Li Hanxiang, E Alkahtani Manal, W Basit Abdul, Elbadawi Moe, Gaisford Simon
UCL School of Pharmacy, University College London, 29-39 Brunswick Square, London WC1N 1AX, UK.
UCL School of Pharmacy, University College London, 29-39 Brunswick Square, London WC1N 1AX, UK; Department of Pharmaceutics, College of Pharmacy, Prince Sattam bin Abdulaziz University, Alkharj 11942, Saudi Arabia.
Int J Pharm. 2023 Oct 28:123561. doi: 10.1016/j.ijpharm.2023.123561.
3D Printing (3DP) of pharmaceuticals could drastically transform the manufacturing of medicines and facilitate the widespread availability of personalised healthcare. However, with increasing awareness of the environmental damage of manufacturing, 3DP must be eco-friendly, especially when it comes to carbon emissions. This study investigated the environmental effects of pharmaceutical 3DP. Using Design of Experiments (DoE) and Machine Learning (ML), we looked at energy use in pharmaceutical Fused Deposition Modeling (FDM). From 136 experimental runs across four common dosage forms, we identified several key parameters that contributed to energy consumption, and consequently CO emission. These parameters, identified by both DoE and ML, were the number of objects printed, build plate temperature, nozzle temperature, and layer height. Our analysis revealed that minimizing trial-and-error by being more efficient in R&D and reducing the build plate temperature can significantly decrease CO emissions. Furthermore, we demonstrated that only the ML pipeline could accurately predict CO emissions, suggesting ML could be a powerful tool in in the development of more sustainable manufacturing processes. The models were validated experimentally on new dosage forms of varying geometric complexities and were found to maintain high accuracy across all three dosage forms. The study underscores the potential of merging sustainability and digitalization in the pharmaceutical sector, aligning with the principles of Industry 5.0. It highlights the comparable learning traits between DoE and ML, indicating a promising pathway for wider adoption of ML in pharmaceutical manufacturing. Through focused efforts to reduce wasteful practices and optimize printing parameters, we can pave the way for a more environmentally sustainable future in pharmaceutical 3DP.
药物的3D打印（3DP）可能会彻底改变药品的制造方式，并促进个性化医疗的广泛普及。然而，随着人们对制造过程中环境破坏的认识不断提高，3DP必须是环保的，尤其是在碳排放方面。本研究调查了药物3DP的环境影响。通过实验设计（DoE）和机器学习（ML），我们研究了药物熔融沉积建模（FDM）中的能源使用情况。从四种常见剂型的136次实验运行中，我们确定了几个导致能源消耗以及二氧化碳排放的关键参数。这些由DoE和ML共同确定的参数包括打印物体的数量、构建平台温度、喷嘴温度和层高。我们的分析表明，通过在研发中提高效率并降低构建平台温度来减少试错次数，可以显著降低二氧化碳排放。此外，我们证明只有ML管道能够准确预测二氧化碳排放，这表明ML可能是开发更可持续制造工艺的有力工具。这些模型在具有不同几何复杂度的新剂型上进行了实验验证，发现在所有三种剂型上都能保持高精度。该研究强调了制药行业将可持续性与数字化相结合的潜力，符合工业5.0的原则。它突出了DoE和ML之间可比的学习特性，表明ML在制药制造中更广泛应用的前景广阔。通过集中精力减少浪费行为并优化打印参数，我们可以为药物3DP的更环境可持续的未来铺平道路。
