentistry

Neurocomputing, Journal Year: 2023, Volume and Issue: 554, P. 126629 - 126629

Published: July 27, 2023

Oral diseases have a significant impact on human health, often going unnoticed in their early stages. Deep learning, promising field artificial intelligence, has shown remarkable success various domains, especially dentistry. This paper aims to provide an overview of recent research deep learning applications dentistry, with focus dental imaging. algorithms perform well difficult tasks such as image segmentation and recognition, enabling accurate identification oral conditions abnormalities. Integration other health data offers holistic understanding the relationship between systemic health. However, there are still many challenges that need be addressed.

Language: Английский

---

Dual-3DM³AD: Mixed Transformer Based Semantic Segmentation and Triplet Pre-Processing for Early Multi-Class Alzheimer’s Diagnosis

IEEE Transactions on Neural Systems and Rehabilitation Engineering, Journal Year: 2024, Volume and Issue: 32, P. 696 - 707

Published: Jan. 1, 2024

Alzheimer's Disease (AD) is a widespread, chronic, irreversible, and degenerative condition, its early detection during the prodromal stage of utmost importance. Typically, AD studies rely on single data modalities, such as MRI or PET, for making predictions. Nevertheless, combining metabolic structural can offer comprehensive perspective staging analysis. To address this goal, paper introduces an innovative multi-modal fusion-based approach named Dual-3DM ³ -AD. This model proposed accurate diagnosis by considering both PET image scans. Initially, we pre-process images in terms noise reduction, skull stripping 3D conversion using Quaternion Non-local Means Denoising Algorithm (QNLM), Morphology function Block Divider Model (BDM), respectively, which enhances quality. Furthermore, have adapted Mixed-transformer with Furthered U-Net performing semantic segmentation minimizing complexity. -AD consisted multi-scale feature extraction module extracting appropriate features from segmented images. The extracted are then aggregated Densely Connected Feature Aggregator Module (DCFAM) to utilize features. Finally, multi-head attention mechanism dimensionality softmax layer applied multi-class diagnosis. compared several baseline approaches help performance metrics. final results unveil that work achieves 98% accuracy, 97.8% sensitivity, 97.5% specificity, 98.2% f-measure, better ROC curves, outperforms other existing models multiclass

Language: Английский

---

Boosting skin cancer diagnosis accuracy with ensemble approach

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Jan. 8, 2025

Skin cancer is common and deadly, hence a correct diagnosis at an early age essential. Effective therapy depends on precise classification of the several skin forms, each with special traits. Because dermoscopy other sophisticated imaging methods produce detailed lesion images, detection has been enhanced. It's still difficult to analyze images differentiate benign from malignant tumors, though. Better predictive modeling are needed since diagnostic procedures used now frequently inaccurate inconsistent results. In dermatology, Machine learning (ML) models becoming essential for automatic lesions image data. With ensemble model, which mix ML approaches take use their advantages lessen disadvantages, this work seeks improve predictions. We introduce new method, Max Voting optimization classification. On HAM10000 ISIC 2018 datasets, we trained assessed three distinct models: Random Forest (RF), Multi-layer Perceptron Neural Network (MLPN), Support Vector (SVM). Overall performance was increased by combined predictions made technique. Moreover, feature vectors that were optimally produced data Genetic Algorithm (GA) given models. demonstrate method greatly improves performance, reaching accuracy 94.70% producing best results F1-measure, recall, precision. The most dependable robust approach turned out be Voting, combines benefits numerous pre-trained provide efficient classifying lesions.

Language: Английский

---

Breast cancer classification based on hybrid CNN with LSTM model

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Feb. 5, 2025

Breast cancer (BC) is a global problem, largely due to shortage of knowledge and early detection. The speed-up process detection classification crucial for effective treatment. Medical image analysis methods computer-aided diagnosis can enhance this process, providing training assistance less experienced clinicians. Deep Learning (DL) models play great role in accurately detecting classifying the huge dataset, especially when dealing with large medical images. This paper presents novel hybrid model DL combined Convolutional Neural Network (CNN) Long Short-Term Memory (LSTM) binary breast on two datasets available at Kaggle repository. CNNs extract mammographic features, including spatial hierarchies malignancy patterns, whereas LSTM networks characterize sequential dependencies temporal interactions. Our method combines these structures improve accuracy resilience. We compared proposed other models, such as CNN, LSTM, Gated Recurrent Units (GRUs), VGG-16, RESNET-50. CNN-LSTM achieved superior performance accuracies 99.17% 99.90% respective datasets. uses prediction evaluation metrics accuracy, sensitivity, specificity, F-score, AUC curve. results showed that our classifiers others second dataset.

