Neural Computing and Applications, Год журнала: 2025, Номер unknown
Опубликована: Фев. 19, 2025
Язык: Английский
Trends in Pharmacological Sciences, Год журнала: 2023, Номер 44(9), С. 561 - 572
Опубликована: Июль 19, 2023
Disease modeling and target identification are the most crucial initial steps in drug discovery, influence probability of success at every step development. Traditional is a time-consuming process that takes years to decades usually starts an academic setting. Given its advantages analyzing large datasets intricate biological networks, artificial intelligence (AI) playing growing role modern identification. We review recent advances focusing on breakthroughs AI-driven therapeutic exploration. also discuss importance striking balance between novelty confidence selection. An increasing number AI-identified targets being validated through experiments several AI-derived drugs entering clinical trials; we highlight current limitations potential pathways for moving forward.
Язык: Английский
Процитировано
157
JCO Clinical Cancer Informatics, Год журнала: 2023, Номер 7
Опубликована: Июнь 1, 2023
Synthetic data are artificial generated without including any real patient information by an algorithm trained to learn the characteristics of a source set and became widely used accelerate research in life sciences. We aimed (1) apply generative intelligence build synthetic different hematologic neoplasms; (2) develop validation framework assess fidelity privacy preservability; (3) test capability clinical/translational hematology.
Язык: Английский
Процитировано
50
Computers in Biology and Medicine, Год журнала: 2024, Номер 179, С. 108734 - 108734
Опубликована: Июль 3, 2024
Язык: Английский
Процитировано
31
Scientific Data, Год журнала: 2024, Номер 11(1)
Опубликована: Фев. 29, 2024
Abstract Large annotated datasets are required for training deep learning models, but in medical imaging data sharing is often complicated due to ethics, anonymization and protection legislation. Generative AI such as generative adversarial networks (GANs) diffusion can today produce very realistic synthetic images, potentially facilitate sharing. However, order share images it must first be demonstrated that they used different with acceptable performance. Here, we therefore comprehensively evaluate four GANs (progressive GAN, StyleGAN 1–3) a model the task of brain tumor segmentation (using two networks, U-Net Swin transformer). Our results show trained on reach Dice scores 80%–90% when real memorization problem models if original dataset too small. conclusion viable option further work required. The generated shared AIDA hub.
Язык: Английский
Процитировано
24
npj Digital Medicine, Год журнала: 2024, Номер 7(1)
Опубликована: Март 20, 2024
Clinical research relies on high-quality patient data, however, obtaining big data sets is costly and access to existing often hindered by privacy regulatory concerns. Synthetic generation holds the promise of effectively bypassing these boundaries allowing for simplified accessibility prospect synthetic control cohorts. We employed two different methodologies generative artificial intelligence - CTAB-GAN+ normalizing flows (NFlow) synthesize derived from 1606 patients with acute myeloid leukemia, a heterogeneous hematological malignancy, that were treated within four multicenter clinical trials. Both models accurately captured distributions demographic, laboratory, molecular cytogenetic variables, as well outcomes yielding high performance scores regarding fidelity usability both cohorts (n = each). Survival analysis demonstrated close resemblance survival curves between original Inter-variable relationships preserved in univariable outcome enabling explorative our data. Additionally, training sample safeguarded mitigating possible re-identification, which we quantified using Hamming distances. provide not only proof-of-concept multimodal rare diseases, but also full public foster further research.
Язык: Английский
Процитировано
22
Korean Journal of Radiology, Год журнала: 2024, Номер 25(3), С. 224 - 224
Опубликована: Янв. 1, 2024
The emergence of Chat Generative Pre-trained Transformer (ChatGPT), a chatbot developed by OpenAI, has garnered interest in the application generative artificial intelligence (AI) models medical field. This review summarizes different AI and their potential applications field medicine explores evolving landscape Adversarial Networks diffusion since introduction models. These have made valuable contributions to radiology. Furthermore, this also significance synthetic data addressing privacy concerns augmenting diversity quality within domain, addition emphasizing role inversion investigation outlining an approach replicate process. We provide overview Large Language Models, such as GPTs bidirectional encoder representations (BERTs), that focus on prominent representatives discuss recent initiatives involving language-vision radiology, including innovative large language vision assistant for biomedicine (LLaVa-Med), illustrate practical application. comprehensive offers insights into wide-ranging clinical research emphasizes transformative potential.
