药学学报

Learning style	Classification	Application
Supervised learning	Multiple linear regression, decision-making tree, random forest, artificial neural network	Classify gene-disease association, pattern recognition of pharmaceutical raw ingredients^[26]
Unsupervised learning	PCA, isometric feature mapping, K-means	Classify P-glycoprotein inhibitors^[27]
Semi-supervised learning	Self-training, transductive learning, generative models	Drugs based on the relocation potential of disease networks^[28]
Reinforcement learning	Standard Markov decision process	Drug reuse based on drug target protein information to construct drug network^[29]
Deep learning	Convolutional neural network, deep belief networks	Drug discovery^[30], prediction of drug prescription in vitro^[31]

Learning style	Classification	Application
Supervised learning	Multiple linear regression, decision-making tree, random forest, artificial neural network	Classify gene-disease association, pattern recognition of pharmaceutical raw ingredients^[26]
Unsupervised learning	PCA, isometric feature mapping, K-means	Classify P-glycoprotein inhibitors^[27]
Semi-supervised learning	Self-training, transductive learning, generative models	Drugs based on the relocation potential of disease networks^[28]
Reinforcement learning	Standard Markov decision process	Drug reuse based on drug target protein information to construct drug network^[29]
Deep learning	Convolutional neural network, deep belief networks	Drug discovery^[30], prediction of drug prescription in vitro^[31]

人工智能在3D打印药物的研究进展

PDF下载

韩晓璐 ¹ , 王珊珊 ¹ , 彭静 ² , 洪晓轩 ¹ , 王增明 ¹^,^* , 王娜 ²^,^* , 郑爱萍 ¹

药学学报 | 综述 2023,58(6): 1577-1585

收起

药学学报 |综述 2023 , 58 (6) : 1577 -1585

人工智能在3D打印药物的研究进展

全屏

[Author(id=1198960145213522029, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1198960145431625847, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960145213522029, language=EN, stringName=Xiao-lu HAN, firstName=Xiao-lu, middleName=null, lastName=HAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹, address=1. Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, Academy of Military Sciences, Beijing 100850, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1198960145591009411, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960145213522029, language=CN, stringName=韩晓璐, firstName=晓璐, middleName=null, lastName=韩, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹, address=1.军事科学院军事医学研究院毒物药物研究所, 北京 100850, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1198960144945086547, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, xref=null, ext=[AuthorCompanyExt(id=1198960144949280852, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960144945086547, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, Academy of Military Sciences, Beijing 100850, China), AuthorCompanyExt(id=1198960144957669461, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960144945086547, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.军事科学院军事医学研究院毒物药物研究所, 北京 100850)])]), Author(id=1198960145704255629, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1198960145821696154, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960145704255629, language=EN, stringName=Shan-shan WANG, firstName=Shan-shan, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹, address=1. Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, Academy of Military Sciences, Beijing 100850, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1198960145955913894, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960145704255629, language=CN, stringName=王珊珊, firstName=珊珊, middleName=null, lastName=王, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹, address=1.军事科学院军事医学研究院毒物药物研究所, 北京 100850, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1198960144945086547, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, xref=null, ext=[AuthorCompanyExt(id=1198960144949280852, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960144945086547, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, Academy of Military Sciences, Beijing 100850, China), AuthorCompanyExt(id=1198960144957669461, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960144945086547, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.军事科学院军事医学研究院毒物药物研究所, 北京 100850)])]), Author(id=1198960146077548722, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1198960146169823416, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960146077548722, language=EN, stringName=Jing PENG, firstName=Jing, middleName=null, lastName=PENG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=², address=2. National Biomedical Analysis Center, Beijing 100850, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1198960146308235463, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960146077548722, language=CN, stringName=彭静, firstName=静, middleName=null, lastName=彭, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=², address=2.国家生物医学分析中心, 北京 100850, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1198960145049944158, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, xref=null, ext=[AuthorCompanyExt(id=1198960145058332766, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960145049944158, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2. National Biomedical Analysis Center, Beijing 100850, China), AuthorCompanyExt(id=1198960145066721375, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960145049944158, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.国家生物医学分析中心, 北京 100850)])]), Author(id=1198960146421481678, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1198960146559893720, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960146421481678, language=EN, stringName=Xiao-xuan HONG, firstName=Xiao-xuan, middleName=null, lastName=HONG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹, address=1. Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, Academy of Military Sciences, Beijing 100850, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1198960146668945633, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960146421481678, language=CN, stringName=洪晓轩, firstName=晓轩, middleName=null, lastName=洪, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹, address=1.军事科学院军事医学研究院毒物药物研究所, 北京 100850, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1198960144945086547, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, xref=null, ext=[AuthorCompanyExt(id=1198960144949280852, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960144945086547, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, Academy of Military Sciences, Beijing 100850, China), AuthorCompanyExt(id=1198960144957669461, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960144945086547, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.军事科学院军事医学研究院毒物药物研究所, 北京 100850)])]), Author(id=1198960146819940587, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=wangzm.1986@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1198960146970935539, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960146819940587, language=EN, stringName=Zeng-ming WANG, firstName=Zeng-ming, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹^,^*, address=1. Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, Academy of Military Sciences, Beijing 100850, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1198960147096764669, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960146819940587, language=CN, stringName=王增明, firstName=增明, middleName=null, lastName=王, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹^,^*, address=1.军事科学院军事医学研究院毒物药物研究所, 北京 100850, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1198960144945086547, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, xref=null, ext=[AuthorCompanyExt(id=1198960144949280852, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960144945086547, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, Academy of Military Sciences, Beijing 100850, China), AuthorCompanyExt(id=1198960144957669461, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960144945086547, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.军事科学院军事医学研究院毒物药物研究所, 北京 100850)])]), Author(id=1198960147302285581, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, orderNo=5, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=wn@proteomics.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1198960147499417884, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960147302285581, language=EN, stringName=Na WANG, firstName=Na, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=²^,^*, address=2. National Biomedical Analysis Center, Beijing 100850, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1198960147650412839, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960147302285581, language=CN, stringName=王娜, firstName=娜, middleName=null, lastName=王, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=²^,^*, address=2.国家生物医学分析中心, 北京 100850, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1198960145049944158, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, xref=null, ext=[AuthorCompanyExt(id=1198960145058332766, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960145049944158, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2. National Biomedical Analysis Center, Beijing 100850, China), AuthorCompanyExt(id=1198960145066721375, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960145049944158, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.国家生物医学分析中心, 北京 100850)])]), Author(id=1198960147776241969, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, orderNo=6, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1198960147889488183, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960147776241969, language=EN, stringName=Ai-ping ZHENG, firstName=Ai-ping, middleName=null, lastName=ZHENG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹, address=1. Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, Academy of Military Sciences, Beijing 100850, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1198960147990151486, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, authorId=1198960147776241969, language=CN, stringName=郑爱萍, firstName=爱萍, middleName=null, lastName=郑, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹, address=1.军事科学院军事医学研究院毒物药物研究所, 北京 100850, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1198960144945086547, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, xref=null, ext=[AuthorCompanyExt(id=1198960144949280852, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960144945086547, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, Academy of Military Sciences, Beijing 100850, China), AuthorCompanyExt(id=1198960144957669461, tenantId=1146029695717560320, journalId=1189982191388893191, articleId=1198628603962425457, companyId=1198960144945086547, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.军事科学院军事医学研究院毒物药物研究所, 北京 100850)])])]