Language: Английский

---

Recent Advancements and Future Prospects in Active Deep Learning for Medical Image Segmentation and Classification

IEEE Access, Journal Year: 2023, Volume and Issue: 11, P. 113623 - 113652

Published: Jan. 1, 2023

Medical images are helpful for the diagnosis, treatment, and evaluation of diseases. Precise medical image segmentation improves diagnosis decision-making, aiding intelligent services better disease management recovery. Due to unique nature images, algorithms based on deep learning face problems such as sample imbalance, edge blur, false positives, negatives. In view these problems, researchers primarily improve network structure but rarely from unstructured aspect. The paper tackles challenges, accentuating limitations convolutional neural network-based methods proposing solutions reduce annotation costs, particularly in complex introduces improvement strategies solve Additionally, article latest learning-based applications analysis, covering segmentation, acquisition, enhancement, registration, classification. Moreover, provides an overview four cutting-edge models, namely (CNN), belief (DBN), stacked autoencoder (SAE), recurrent (RNN). study selection involved searching benchmark academic databases, collecting relevant literature appropriate indicator emphasizing DL-based classification approaches, evaluating performance metrics. research highlights clinicians' scholars' obstacles developing efficient accurate malignancy prognostic framework state-of-the-art deep-learning algorithms. Furthermore, future perspectives explored overcome challenges advance field analysis.

Language: Английский

---

Deep Learning Approaches for Medical Image Analysis and Diagnosis

Cureus, Journal Year: 2024, Volume and Issue: unknown

Published: May 2, 2024

In addition to enhancing diagnostic accuracy, deep learning techniques offer the potential streamline workflows, reduce interpretation time, and ultimately improve patient outcomes. The scalability adaptability of algorithms enable their deployment across diverse clinical settings, ranging from radiology departments point-of-care facilities. Furthermore, ongoing research efforts focus on addressing challenges data heterogeneity, model interpretability, regulatory compliance, paving way for seamless integration solutions into routine practice. As field continues evolve, collaborations between clinicians, scientists, industry stakeholders will be paramount in harnessing full advancing medical image analysis diagnosis. with other technologies, including natural language processing computer vision, may foster multimodal decision support systems care. future diagnosis is promising. With each success advancement, this technology getting closer being leveraged purposes. Beyond analysis, care pathways like imaging, imaging genomics, intelligent operating rooms or intensive units can benefit models.

Language: Английский

---

Integrating artificial intelligence in industry 4.0: insights, challenges, and future prospects–a literature review

Annals of Operations Research, Journal Year: 2024, Volume and Issue: unknown

Published: May 8, 2024

Language: Английский

---

Estimation of Muscle Forces of Lower Limbs Based on CNN–LSTM Neural Network and Wearable Sensor System

Sensors, Journal Year: 2024, Volume and Issue: 24(3), P. 1032 - 1032

Published: Feb. 5, 2024

Estimation of vivo muscle forces during human motion is important for understanding control mechanisms and joint mechanics. This paper combined the advantages convolutional neural network (CNN) long-short-term memory (LSTM) proposed a novel force estimation method based on CNN–LSTM. A wearable sensor system was also developed to collect angles angular velocities hip, knee, ankle joints in sagittal plane walking, collected kinematic data were used as input model. In this paper, calculated using OpenSim Static Optimization (SO) standard value train Four lower limb muscles left leg, including gluteus maximus (GM), rectus femoris (RF), gastrocnemius (GAST), soleus (SOL), selected studying objects paper. The experiment results showed that compared CNN LSTM, CNN–LSTM performed better under slow (1.2 m/s), medium (1.5 fast walking speeds (1.8 m/s). average correlation coefficients between true estimated values four slow, medium, 0.9801, 0.9829, 0.9809, respectively. had smaller fluctuations different speeds, which indicated model good robustness. external testing generalization. well when object not included training sample. article convenient estimating forces, could provide theoretical assistance quantitative analysis injury. has established relationship signals model; SO calculate OpenSim, it more efficient clinical or engineering applications.

Language: Английский

---

Vision-BioLLM: Large vision language model for visual dialogue in biomedical imagery

Biomedical Signal Processing and Control, Journal Year: 2025, Volume and Issue: 103, P. 107437 - 107437

Published: Jan. 5, 2025

---

Enhancing Prognosis Accuracy for Ischemic Cardiovascular Disease Using K Nearest Neighbor Algorithm: A Robust Approach

IEEE Access, Journal Year: 2023, Volume and Issue: 11, P. 97879 - 97895

Published: Jan. 1, 2023

Ischemic Cardiovascular diseases are one of the deadliest in world. However, mortality rate can be significantly reduced if we detect disease precisely and effectively. Machine Learning (ML) models offer substantial assistance to individuals requiring early treatment detection realm cardiovascular health. In response this critical need, study developed a robust system predict ischemic accurately using ML-based algorithms. The dataset obtained from Kaggle encompasses comprehensive collection over 918 observations, encompassing 12 essential features crucial for predicting disease. contrast, much-existing research relies primarily on datasets comprising only 303 instances UCI repository. Six algorithms, including K Nearest Neighbors (KNN), Random Forest (RF), Logistic Regression (LR), Support Vector (SVM), Gaussian Naïve Bayes (GNB), Decision Trees (DT), trained heart data. effectiveness proposed methodologies is meticulously evaluated benchmarked against cutting-edge techniques, employing range performance criteria. empirical findings manifest that KNN classifier produced optimized results with 91.8% accuracy, 91.4% recall, 91.9% F1 score, 92.5% precision, AUC 90.27%.

Language: Английский

---