A Systematic Review of Synthetic Data Generation Techniques Using Generative AI

Electronics, Journal Year: 2024, Volume and Issue: 13(17), P. 3509 - 3509

Published: Sept. 4, 2024

Synthetic data are increasingly being recognized for their potential to address serious real-world challenges in various domains. They provide innovative solutions combat the scarcity, privacy concerns, and algorithmic biases commonly used machine learning applications. preserve all underlying patterns behaviors of original dataset while altering actual content. The methods proposed literature generate synthetic vary from large language models (LLMs), which pre-trained on gigantic datasets, generative adversarial networks (GANs) variational autoencoders (VAEs). This study provides a systematic review techniques that can be identify limitations suggest future research areas. findings indicate these technologies specific types, they still have some drawbacks, such as computational requirements, training stability, privacy-preserving measures limit usability. Addressing issues will facilitate broader adoption generation across disciplines, thereby advancing data-driven solutions.

Language: Английский

---

Exploring the Landscape of Explainable Artificial Intelligence (XAI): A Systematic Review of Techniques and Applications

Big Data and Cognitive Computing, Journal Year: 2024, Volume and Issue: 8(11), P. 149 - 149

Published: Oct. 31, 2024

Artificial intelligence (AI) encompasses the development of systems that perform tasks typically requiring human intelligence, such as reasoning and learning. Despite its widespread use, AI often raises trust issues due to opacity decision-making processes. This challenge has led explainable artificial (XAI), which aims enhance user understanding by providing clear explanations decisions paper reviews existing XAI research, focusing on application in healthcare sector, particularly medical medicinal contexts. Our analysis is organized around key properties XAI—understandability, comprehensibility, transparency, interpretability, explainability—providing a comprehensive overview techniques their practical implications.

Language: Английский

---

Machine Learning-Based Process Optimization in Biopolymer Manufacturing: A Review

Polymers, Journal Year: 2024, Volume and Issue: 16(23), P. 3368 - 3368

Published: Nov. 29, 2024

The integration of machine learning (ML) into material manufacturing has driven advancements in optimizing biopolymer production processes. ML techniques, applied across various stages production, enable the analysis complex data generated throughout identifying patterns and insights not easily observed through traditional methods. As sustainable alternatives to petrochemical-based plastics, biopolymers present unique challenges due their reliance on variable bio-based feedstocks processing conditions. This review systematically summarizes current applications techniques aiming provide a comprehensive reference for future research while highlighting potential enhance efficiency, reduce costs, improve product quality. also shows role algorithms, including supervised, unsupervised, deep

Language: Английский

---

The potential of generative AI with prostate-specific membrane antigen (PSMA) PET/CT: challenges and future directions

Medical Review, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 23, 2025

Abstract The diagnosis and prognosis of Prostate cancer (PCa) have undergone a significant transformation with the advent prostate-specific membrane antigen (PSMA)-targeted positron emission tomography (PET) imaging. PSMA-PET imaging has demonstrated superior performance compared to conventional methods by detecting PCa, its biochemical recurrence, sites metastasis higher sensitivity specificity. That now intersects rapid advances in artificial intelligence (AI) – including emergence generative AI. However, there are unique clinical challenges associated that still need be addressed ensure continued widespread integration into care research trials. Some those very wide dynamic range lesion uptake, benign uptake organs may adjacent disease, insufficient large datasets for training AI models, as well artifacts images. Generative e.g., adversarial networks, variational autoencoders, diffusion language models played crucial roles overcoming many such across various modalities, PET, computed tomography, magnetic resonance imaging, ultrasound, etc. In this review article, we delve potential role enhancing robustness utilization image analysis, drawing insights from existing literature while also exploring current limitations future directions domain.

Language: Английский

---

Advances in 3D fusion of multimodal medical images: 3D reconstruction of bone, muscle, and ligament structures under load from radiographs and magnetic resonance imaging

Research on Biomedical Engineering, Journal Year: 2025, Volume and Issue: 41(1)

Published: Feb. 14, 2025

Language: Английский

---

Generating Pseudo-Subtracted Image in Dual-Energy Contrast-Enhanced Spectral Mammography Using Transfer Learning