韩晓璐¹, 王珊珊¹, 彭静², 洪晓轩¹, 王增明¹^,^*, 王娜²^,^*, 郑爱萍¹

作者信息

1.军事科学院军事医学研究院毒物药物研究所, 北京 100850

2.国家生物医学分析中心, 北京 100850

通讯作者:

*王增明, Tel: 86-10-66874665, E-mail: wangzm.1986@163.com;

王娜, Tel: 86-10-55498704, E-mail: wn@proteomics.cn

Research progress of artificial intelligence in 3D printed drugs

Xiao-lu HAN¹, Shan-shan WANG¹, Jing PENG², Xiao-xuan HONG¹, Zeng-ming WANG¹^,^*, Na WANG²^,^*, Ai-ping ZHENG¹

Affiliations

1. Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, Academy of Military Sciences, Beijing 100850, China

2. National Biomedical Analysis Center, Beijing 100850, China

出版时间: 2023-06-12 doi: 10.16438/j.0513-4870.2022-1259

文章导航

摘要

收起

2015年美国提出了精准医学的医疗理念, 将医药治疗从“一刀切”转向个性化, 更加强调患者个性化及药物定制化。同年, 全球首个3D打印药片Spritam^®上市, 标志着药物3D打印这种新兴技术获得监管部门认可, 同时也为个性化药物定制提供了一种新途径。3D打印技术学科交叉性强、灵活性高, 应用于制剂行业对从业人员提出了更高要求。随着人工智能(artificial intelligence, AI) 的发展, 现代社会能以超人类的速度和智力执行各项任务, 如疾病诊断、机器人手术等。机器学习(machine learning, ML) 作为主要的AI技术已在3D打印药物研发的多个环节得到广泛应用, 加快了3D打印药物的研发、生产及临床应用, 并推动全球个性化医学和工业4.0新进程。本综述介绍了药物3D打印技术、非AI药物优化技术和ML等AI关键技术的基本概念和主要分类, 重点分析了ML在药物3D打印中的应用及研究进展, 阐明AI如何赋能药物3D打印在前处理、打印过程及后处理过程的智能化水平, 为加速3D打印药物发展提供了新思路。

关键词

3D打印 / 人工智能 / 机器学习 / 实验设计 / 有限元分析

Abstract

收起

In 2015, the United States put forward the concept of precision medicine, which changed medical treatment from "one size fits all" to personalization, and paid more attention to personalization and drug customization. In the same year, Spritam^®, the world's first 3D printed tablet, was in the market, marking the emerging pharmaceutical 3D printing technology was recognized by regulatory authorities, and it also provided a new way for drug customization. 3D printing technology has strong interdisciplinary and high flexibility, which puts forward higher requirements for pharmaceutical staffs. With the development of artificial intelligence (AI), modern society can perform various tasks, such as disease diagnosis and robotic surgery, with superhuman speed and intelligence. As a major AI technology, machine learning (ML) has been widely used in many aspects of 3D printing drug, accelerating the research and development, production, and clinical application, and promoting the new process of global personalized medicine and industry 4.0. This paper introduces the basic concepts and main classifications of 3D printing drug, non-AI drug optimization technology and ML. It focuses on the analysis of the research progress of ML in 3D printing drug, and elucidates how AI can empower the intelligent level of 3D printing drug in pre-processing, printing, and post-processing process. It provides a new idea for accelerating the development of 3D printed drug.

Key words

3D printing / artificial intelligence / machine learning / design of experiment / finite element analysis

引用本文

韩晓璐, 王珊珊, 彭静, 洪晓轩, 王增明, 王娜, 郑爱萍. 人工智能在3D打印药物的研究进展. 药学学报, 2023 , 58 (6) : 1577 -1585 . DOI: 10.16438/j.0513-4870.2022-1259

Xiao-lu HAN, Shan-shan WANG, Jing PENG, Xiao-xuan HONG, Zeng-ming WANG, Na WANG, Ai-ping ZHENG. Research progress of artificial intelligence in 3D printed drugs[J]. Acta Pharmaceutica Sinica, 2023 , 58 (6) : 1577 -1585 . DOI: 10.16438/j.0513-4870.2022-1259

正文

收起

传统医学领域仍处于“一刀切”的医疗模式中, 商业化药品依赖于大规模生产, 在剂量和组成上缺乏针对患者的灵活性策略。随着新技术发展, 人们逐渐认识到这种治疗模式的僵化, 并大力推动个性化药物^[1]。3D打印技术通过计算机将二维图像转换成三维数据, 然后通过特定成型设备, 逐层打印、层层叠加, 最终把蓝图变成实物。2015年, 美国食品药品监督管理局(Food and Drug Administration, FDA) 批准了全球首个3D打印药片Spritam^®; 2020年, 中国南京Triastek公司的T₁₉新药临床试验(investigational new drug, IND) 申请获得批准^[2], 德国默克公司也开始布局通过选择性激光烧结3D打印技术开发可商业化的药物, 这3个标志性事件为药物3D打印的可行性奠定了基础。2019年, 首个3D打印药物优于传统药物的临床研究发表, 也加速了其向临床实践的过渡^[3]。