Язык: Английский
Процитировано
20
Nature, Год журнала: 2025, Номер 637(8045), С. 319 - 326
Опубликована: Янв. 8, 2025
Язык: Английский
Процитировано
20
Computers in Biology and Medicine, Год журнала: 2025, Номер 189, С. 109834 - 109834
Опубликована: Март 1, 2025
This paper presents a comprehensive systematic review of generative models (GANs, VAEs, DMs, and LLMs) used to synthesize various medical data types, including imaging (dermoscopic, mammographic, ultrasound, CT, MRI, X-ray), text, time-series, tabular (EHR). Unlike previous narrowly focused reviews, our study encompasses broad array modalities explores models. Our aim is offer insights into their current future applications in research, particularly the context synthesis applications, generation techniques, evaluation methods, as well providing GitHub repository dynamic resource for ongoing collaboration innovation. search strategy queries databases such Scopus, PubMed, ArXiv, focusing on recent works from January 2021 November 2023, excluding reviews perspectives. period emphasizes advancements beyond GANs, which have been extensively covered reviews. The survey also aspect conditional generation, not similar work. Key contributions include broad, multi-modality scope that identifies cross-modality opportunities unavailable single-modality surveys. While core techniques are transferable, we find methods often lack sufficient integration patient-specific context, clinical knowledge, modality-specific requirements tailored unique characteristics data. Conditional leveraging textual conditioning multimodal remain underexplored but promising directions findings structured around three themes: (1) Synthesis highlighting clinically valid significant gaps using synthetic augmentation, validation evaluation; (2) Generation identifying personalization innovation; (3) Evaluation revealing absence standardized benchmarks, need large-scale validation, importance privacy-aware, relevant frameworks. These emphasize benchmarking comparative studies promote openness collaboration.
Язык: Английский
Процитировано
6
International Journal of Medical Informatics, Год журнала: 2023, Номер 173, С. 105040 - 105040
Опубликована: Март 8, 2023
Clinical decision support systems (CDSSs) can provide various functions and advantages to healthcare delivery. Quality during pregnancy childbirth is of vital importance, machine learning-based CDSSs have shown positive impact on care. This paper aims investigate what has been done in the context care using learning, aspects require attention from future researchers. We conducted a systematic review existing literature following structured process search, selection filtering, data extraction synthesis. 17 research papers were identified topic CDSS development for different learning algorithms. discovered an overall lack explainability proposed models. also observed experimentation, external validation discussion around culture, ethnicity race source data, with most studies single centre or country, awareness applicability generalisability regarding populations. Finally, we found gap between practices implementation, user testing. Machine are still under-explored Despite open problems that remain, few tested reported effects, reinforcing potential such improve clinical practice. encourage researchers take into consideration order their work translate use.
Язык: Английский
Процитировано
34
International Journal of Medical Informatics, Год журнала: 2024, Номер 185, С. 105413 - 105413
Опубликована: Март 12, 2024
Ensuring safe adoption of AI tools in healthcare hinges on access to sufficient data for training, testing and validation. Synthetic has been suggested response privacy concerns regulatory requirements can be created by training a generator real produce dataset with similar statistical properties. Competing metrics differing taxonomies quality evaluation have proposed, resulting complex landscape. Optimising entails balancing considerations that make the fit use, yet relevant dimensions are left out existing frameworks. We performed comprehensive literature review use synthetic within scope tabular using deep generative methods. Based this collective team experiences, we developed conceptual framework assurance. The applicability was benchmarked against practical case from Dutch National Cancer Registry. present assurance applications aligns diverging taxonomies, expands common include Fairness Carbon footprint, proposes stages necessary support real-life applications. Building trust increasing transparency reducing safety risk will accelerate development uptake trustworthy benefit patients. Despite growing emphasis algorithmic fairness carbon these were scarce review. overwhelming focus similarity distance while sequential logic detection scarce. A consensus-backed includes all provide responsible data. As choice appropriate highly context dependent, further research is needed validation studies guide metric choices technical standards.
Язык: Английский
Процитировано
15