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: March 26, 2025

Background: Dual-energy contrast-enhanced spectral mammography (CESM) enhances breast cancer detection but increases radiation exposure, especially for high-risk patients like BRCA1 mutation carriers. Additionally, the dual-energy acquisition process can be time-consuming. This study uses deep learning to convert low-energy images into subtracted images, reducing and contrast-related risks, while also addressing time consumption challenge of traditional CESM procedure. Methods: The utilized Categorized Digital Database Low-energy Subtracted Contrast-Enhanced Spectral Mammography Images (CDD-CESM), which contains 7600 image pairs after augmentation. dataset was divided 70% training 30% testing. CycleGAN's performance evaluated compared against U-Net, Pix2Pix, ResNet18. Key metrics comparison included Structural Similarity Index Peak Signal-to-Noise Ratio. models were tested their ability generate high-quality without need paired data. Results: CycleGAN outperformed ResNet18 in generating pseudo-subtracted images. SSIM score 0.961, close that real indicates successfully preserves structural details. achieved this at a lower computational cost Conclusions: effectively generates from data, presenting viable alternative imaging. method has potential reduce additional imaging, minimize simplify imaging procedures. high highlights maintain strong similarities generated making it promising tool detecting lesions mammography.

Language: Английский

---

Comprehensive Review on the Impact of Artificial Intelligence on Diagnosis and Personalized Treatment in Nuclear Medicine

SHS Web of Conferences, Journal Year: 2025, Volume and Issue: 214, P. 01006 - 01006

Published: Jan. 1, 2025

Artificial intelligence (AI) continues to advance nuclear medicine in all areas, including treatment planning, resource allocation, and precision. The imaging techniques powered by AI enable faster more accurate diagnosis of diseases machine learning models improve individual-specific dosimetry. Additionally, increases operational efficiency, reduces costs, lower radiation exposure for patients. Despite these improvements, issues such as ethical concerns, bias data, clinical integration difficulties still exist. This review paper discusses the role changing practice, emphasizing pros cons, anticipated future. As field proves its further value, multidisciplinary collaborations are invited help ensure AI’s value future treatment.

Language: Английский

---

Supervised vs. Unsupervised GAN for Pseudo-CT Synthesis in Brain MR-Guided Radiotherapy

Published: April 4, 2025

Purpose Radiotherapy is a crucial treatment for brain tumor malignancies. To address the limitations of CT-based planning, recent research has explored MR-only radiotherapy, requiring precise MR-to-CT synthesis. This study compares two deep learning approaches, supervised (Pix2Pix) and unsupervised (CycleGAN), generating pseudo-CT (pCT) images from T1- T2-weighted MR sequences. Methods Materials: 3270 paired T1 T2 weighted MRI are collected registered with corresponding CT images. After preprocessing pCT generative model was trained using "pix2pix" model, an network (CycleGan), also purpose comparing quality against pix2pix. assess differences between reference images, three key metrics (SSIM, PSNR MAE) used. Results The average SSIM, MAE pix2pix on 0.964 ± 0.03, 32.812 5.21 79.681 9.52 HU respectively. Statistical analysis revealed that Pix2Pix significantly outperformed CycleGAN in high-fidelity (p < 0.05). There no notable difference effectiveness T1-weighted versus > Conclusion Both methods demonstrated capability to generate accurate conventional While like achieve higher accuracy, approaches such as offer greater flexibility by eliminating need training data, making them suitable applications where data unavailable.

Language: Английский

---

Advancing pancreatic cancer management: the role of artificial intelligence in diagnosis and therapy

Beni-Suef University Journal of Basic and Applied Sciences, Journal Year: 2025, Volume and Issue: 14(1)

Published: April 7, 2025

Abstract Background Pancreatic cancer is the deadliest form of with a low survival rate due to its late diagnosis. Hence, early detection and swift intervention are very crucial for management. However, current diagnostic markers lack sufficient precision, effectiveness treatment options remains imprecise, emphasizing need more advanced approaches. Main body Artificial intelligence (AI) technology enables rapid high-risk groups pancreatic using various techniques such as medical imaging, pathological examination, biomarkers, other methods, facilitating cancer. Simultaneously, AI algorithms may also be used forecast duration survival, likelihood recurrence, metastasis, response treatment, all which can impact prognosis. Moreover, applied in handling cases oncology departments, particular, creating computer-assisted systems. Conclusion The end-to-end application management calls multidisciplinary collaboration among doctors, laboratory scientists, data analysts, engineers. Despite limitations, powerful computational capabilities will soon combating health conditions.

Language: Английский

---

Hierarchical multi-scale Mamba generative adversarial network for multi-modal medical image synthesis

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127451 - 127451

Published: April 1, 2025

Language: Английский

---