3D打印药物以患者为中心, 由此3D打印制剂的设计和生产过程也比传统制剂多了更多抉择。ML是AI的一个子集, 利用ML从已有的大量数据中学习预测新结果, 可不被分析变量的数量所限制, 可在数据可视化、图像识别和系统建模方面发挥作用, 更好地理解3D打印复杂的加工过程, 为任何给定场景的3D打印药物设计、生产与调配提供最优化方案。本综述围绕非AI药物优化技术和AI关键技术在药物3D打印的应用现状和前景进行论述, 综合对比了不同技术的优缺点, 为加速3D打印药物研发提供了新思路。

1 药物3D打印技术及传统优化技术

收起

1.1 药物3D打印技术

与传统压片法不同, 药物3D打印技术系通过计算机辅助设计(computer aided design, CAD) 软件设计创建制剂结构, 具备良好的微观精确控制与空间精准调控能力, 依托CAD可制备出具有特定内部结构和性能的医疗产品^{[4, 5]}, 可灵活定制药物剂量、组合和释放^[2]。经历20多年的发展历程, 药物3D打印行业已衍生出不同技术种类, 如黏结剂喷射、熔融沉积、半固体挤出、选择性激光烧结和立体光刻等^[6]。各子技术存在异质性, 具体特点及原理如下。

1.1.1 黏结剂喷射技术(binder jet 3D printing, BJ-3DP)

BJ-3DP与湿法制粒技术相似, 打印原理如图 1A所示。黏合剂液滴从各自的打印头喷射出来, 根据三维结构要求在粉床上沉积, 利用粉液黏结作用, 逐层叠加形成药物实体。1996年, 麻省理工学院的Michael J. Cima教授首次报道了BJ-3DP技术可应用于制药。2015年, Aprecia公司基于此技术开发了全球第一款3D打印药品Spritam^{® [7]}。该技术是最早被应用到制药领域的3D打印技术, 已成功实现了产业化和商业化, 目前主要用于口服固体制剂和植入剂的开发。

1.1.2 熔融沉积技术(fused deposition modeling, FDM)

FDM技术设备简单、成本低、产品强度高, 是最普及的3D打印技术。该技术利用丝状热熔性材料加热融化后按指定位置凝固成型, 逐层构建三维实体, 打印原理见图 1B。FDM技术可通过丝材载药直接打印药物, 或后填充的方式制备药物胶囊, 通过三维结构设计和不同种类聚合物来调控药物释放。FDM技术也存在一定局限性, 打印过程需加热, 适合于热稳定的药物, 被用于复方制剂、缓释制剂、植入剂和药物洗脱装置的研究。目前, 默沙东和Multiply Labs公司采用该技术快速制备小批量、个性化药物。

1.1.3 半固体挤出技术(semi solid extrusion, SSE)

SSE是材料挤压技术的一种变种, 通过气动、活塞或螺杆的方式挤出凝胶或糊剂材料, 层层堆叠形成三维结构, 打印原理见图 1B。与FDM技术相比, SSE采用低打印温度, 更适用于热敏感性药物。SSE技术被用于复方制剂、控释制剂、速释片剂、咀嚼片、口腔分散薄膜和固体自乳化制剂等方面的研究。目前, 英国FabRx和新加坡Craft Health两家药物公司对该技术进行了商业化布局。

1.1.4 立体光刻技术(stereolithography, SLA)

SLA基于光聚合固化原理, 利用液态光敏树脂在紫外激光束照射下快速固化的特性进行3D打印产品的制备, 打印原理见图 1C。SLA技术因成型速度快、打印精度高、实体质量高等优势, 被广泛用于制造具有多种功能的生物医学设备和植入体, 片剂、透皮贴剂及微针也有相关研究报道^[8]。由于光敏树脂的安全性问题等原因, 光固化3D打印制剂实现商业化, 仍需解决诸多问题。

1.1.5 选择性激光烧结技术(selective laser sintering, SLS)

SLS基于粉末床熔融成型原理, 通过控制激光光源扫描, 有选择地烧结固体粉末材料形成药物, 打印原理见图 1D。SLS技术打印药片后处理较为繁琐, 还需在粉末中加入激光吸收剂, 使得工艺难度增加。2020年, 默克公司利用该技术针对孤儿药和肿瘤适应症群体开发药物, 并进行商业化生产。

1.2 非ML药物优化技术

新型3D打印药物在提供多样化的设计组合的同时, 也对制剂研究人员提出更高要求, 传统正交实验方式显然是不适用的。制药领域已建立了几种较为成熟的工艺优化技术, 这些技术已用于规范药物3D打印的开发过程。本综述主要讨论3种在药物3D打印中开展的较为广泛的非ML药物优化技术。

1.2.1 实验设计(design of experiment, DoE)

质量源于设计(quality by design, QbD) 理念在制药领域中的重要性越来越高, 质量不是通过检验注入到产品中, 而是通过设计赋予的。DoE是实施QbD研发理念的重要工具之一, 是探究对药品关键质量有显著影响的物料属性、工艺参数及其设计空间的最有效途径。在DoE模型中, 需明确研究目标、变量及变量水平^[9], 允许映射各变量单独和组合影响实验结果的程度, 建模过程将受到可用资源、统计能力和操作过程的影响。研究人员也可使用该模型来预测最佳设计空间。全阶乘设计需执行所有的实验设计; 分阶乘设计是全阶乘的一个抽样子集, 以降低统计功率为代价减少实验次数。目前为止, 药物3D打印中的DoE^[10]已被用于评价FDM^[11]、SLS^[12]和SSE^[13]中各种材料及工艺参数与药物性能之间的关系。本课题组Wang等^[14]运用质量源于设计理念, 采用3个中心点的2³全因子DoE, 对BJ-3DP口崩片的打印液组成进行优化, 如图 2A所示。研究使用40% (v/v) 异丙醇水溶液作为基础溶剂, 以聚乙烯吡咯烷酮K30、甘油及聚山梨酯20用量作为自变量, 对打印片剂重量差异、硬度、脆碎度及分散均匀性进行分析。通过DoE模型获得了打印液处方的设计空间, 并利用响应优化器获得了最优的打印液处方, 为黏结剂喷射型3D打印制剂中打印液的研究提供了参考和依据。

