Traditional 2D cell and animal models are the most widely used disease models in biological studies, which are the "compulsory courses" for the vast majority of drugs before entering clinical trials. Traditional cell-based models lack essential features, such as complex multiple cultures, physiological microenvironments, and tissue mechanical forces. Animal models, although regarded as the current gold standard in many biological studies, have problems such as high cost, ethical issues, low throughput, and interspecific differences, which significantly limit the progress of drug development and other biological research.
Microfluidic-based organs-on-chips (OoCs) technology is proposed to fill the gap between traditional 2D cell culture and animal models and to gradually replace animal studies. As a product of the progressive development of microfluidic technology, OoCs combine microfluidic technology with cell biology; they faithfully mimic the physiological microenvironment of the in vivo target organs. These novel in vitro biological models can replicate the local characteristics of a disease and control the environmental parameters of cell survival, making them a cost-efficient and high-throughput platform for a wide range of biological studies, such as disease modeling, drug discovery, drug pharmacokinetics and toxicity prediction, and personalized medicine. In 2016, OoC technology was selected as one of the top ten emerging technologies of the year at the Davos Forum.
But the massive amount of image-based data generated by the high-throughput OoC devices, combined with the automation required to precisely control the tissue microenvironment, goes far beyond the manual analysis of researchers with a biomedical background. Therefore, OoCs urgently need to find a tool that can assist, or even replace, manual analysis and subjective judgment, so as to improve the efficiency and accuracy of the system. Artificial Intelligence (AI) has been widely used in computer vision, natural language processing, speech recognition and other fields in recent years, and has successfully achieved commercialization, which is a key technology in the "Fourth Industrial Revolution". The most popular algorithm in the field of AI is the neural network-based deep learning, the behavior of neural network-based deep learning mimics many characteristics of the human brain. Due to its strong feature representation and data mining ability, deep learning has been widely used in the fields of computer vision, natural language processing, and speech recognition.
Therefore, utilizing deep learning as a powerful tool to explore and analyze experimental data generated by OoCs can effectively explore the inherent laws behind massive data, improve the intelligence level of OoCs, and stimulate its great potential in drug development, disease modeling and personalized medicine.
On January 19, 2022, Professor Lin LI and Associate Professor Bo PENG from Professor Wei HUANG's team in Northwestern Polytechnical University published a review article on Research under the title "An Overview of Organs-on-Chips Based on Deep Learning" On the first issue of 2022. This review comprehensively summarized the latest research progress of applying deep learning algorithms to OoCs. It also envisions the future direction of this new interdisciplinary area.
The content of the article is divided into four sections. The first section briefly introduces microfluidic technology and OoC devices, showing advantages of OoCs over traditional disease models. At the same time, one of the bottlenecks in the development of organ chips is pointed out: high-throughput platforms bring massive amounts of image-based data and artificial experimental errors. The second section systematically describes the development of deep learning algorithms, interspersed with the principles of algorithms and some classic neural network models that implement deep learning. The third section comprehensively summarizes and analyzes the various deep learning algorithms that are currently suitable for organ-on-chip or have already been used in OoC data mining. Based on the classification of different application scenarios, and the complexity of deep learning algorithms, the third section introduces the relevant applications in detail, which is helpful for comparing and analyzing between different applications. The fourth section predicts the development direction of the integration of deep learning with OoCs from different aspects such as organelle segmentation and tracking in OoCs, automatic cell culture systems, drug development, rare disease-on-chips, and human-on-chips. This article is expected to become a foundational guidance paper on the automation, standardization and quantitative development of OoC technology.
From China Daily