Therapeutic regimens in regenerative medicine are witnessing an unprecedented interest in stem cells. Stem cells are undifferentiated human cells capable of self-renewal division, as creating more cells, setting the foundation for a pool of unspecialized stem cells. Such division generates two cells where one maintains its self-renewal ability while the other becomes more differentiated into a special type of cell1. Stem cells are organized into three categories, induced pluripotent stem cells (iPSCs), embryonic stem cells (ECSs), and adult stem cells (ACSs)2. Both iPSCs and ECSs fall under the category of pluripotent stem cells (PSCs) that are characterized by their ability to differentiate into the three germ cell layers.
While ESCs can differentiate into all the germ layer derivatives, the endoderm, mesoderm, and ectoderm, iPSCs are reprogrammed somatic cells generating pluripotent patient-specific cell lineages capable of aiding model human diseases3. Unlike iPSCs and ECSs, ACSs have a lower differentiation level, termed multipotent, and hence, differentiate into more tissue-specific stem cells4. ACSs are rare undifferentiated cells spreading within the body that transform from their quiescent state into proliferative and dividing cells when naturally dying cells are to be replaced.
Artificial Intelligence (AI) is a computer engineering and science advancing field already implemented in multiple disciplines such as home automation, robotics, health care, agriculture, banking, and transportation5. AI has achieved such momentum by mimicking human cognitive intelligence in multiple areas, including face and speech recognition. Currently, AI is further advanced by deep learning (DL) and machine learning (ML) in numerous domains like text analysis, autonomous automobiles, image classification, and medicine6. Its role in medicine has been tremendous due to its ability to analyze complex medical data and utilize meaningful connections within a dataset to produce results that can help in diagnosis and outcome prediction7.
In the context of artificial intelligence (AI), stem cell research has gained massive traction in recent years. AI has been utilized in multiple forms, including the DL algorithm, a subset of ML in which the human neural circuit is simulated into a digital multilayered neural network, convoluted neural networks (CNNs), capable of automatically obtaining the features of a specific image8. The branch of induced pluripotent stem cells (iPSC) in the field of regenerative medicine has mainly witnessed the merits of AI technology by facilitating the process of image classification through CNNs8. Basic data from image features are processed by CNNs, where the image passes through a series of convolutional layers followed by layers of feature extraction and, finally, the classification layers. These layers are responsible for identifying the basic structures, such as lines, blobs, and edges present in the image9. Therefore, in this review, we highlight the analysis of stem cell imaging under the field of AI covering multiple algorithms and techniques utilized in clinical settings. Moreover, the possibilities of application of AI in the characterization of various pathways of stem cells differentiation were also presented.
Review Question:
What are the available applications of AI-based imaging analysis for various types of stem cells?
Inclusion and exclusion criteria:
The inclusion criteria have been set based on the mnemonics PCC¹⁰ (Population, Concept, and Context). Additionally, the type of study, language of publication, and date will be considered.
Inclusion criteria:
- Studies involving any type of Stem cells (iPSCs, ECSs, and ACSs, PSCs)
- Studies using AI-based imaging analysis
- Published studies in any language and the full text is accessible
- No date restriction
Exclusion criteria:
- Studies investigating different types of cells rather than stem cells
- Studies using AI technology for other purposes than imaging analysis
- Preprints, reviews, and conference papers
- Full text not accessible