1.2.2 机制模型

机制模型是利用物理原理构建数学模型解释过程变量的方法, 目前已在3D打印领域开展了应用^[15]。这些模型的构建需一系列热力学、粒子物理和流体动力学等专业知识^[15]。对3D打印过程中的力学模型进行探索, 相关模型包括力学性能预测^[15]、光聚合动力学和粉末床技术中的吸放热过程等^[16]。然而, 对于药物3D打印的机制建模还没有得到充分的应用。与其他技术(如DoE) 不同, 目前还没有现成的软件来进行模拟3D打印的机制模型。2019年, Zidan等^[17]将药物微挤压3D打印过程与机制模型相结合进行了研究, 对卡波姆基质的3D打印糊剂进行挤压性测试, 采用符合Herschel-Bulkley方程对微挤压分析模型进行19组挤出性数据的拟合, 建立了糊剂的流量、喷嘴和筒体直径、打印压力和滑流角的本构方程, 计算了膏体的关键材料参数, 即屈服应力σ₀、稠度系数k和流动系数n (图 2B)。研究结果显示, 材料参数受可溶性和溶胀性辅料的比例影响显著, 喷嘴直径对n有显著影响, 而对k无显著影响。此研究利用机制建模帮助调整糊剂的流变性, 后续在此基础上指导了缓释3D打印片剂的开发^[18]。

1.2.3 仿真模拟技术

与DoE相似, 有限元分析(finite element analysis, FEA) 与计算流体力学(computational fluid dynamics, CFD) 是学术界和工业界使用的另一种基于仿真模拟的优化技术, 利用计算机和数学方法对真实物理系统进行模拟和分析, 已广泛应用于航空航天、电子和生物力学等领域。这两种建模技术都能模拟产品受到的一系列力的变化, 包括机械应力、热和流体动力学等, 由此来执行复杂的模拟实验, 并具有高精度计算和预测能力^[19], 如加载新的模型和施加预期应力, 在大小和方向上进行因式分解。由于设计结果可通过应力变化图及流速变化图等方式具象地表现出来, 所以这两种技术被认为属于“白盒”测试的一种。3D打印领域的药学研究人员利用这类技术来预测FDM丝材的热力学行为^[20]、可视化微针中的应力分布^[21]等方面的研究。Chen等^[22]制造了一种由中空微针和超声换能器组成的集成3D打印超声微针阵列装置(图 2C)。中空微针使药物能穿透角质层, 通过MNs传输的超声为药物在组织中的运输提供动力。该研究利用FEA进行了单针应力分析, 构建了角型、圆柱型和箭型的微针, 有限元模型优化了不同形状微针贴片的渗透强度, 并对单针应变进行了数值模拟, 优化了空心微针的几何形状设计。

2 ML及其在药物3D打印中的应用

收起

2.1 AI与ML

AI是研究和开发用于模拟、延伸和扩展人类智能理论、方法、技术及应用系统的一门新的技术科学^[23]。AI可使针对单一任务, 如新药发现或医学图像诊断^[21]; 也可提供类似于人脑的认知能力, 如无人驾驶汽车^[24]。ML是AI的主要技术之一, 能适应训练数据的变化, 以远远超过人类智力的速度工作, 并大幅降低风险错误。在进行任何学习之前, 必须对数据进行预处理, 并对数据进行矢量化, 以确保数据的准确和ML的适用性。在数据处理过程中, 一般都需进行数据的清洗, 如数据集是否存在重复、是否存在缺失、数据是否完整、数据中是否存在异常值等。数据异常会影响ML模型的性能。一旦数据被输入ML算法并形成了一个模型, 就可对新数据做出预测。通常将规则附加到模型上, 算法就可被训练来识别数据, 从而“学习”数据特征如何映射到结果, ML的学习方式因具体ML技术而异。为全面评估ML模型预测性能, 可采用度量标准进行评价, 如准确度、精确度、查全率、特异性等^[25]。根据不同学习方式, ML可被分为五大类, 其具体分类及在药学领域的应用情况见表 1^[26-31]。

2.2 ML在药物3D打印中的应用

AI和3D打印都是颠覆性的新技术, 在第四次工业革命中, 药物3D打印和ML技术都推进了传统制造业的自动化^[32], 直到2020年, 相关学者将两者结合进行研究, 验证了ML可能进一步加速3D打印在个性化药物研发的应用, 两者的发展时间轴如图 3所示。Gongora等^[33]研究表明ML可将FDM的实验次数减少至原实验数量的17%。ML已在预打印、实时打印和打印后阶段全流程展开了应用。

2.2.1 预打印阶段的ML

在药物3D打印设计过程中, 必须依靠制剂经验来调整处方及工艺, 实现个性化药物需求。但药品研发是一个复杂的过程, 细微的偏差也会显著影响药物特性^[34]。在个性化药物设计中, 需考虑患者的疾病情况、性别、年龄、吞咽能力、口味偏好、所需药物剂量、药物释放动力学和联合用药等情况^[35-38]。在预打印阶段, ML已被用于预测药物的稳定性、载药量、材料的可打印性等^{[31, 39-41]}。

可打印性是3D打印技术的关键影响因素之一, 但可打印性一直没有建立良好的评价标准。ML可被用来预测处方的可打印性^[25]。在第一项研究中, Elbadawi等^[25]利用热熔挤出技术(hot melt extrusion, HME) 生产了614种载药丝材的处方数据集, 其中包含了145种辅料。每种处方都根据丝材的机械特性、可打印性及挤出和打印温度进行了标记。研究建立的ML模型能预测丝材的力学性能, 仅用处方材料的加权分数作为输入, 研究了不同的ML技术的预测性能, 即多元线性回归(multi-linear regression, MLR)、K近邻算法(K-nearest neighbors, kNN)、支持向量机(support vector machines, SVM)、随机森林(random forests, RF)、神经网络(neural networks, NN) 和深度学习(使用分层神经网络)。结果显示, 深度学习、RF和SVM预测总体精度能达到70%以上。其中SVM最高, 实现了76%的总体精度(图 4A)。使用物理属性进行训练时, 模型精度下降到70%, 但该模型不必重新训练模型就可用于新材料预测。为保证ML模型学习信息的透明度, 用RF对响应变量公式中材料的重要性进行排序, 发现载体浓度是预测可打印性的决定因素, 其次是增塑剂浓度, 这一结果更符合对于可打印性处方变量的经验判断。不同技术建模的表现取决于维度大小、样本和数据的线性等因素。该研究团队开发了一个基于计算网络的制药软件“M3DISEEN”, 希望可利用AI和ML技术加速FDM 3D打印的发展。

