Researchers from the Centre for Genomic Regulation (CRG), the University of the Basque Country (UPV/EHU), Donostia International Physics Center (DIPC), and the Fundación Biofisica Bizkaia (FBB) have developed an artificial intelligence (AI) tool called AINU (AI of the NUcleus) that can differentiate cancer cells from normal cells and detect early stages of viral infection inside cells. The findings, published in the journal Nature Machine Intelligence, have significant implications for improved diagnostic techniques and disease monitoring strategies.
AINU utilizes a special microscopy technique called STORM to obtain high-resolution images of cells, capturing finer details than traditional microscopes. With nanoscale resolution, the AI can detect rearrangements inside cells as small as 20nm, enabling the recognition of specific patterns and differences in how DNA is arranged within cells. This capability allows for the early detection of alterations, providing doctors with valuable time to monitor diseases, personalize treatments, and improve patient outcomes.
The AI tool, based on a convolutional neural network, analyzes visual data like images and has been successfully used in various applications such as facial recognition. In the medical field, convolutional neural networks are employed to analyze medical images and identify signs of cancer or abnormalities in MRI and X-ray scans, aiding in faster and more accurate diagnoses.
To train AINU, researchers fed it with nanoscale-resolution images of cell nuclei from different types of cells in various states. The AI learned to recognize specific patterns by analyzing the distribution and arrangement of nuclear components in three-dimensional space. By identifying distinct changes in nuclear structure, such as alterations in DNA organization or enzyme distribution, AINU can classify new images of cell nuclei as cancerous or normal.
The AI’s nanoscale resolution also enables the detection of viral infections within an hour of infection. By identifying slight differences in DNA packing, AINU can detect the presence of viruses and monitor how they affect cells immediately after entry. This capability could aid in the development of better treatments, vaccines, and faster diagnostic processes in hospitals and clinics.
While the technology shows promise, there are limitations to address before clinical implementation. STORM imaging requires specialized equipment typically found in biomedical research labs, and the throughput of analyzing only a few cells at a time needs to be increased for diagnostic purposes. However, advancements in STORM imaging may soon make the technology more accessible and applicable in clinical settings.
In the short term, AINU is expected to accelerate scientific research by accurately identifying pluripotent stem cells, which have the potential to develop into any type of cell in the body. This capability could contribute to safer and more effective stem cell therapies, reducing the reliance on animal testing.