Drug Discovery Today, Год журнала: 2023, Номер 29(1), С. 103823 - 103823
Опубликована: Ноя. 8, 2023
Язык: Английский
Advanced Drug Delivery Reviews, Год журнала: 2022, Номер 182, С. 114098 - 114098
Опубликована: Янв. 5, 2022
Язык: Английский
Процитировано
132
Advanced Materials, Год журнала: 2024, Номер 36(34)
Опубликована: Март 8, 2024
In recent years, there has been widespread adoption of machine learning (ML) technologies to unravel intricate relationships among diverse parameters in various additive manufacturing (AM) techniques. These ML models excel at recognizing complex patterns from extensive, well-curated datasets, thereby unveiling latent knowledge crucial for informed decision-making during the AM process. The collaborative synergy between and holds potential revolutionize design production AM-printed parts. This review delves into challenges opportunities emerging intersection these two dynamic fields. It provides a comprehensive analysis publication landscape ML-related research field AM, explores common applications (such as quality control, process optimization, microstructure analysis, material formulation), concludes by presenting an outlook that underscores utilization advanced models, development sensors, AM-related Notably, garnered increased attention due its superior performance across applications. is envisioned integration processes will significantly enhance 3D printing capabilities areas.
Язык: Английский
Процитировано
77
Advanced Drug Delivery Reviews, Год журнала: 2023, Номер 202, С. 115108 - 115108
Опубликована: Сен. 27, 2023
Язык: Английский
Процитировано
56
International Journal of Pharmaceutics, Год журнала: 2023, Номер 633, С. 122628 - 122628
Опубликована: Янв. 20, 2023
Three-dimensional (3D) printing is drastically redefining medicine production, offering digital precision and personalized design opportunities. One emerging 3D technology selective laser sintering (SLS), which garnering attention for its high precision, compatibility with a wide range of pharmaceutical materials, including low-solubility compounds. However, the full potential SLS medicines yet to be realized, requiring expertise considerable time-consuming resource-intensive trial-and-error research. Machine learning (ML), subset artificial intelligence, an in silico tool that accomplishing remarkable breakthroughs several sectors ability make highly accurate predictions. Therefore, present study harnessed ML predict printability formulations. Using dataset 170 formulations from 78 models were developed inputs included formulation composition characterization data retrieved Fourier-transformed infrared spectroscopy (FT-IR), X-ray powder diffraction (XRPD) differential scanning calorimetry (DSC). Multiple explored, supervised unsupervised approaches. The results revealed can achieve accuracies, by using leading maximum F1 score 81.9%. FT-IR, XRPD DSC as resulted 84.2%, 81.3%, 80.1%, respectively. A subsequent pipeline was built combine predictions into one consensus model, where found further increase 88.9%. it determined first time benefit multi-modal data, combining numeric, spectral, thermogram data. lays groundwork leveraging existing developing high-performing computational accelerate development.
Язык: Английский
Процитировано
49
International Journal of Pharmaceutics, Год журнала: 2024, Номер 652, С. 123741 - 123741
Опубликована: Янв. 3, 2024
Artificial intelligence (AI) is a revolutionary technology that finding wide application across numerous sectors. Large language models (LLMs) are an emerging subset of AI and have been developed to communicate using human languages. At their core, LLMs trained with vast amounts information extracted from the internet, including text images. Their ability create human-like, expert in almost any subject means they increasingly being used as aid presentation, particularly scientific writing. However, we wondered whether could go further, generating original research preparing results for publication. We tasked GPT-4, LLM, write pharmaceutics manuscript, on topic itself novel. It was able conceive hypothesis, define experimental protocol, produce photo-realistic images printlets, generate believable analytical data range instruments convincing publication-ready manuscript evidence critical interpretation. The model achieved all this less than 1h. Moreover, generated were multi-modal nature, thermal analyses, vibrational spectroscopy dissolution testing, demonstrating multi-disciplinary expertise LLM. One area which failed, however, referencing literature. Since appeared though, suggest certainly play role but input, interpretation validation. discuss potential benefits current bottlenecks realising ambition here.
Язык: Английский
Процитировано
28
Applied Sciences, Год журнала: 2024, Номер 14(4), С. 1471 - 1471
Опубликована: Фев. 11, 2024
The Fourth Industrial Revolution combined with the advent of artificial intelligence brought significant changes to humans’ daily lives. Extended research in field has aided both documenting and presenting these changes, giving a more general picture this new era. This work reviews application scientific literature on presence machine vision it each sector which contributed, determining exact extent its influence. Accordingly, an attempt is made present overview use Fifth identify between two consequent periods. uses PRISMA methodology follows form Scoping Review using sources from Scopus Google Scholar. Most publications reveal emergence almost every human life influence performance results. Undoubtedly, review highlights great offer many sectors, establishing searching for ways it. It also proven that systems can help industries gain competitive advantage terms better product quality, higher customer satisfaction, improved productivity.
Язык: Английский
Процитировано
28
Applied Materials Today, Год журнала: 2024, Номер 36, С. 102061 - 102061
Опубликована: Янв. 16, 2024
Formulation development is a critical step in the of medicines. The process requires human creativity, ingenuity and in-depth knowledge formulation processing optimization, which can be time-consuming. Herein, we tested ability artificial intelligence (AI) to create de novo formulations for three-dimensional (3D) printing. Specifically, conditional generative adversarial networks (cGANs), are models known their were trained on dataset consisting 1437 fused deposition modelling (FDM) printed that extracted from both literature in-house data. In total, 27 different cGANs architectures explored with varying learning rate, batch size number hidden layers parameters generate 270 formulations. After comparison between characteristics AI-generated human-generated formulations, it was discovered medium rate (10−4) could strike balance generating novel realistic. Four these fabricated using an FDM printer, first successfully printed. Our study represents milestone, highlighting capacity AI undertake creative tasks its potential revolutionize drug process.
Язык: Английский
Процитировано
19
Results in Engineering, Год журнала: 2024, Номер 25, С. 103676 - 103676
Опубликована: Дек. 8, 2024
Язык: Английский
Процитировано
18
Journal of Controlled Release, Год журнала: 2022, Номер 352, С. 1071 - 1092
Опубликована: Ноя. 18, 2022
Язык: Английский
Процитировано
52
International Journal of Pharmaceutics, Год журнала: 2023, Номер 639, С. 122926 - 122926
Опубликована: Апрель 7, 2023
Achieving carbon neutrality is seen as an important goal in order to mitigate the effects of climate change, dioxide a major greenhouse gas that contributes global warming. Many countries, cities and organizations have set targets become neutral. The pharmaceutical sector no exception, being contributor emissions (emitting approximately 55% more than automotive for instance) hence need strategies reduce its environmental impact. Three-dimensional (3D) printing advanced fabrication technology has potential replace traditional manufacturing tools. Being new technology, impact 3D printed medicines not been investigated, which barrier uptake by industry. Here, energy consumption (and emission) printers considered, focusing on technologies successfully demonstrated produce solid dosage forms. 6 benchtop was measured during standby mode printing. On standby, ranged from 0.03 0.17 kWh. required producing 10 printlets 0.06 3.08 kWh, with using high temperatures consuming energy. Carbon between 11.60 112.16 g CO2 (eq) per printlets, comparable tableting. Further analyses revealed decreasing temperature found demand considerably, suggesting developing formulations are printable at lower can emissions. study delivers key initial insights into potentially transformative provides encouraging results demonstrating deliver quality without environmentally detrimental.
Язык: Английский
Процитировано
40