主成分分析技术(principal component analysis, PCA) 已被用于FDM载药丝材的可打印性预测^[42]。以市售打印丝材的力/距离图(以下简称柔度图) 为对照, 采用相关分析和PCA确定不同丝材的柔度图与其供料性(被标记为“可送丝”、“可调整”或“不可送丝”, 决定可打印性的关键因素) 是否存在统计学相关性。结果显示, 机械缺陷丝材与适合打印的丝材间存在明显的聚类差异, 且采用PCA结果聚类后具有更佳的可视化水平。此研究深入剖析了处方组成和丝材供料性之间的关系, 为FDM处方开发带来新的见解。

2.2.2 自动3D打印药品

在个性化药品生产中, 打印参数直接影响了产品的最终性能, 药物3D打印过程中期望利用自主学习, 创建智能的3D打印过程。监督和非监督ML已开始用于预测药物的最佳打印条件。一项研究^[43]证明了PCA可用于确定FDM打印参数多变量的快速分析。该研究中聚己内酯被用于打印内部的网格, 从而检查其宏观和微观结构对FDM的重现性。根据PCA双曲面图发现, 打印速度比打印温度更能解释数据方差, 且打印速度与打印线宽和产品质量呈负相关。该研究揭示了FDM过程中片重、给药精度和维度真实性之间的相互关系(图 4B), 提出了一种标准偏差求和的打印参数提取方法, 该方法在给定所有质量参数的前提下, 可提取出最优打印参数。此外, Elbadawi等^[25]也利用监督学习预测最佳HME和FDM打印温度。在开发中, HME的起始温度通常根据经验指导, 建议从高于处方甘油三酯15~60 ℃开始, 利用ML技术可不必使用差示扫描量热法确定玻璃化温度, 预测其精度达到± 8.9 ℃。到目前为止, 对于FDM打印温度还没有标准化的预测方法, 这项研究也是第一次建立理想的打印温度算法。

在药物3D打印过程的另一个应用是打印偏差的智能识别, 利用ML与计算机视觉结合实现打印检测。计算机视觉也被称为机器视觉, 可实现对图像或视频的超人类解释^[25]。一项研究^[44]将ML和机器视觉结合, 用于检测SLS打印过程中的异常。打印机通过安装摄像头监控打印过程, 通过算法自适应训练来检测异常。结果表明, 该算法在检测不存在异常时达到了100%的正确率, 在检测异常时达到89%的正确率。除照片图像外, 视频也可通过ML技术进行处理。将深度学习与实时视频监控相结合, 可自动校正“过挤压”和“欠挤压”的FDM打印过程。为实现自主检测, 离线训练模型被用来分类“过挤压”、“欠挤压”或“高质量”打印。该模型可被实时应用, 检测出不规则挤出, 从而使打印机能调整打印速度、流量或喷嘴高度, 预测打印质量达到了98%以上的准确率, 且预测速度比人类反应速度快得多。

2.2.3 打印后阶段的ML

在3D打印药物生产后端, ML被作为一种质量控制手段, 如含量及体外溶出的预测。ML可用来支持过程分析技术(process analysis technology, PAT), PAT是一种实时监测原材料、中间体和过程的关键质量和性能特征为手段建立起来的设计、分析和控制生产的技术, 已在制药行业广泛应用。无损检测技术, 如拉曼光谱或近红外光谱(near-infrared, NIR), 可利用多元分析建立预测模型, 将光谱与不同参数(如药物浓度) 联系起来^[45]。无损技术和智能PAT的发展也为实现3D打印药品实时监测提供了技术基础。

偏最小二乘(partial least squares, PLS) 是以监督方式确定新特征, 并计算变量之间的协方差^[46]。有研究将NIR与PLS结合, 用于单个SLS打印药片中两种不同药物的快速定量^[47]。该团队的另一项研究^[48], 再次使用PLS建立模型, 利用便携式近红外设备预测药片中药物浓度(图 4C), 该方法能实现不同几何形状和处方类型的药物浓度高精度检测, 检测范围为4%~40% (w/w)。将便携式NIR与高效液相色谱法测定的药物浓度进行比较, t检验显示两种方法无显著性差异。

高光谱成像技术是基于非常多窄波段的影像数据技术, 将成像技术与光谱技术相结合, 可用于药物3D打印的质量控制^[49]。该技术结合光谱和空间信息, 通过每个像素生成光谱的原始图像来映射药物浓度、分布情况及分布状态^[50]。PCA可用来消除数据中的噪声、识别光谱数据, 提高光谱图像的解析速度。PCA和高光谱成像结合已被用于喷墨打印药物中茶碱浓度^[51]、SLS药物中盐酸克林霉素棕榈酸酯^[52]和FDM药物中吲哚美辛^[50]的检测。一系列研究证明ML可进一步帮助解析3D打印药物性能, 也为复杂制剂的在线检测提供了可能。

药物溶出是药剂学的一项基本表征手段, 传统溶出评价手段费时费力。一些研究已使用ML预测3D打印药物的溶出曲线^[53-55]。第一项研究^[53]中, 利用人工神经网络(artificial neural network, ANN) 来预测DLP打印药物的溶解行为, 这项研究比较了单独使用材料组成, 和同时使用材料组成及曝光时间作为输入的两个ANN模型的预测差异。结果表明, 单独使用材料组成的ANN模型r²为0.981, 而考虑曝光时间的ANN模型r²为0.996, 从而推断曝光时间是一个相关变量。另一项研究^[56]中, 以流变特性作为输入, 比较了4种不同ML技术预测FDM药品溶出特性的准确性。该研究表明, 在预测药物释放方面, 非线性技术决策树优于其他线性技术。

3 总结与展望

收起

药物3D打印技术开发过程涉及医药用3D打印药物设备的设计和制造、制药工艺和药物剂型专用的辅料及打印材料的开发, 以及为药物三维结构剂型的释放机制的体内外研究, 整体开发难度大, 技术壁垒高, 对专业人员要求高, 跨越工程学、材料学和药学等诸多专业, 也限制了一些优化技术(如DoE、机制建模和FEA等) 在药物3D打印过程的应用。DoE数据容量限制和模型狭窄, 无法对许多具有相互作用因素的复杂过程进行建模。FEA优化周期成本高耗时长, 并且需要大量专业知识。而ML不需深入的专业知识, 只要有足够数量数据, 就可无缝处理数千个条目的数据集。ML作为更先进的流程优化技术, 可能会与其他优化技术结合或完全取代它们^[52], 为3D打印药物的自主连续生产和智能制造提供了机会^[57]。

本综述强调了人工智能在药物3D打印技术领域的相关应用, 为医学摆脱“一刀切”的供应范式, 拥抱个性化药物管理提供了借鉴意义。药物3D打印可为临床实践提供符合患者需求的个性化药物, 但需有经验的3D打印从业者及制剂研发人员的专业知识。过程优化技术在药学领域已开展了广泛应用, 如有限元分析、机械建模和DoE等技术。然而, 没有一个技术手段可优化药物3D打印的各个阶段。相比较而言, ML则可为3D打印药物生产的各个阶段提供智能优化策略。而且, 药物3D打印的每个阶段都可构建成一个智能物联网, 处理药物开发的每一个阶段: 从配方设计到最终产品上市。这样的管理模式将完全消除药物3D打印中的人力需求, 向患者提供全时段定制的个性化药物。

作者贡献: 韩晓璐和王珊珊具体完成文章撰写; 彭静和洪晓轩完成文献检索和筛选; 王增明和王娜确定文章方向及内容组成; 郑爱萍对文章撰写质量进行把关和审校。

利益冲突: 文章内容不涉及相关利益冲突, 未涉及任何厂家及相关经济组织直接或间接的经济或利益的赞助。

基金

收起

国家自然科学基金资助项目(82073793)
北京市自然科学基金资助项目(L202043)

参考文献

收起

文献

收起

参考文献引证文献

排序方式：

[1]

Trenfield SJ, Awad A, Madla CM, et al. Shaping the future: recent advances of 3D printing in drug delivery and healthcare[J]. Expert Opin Drug Deliv, 2019, 16: 1081-1094.

https://doi.org/10.1080/17425247.2019.1660318

[2]

Zheng Y, Deng F, Wang B, et al. Melt extrusion deposition (MED^TM) 3D printing technology - a paradigm shift in design and development of modified release drug products[J]. Int J Pharm, 2021, 602: 120639.

https://doi.org/10.1016/j.ijpharm.2021.120639

[3]

Goyanes A, Madla CM, Umerji A, et al. Automated therapy preparation of isoleucine formulations using 3D printing for the treatment of MSUD: first single-centre, prospective, crossover study in patients[J]. Int J Pharm, 2019, 567: 118497.

https://doi.org/10.1016/j.ijpharm.2019.118497

[4]

Vaz VM, Kumar L. 3D printing as a promising tool in persona-lized medicine[J]. AAPS PharmSciTech, 2021, 22: 49-54.

https://doi.org/10.1208/s12249-020-01905-8

[5]

Trenfield SJ, Madla CM, Basit AW, et al. Binder jet printing in pharmaceutical manufacturing [M] // Abdul WB, Simon G. 3D Printing of Pharmaceuticals. Cham: Springer, 2018, 31: 41-54.

[6]

Trenfield SJ, Awad A, Goyanes A, et al. 3D printing pharmaceuticals: drug development to frontline care[J]. Trends Pharmacol Sci, 2018, 39: 440-451.

https://doi.org/10.1016/j.tips.2018.02.006

[7]

Shi J, Wang Z, Zheng A. Applications and challenges of 3D printing technique in manufacturing pharmaceutical preparations[J]. Prog Pharm Sci Mar (药学进展), 2019, 43: 164-173. https://www.cnki.com.cn/Article/CJFDTOTAL-TAIY202303012.htm

[8]

Li H, Xu L, Zhong W, et al. Recent advances of photocurable 3D printed pharmaceutical preparations[J]. Chin Pharm J (中国药学杂志), 2021, 56: 1189-1195. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGYX202310003.htm

[9]

Moldenhauer D, Nguyen DCY, Jescheck L, et al. 3D screen printing - an innovative technology for large-scale manufacturing of pharmaceutical dosage forms[J]. Int J Pharm, 2021, 592: 120096.

https://doi.org/10.1016/j.ijpharm.2020.120096

[10]

Zhang J, Thakkar R, Zhang Y, et al. Structure-function correlation and personalized 3D printed tablets using a quality by design (QbD) approach[J]. Int J Pharm, 2020, 590: 119945.

https://doi.org/10.1016/j.ijpharm.2020.119945

[11]

Carlier E, Marquette S, Peerboom C, et al. Investigation of the parameters used in fused deposition modeling of poly (lactic acid) to optimize 3D printing sessions[J]. Int J Pharm, 2019, 565: 367-377.

https://doi.org/10.1016/j.ijpharm.2019.05.008

[12]

Barakh Ali SF, Mohamed EM, Ozkan T, et al. Understanding the effects of formulation and process variables on the printlets quality manufactured by selective laser sintering 3D printing[J]. Int J Pharm, 2019, 570: 118651.

https://doi.org/10.1016/j.ijpharm.2019.118651

[13]

Tsintavi E, Rekkas DM, Bettini R. Partial tablet coating by 3D printing[J]. Int J Pharm, 2020, 581: 119298.

https://doi.org/10.1016/j.ijpharm.2020.119298

[14]

Wang Z, Li X, Han X, et al. Full-factor design of experiment application in the development of printing ink for binder jetting 3D printed preparation[J]. Acta Pharm Sin (药学学报), 2021, 56: 1460-1469. https://www.cnki.com.cn/Article/CJFDTOTAL-YISY202303019.htm

[15]

Wei HL, Mukherjee T, Zhang W, et al. Mechanistic models for additive manufacturing of metallic components[J]. Prog Mater Sci, 2021, 116: 100703.

https://doi.org/10.1016/j.pmatsci.2020.100703

[16]

Wang HY, Qin XD. A mechanistic model for cutting force in helical milling of carbon fiber-reinforced polymers[J]. Int J Adv Manuf Tech, 2016, 82: 1485-1494.

https://doi.org/10.1007/s00170-015-7460-0

[17]

Zidan A, Alayoubi A, Coburn J, et al. Extrudability analysis of drug loaded pastes for 3D printing of modified release tablets[J]. Int J Pharm, 2019, 554: 292-301.

https://doi.org/10.1016/j.ijpharm.2018.11.025

[18]

Zidan A, Alayoubi A, Asfari S, et al. Development of mechanistic models to identify critical formulation and process variables of pastes for 3D printing of modified release tablets[J]. Int J Pharm, 2019, 555: 109-123.

https://doi.org/10.1016/j.ijpharm.2018.11.044

[19]

Feng JR, Yu GH, Zhao MK, et al. Dynamic risk assessment framework for industrial systems based on accidents chain theory: the case study of fire and explosion risk of UHV converter transformer[J]. Reliab Eng Syst Safe, 2022, 228: 2108760.

[20]

Ramanath HS, Chua CK, Leong KF, et al. Melt flow behaviour of poly-epsilon-caprolactone in fused deposition modelling[J]. J Mater Sci Mater Med, 2008, 19: 2541-2550.

https://doi.org/10.1007/s10856-007-3203-6

[21]

Xenikakis I, Tzimtzimis M, Tsongas K, et al. Fabrication and finite element analysis of stereolithographic 3D printed microneedles for transdermal delivery of model dyes across human skin in vitro[J]. Eur J Pharm Sci, 2019, 137: 104976.

https://doi.org/10.1016/j.ejps.2019.104976

[22]

Chen Z, Wu H, Zhao S, et al. 3D-printed integrated ultrasonic microneedle array for rapid transdermal drug delivery[J]. Mol Pharm, 2022, 19: 3314-3322.

https://doi.org/10.1021/acs.molpharmaceut.2c00466

[23]

Chen JJF. Data Mining/Machine Learning Techniques for Drug Discovery: Computational and Experimental Pipeline Development [D]. Akron: The University of Akron, 2018.

[24]

Mills J. Autonomous vehicles[J]. Motor Pro, 2019, 2: 114-122.

[25]

Elbadawi M, Muñiz Castro B, Gavins FKH, et al. M3DISEEN: a novel machine learning approach for predicting the 3D printability of medicines[J]. Int J Pharm, 2020, 590: 119837.

https://doi.org/10.1016/j.ijpharm.2020.119837

[26]

Bravo À, Piñero J, Queralt-Rosinach N, et al. Extraction of relations between genes and diseases from text and large-scale data analysis: implications for translational research[J]. BMC Bioinformatics, 2015, 16: 55.

https://doi.org/10.1186/s12859-015-0472-9

[27]

Wang YH, Li Y, Yang SL, et al. Classification of substrates and inhibitors of P-glycoprotein using unsupervised machine learning approach[J]. J Chem Inf Model, 2005, 45: 750-757.

https://doi.org/10.1021/ci050041k

[28]

Park S, Lee DG, Shin H. Network mirroring for drug repositioning[J]. BMC Med Inform Decis Mak, 2017, 17: 55-66.

https://doi.org/10.1186/s12911-017-0449-x

[29]

Nam Y, Kim M, Chang HS, et al. Drug repurposing with network reinforcement[J]. BMC Bioinformatics, 2019, 20: 383.

https://doi.org/10.1186/s12859-019-2858-6

[30]

Gupta R, Srivastava D, Sahu M, et al. Artificial intelligence to deep learning: machine intelligence approach for drug discovery[J]. Mol Divers, 2021, 25: 1315-1360.

https://doi.org/10.1007/s11030-021-10217-3

[31]

Yang Y, Ye Z, Su Y, et al. Deep learning for in vitro prediction of pharmaceutical formulations[J]. Acta Pharm Sin B, 2019, 9: 177-185.

https://doi.org/10.1016/j.apsb.2018.09.010

[32]

Arden NS, Fisher AC, Tyner K, et al. Industry 4.0 for pharmaceutical manufacturing: preparing for the smart factories of the future[J]. Int J Pharm, 2021, 602: 120554.

https://doi.org/10.1016/j.ijpharm.2021.120554

[33]

Gongora AE, Xu B, Perry W, et al. A Bayesian experimental autonomous researcher for mechanical design[J]. Sci Adv, 2020, 6: 1708-1714.

[34]

Karakurt I, Aydoğdu A, Çıkrıkcı S, et al. Stereolithography (SLA) 3D printing of ascorbic acid loaded hydrogels: a controlled release study[J]. Int J Pharm, 2020, 584: 119428.

https://doi.org/10.1016/j.ijpharm.2020.119428

[35]

Stillhart C, Vučićević K, Augustijns P, et al. Impact of gastrointestinal physiology on drug absorption in special populations-an UNGAP review[J]. Eur J Pharm Sci, 2020, 147: 105280.

https://doi.org/10.1016/j.ejps.2020.105280

[36]

Merchant HA, Liu F, Orlu Gul M, et al. Age-mediated changes in the gastrointestinal tract[J]. Int J Pharm, 2016, 512: 382-395.

https://doi.org/10.1016/j.ijpharm.2016.04.024

[37]

Dou L, Gavins FKH, Mai Y, et al. Effect of food and an animal's sex on P-glycoprotein expression and luminal fluids in the gastrointestinal tract of Wistar rats[J]. Pharmaceutics, 2020, 12: 296-311.

https://doi.org/10.3390/pharmaceutics12040296

[38]

Vallet T, Michelon H, Orlu M, et al. Acceptability in the older population: the importance of an appropriate tablet size[J]. Pharmaceutics, 2020, 12: 746-757.

https://doi.org/10.3390/pharmaceutics12080746

[39]

Han R, Xiong H, Ye Z, et al. Predicting physical stability of solid dispersions by machine learning techniques[J]. J Control Release, 2019, 311-312: 16-25.

https://doi.org/10.1016/j.jconrel.2019.08.030

[40]

Huang C, Clayton EA, Matyunina LV, et al. Machine learning predicts individual cancer patient responses to therapeutic drugs with high accuracy[J]. Sci Rep, 2018, 8: 16444.

https://doi.org/10.1038/s41598-018-34753-5

[41]

King AC, Woods M, Liu W, et al. High-throughput measurement, correlation analysis, and machine-learning predictions for pH and thermal stabilities of Pfizer-generated antibodies[J]. Protein Sci, 2011, 20: 1546-1557.

https://doi.org/10.1002/pro.680

[42]

Nasereddin JM, Wellner N, Alhijjaj M, et al. Development of a simple mechanical screening method for predicting the feedability of a pharmaceutical FDM 3D printing filament[J]. Pharm Res, 2018, 35: 151-164.

https://doi.org/10.1007/s11095-018-2432-3

[43]

Alhijjaj M, Nasereddin J, Belton P, et al. Impact of processing parameters on the quality of pharmaceutical solid dosage forms produced by fused deposition modeling (FDM)[J]. Pharmaceutics, 2019, 11: 633-654.

https://doi.org/10.3390/pharmaceutics11120633

[44]

Scime L, Beuth J. Anomaly detection and classification in a laser powder bed additive manufacturing process using a trained computer vision algorithm[J]. Addit Manufact, 2018, 19: 114-126.

https://doi.org/10.1016/j.addma.2017.11.009

[45]

Sampaio PS, Soares A, Castanho A, et al. Optimization of rice amylose determination by NIR-spectroscopy using PLS chemometrics algorithms[J]. Food Chem, 2018, 242: 196-204.

https://doi.org/10.1016/j.foodchem.2017.09.058

[46]

Schwendicke F, Samek W, Krois J. Artificial intelligence in dentistry: chances and challenges[J]. J Dent Res, 2020, 99: 769-774.

https://doi.org/10.1177/0022034520915714

[47]

Trenfield SJ, Tan HX, Goyanes A, et al. Non-destructive dose verification of two drugs within 3D printed polyprintlets[J]. Int J Pharm, 2020, 577: 119066.

https://doi.org/10.1016/j.ijpharm.2020.119066

[48]

Trenfield SJ, Goyanes A, Telford R, et al. 3D printed drug products: non-destructive dose verification using a rapid point-and-shoot approach[J]. Int J Pharm, 2018, 549: 283-292.

https://doi.org/10.1016/j.ijpharm.2018.08.002

[49]

Sacré PY, De Bleye C, Chavez PF, et al. Data processing of vibrational chemical imaging for pharmaceutical applications[J]. J Pharm Biomed Anal, 2014, 101: 123-140.

https://doi.org/10.1016/j.jpba.2014.04.012

[50]

Scoutaris N, Ross SA, Douroumis D. 3D printed "Starmix" drug loaded dosage forms for paediatric applications[J]. Pharm Res, 2018, 35: 34-41.

https://doi.org/10.1007/s11095-017-2284-2

[51]

Vakili H, Kolakovic R, Genina N, et al. Hyperspectral imaging in quality control of inkjet printed personalised dosage forms[J]. Int J Pharm, 2015, 483: 244-249.

https://doi.org/10.1016/j.ijpharm.2014.12.034

[52]

Mohamed EM, Barakh Ali SF, Rahman Z, et al. Formulation optimization of selective laser sintering 3D-printed tablets of clindamycin palmitate hydrochloride by response surface methodology[J]. AAPS PharmSciTech, 2020, 21: 232-240.

https://doi.org/10.1208/s12249-020-01775-0

[53]

Madzarevic M, Medarevic D, Vulovic A, et al. Optimization and prediction of ibuprofen release from 3D DLP printlets using artificial neural networks[J]. Pharmaceutics, 2019, 11: 544-561.

https://doi.org/10.3390/pharmaceutics11100544

[54]

Hernandez E, Pawar P, Keyvan G, et al. Prediction of dissolution profiles by non-destructive near infrared spectroscopy in tablets subjected to different levels of strain[J]. J Pharm Biomed Anal, 2016, 117: 568-576.

https://doi.org/10.1016/j.jpba.2015.10.012

[55]

Horkovics-Kovats S, László Galata D, Zlatoš P, et al. Raman-based real-time dissolution prediction using a deterministic permeation model[J]. Int J Pharm, 2022, 617: 121624.

https://doi.org/10.1016/j.ijpharm.2022.121624

[56]

Elbadawi M, Gustaffson T, Gaisford S, et al. 3D printing tablets: predicting printability and drug dissolution from rheological data[J]. Int J Pharm, 2020, 590: 119868.

https://doi.org/10.1016/j.ijpharm.2020.119868

[57]

Gioiello A, Piccinno A, Lozza AM, et al. The medicinal chemistry in the era of machines and automation: recent advances in continuous flow technology[J]. J Med Chem, 2020, 63: 6624-6647.

https://doi.org/10.1021/acs.jmedchem.9b01956

2023年第58卷第6期

PDF下载

465

206

引用本文

BibTeX

文章信息

doi: 10.16438/j.0513-4870.2022-1259

接收时间：2022-11-23
首发时间：2025-11-21
出版时间：2023-06-12

补充材料

相关文章

文章信息

作者

出版历史

收稿日期：2022-11-23
修回日期：2022-12-22

基金

国家自然科学基金资助项目(82073793)

北京市自然科学基金资助项目(L202043)

作者信息

1.军事科学院军事医学研究院毒物药物研究所, 北京 100850

2.国家生物医学分析中心, 北京 100850

通讯作者:

*王增明, Tel: 86-10-66874665, E-mail: wangzm.1986@163.com;

王娜, Tel: 86-10-55498704, E-mail: wn@proteomics.cn

参考文献

分享链接

https://castjournals.cast.org.cn/joweb/yxxb/CN/10.16438/j.0513-4870.2022-1259

分享至

全文二维码

扫描看全文

引用本文

BibTeX

本文的引用情况

2种不同金属材料的力学参数

科 Family	属数 Number of genus	种数 Number of species	占总种数比例 Percentage of total species (%)	属 Genus	种数 Number of species	占总种数比例 Percentage of total species (%)
鹅膏菌科Amanitaceae	2	11	5.26	鹅膏菌属 Amanita	10	4.78
小菇科 Mycenaceae	2	12	5.74	丝盖伞属 Inocybe	5	2.39
多孔菌科 Polyporaceae	8	14	6.70	蜡蘑属 Laccaria	5	2.39
红菇科 Russulaceae	3	23	11.00	小皮伞属 Marasmius	6	2.87
				小菇属 Mycena	11	5.26
				光柄菇属 Pluteus	5	2.39
				红菇属 Russula	17	8.13
				栓菌属 Trametes	5	2.39

关闭全屏

BibTeX
EndNote
RefWorks
TxT