In-vitro blood purification using tiny pinch holographic optical tweezers based on deep learning

Area of Science:

• Biomedical Engineering
• Medical Devices
• Cell Biology

Background:

• In-vitro blood purification is critical for medical treatments, necessitating precise component separation.
• Current methods for blood component extraction lack a general, on-demand approach.
• Efficient and accurate blood filtering remains a significant challenge in medical applications.

Purpose of the Study:

• To introduce the first programmable non-contact blood purification system.
• To enable accurate detection and extraction of diverse blood components.
• To overcome limitations of existing blood purification techniques.

Main Methods:

• Development of a novel blood component object detection dataset and deep learning models.
• Fabrication of a microfluidic chip for controlled fluid handling.
• Implementation of a customized holographic optical tweezer for non-contact cell manipulation.

Main Results:

• Real-time blood fractionation with high precision (up to 96.89%) achieved.
• Demonstrated high efficiency in detecting and extracting target blood components.
• Validated the system's capability for accurate cell and particle identification.

Conclusions:

• The developed system offers a programmable, non-contact solution for blood purification.
• High precision and efficiency in blood component separation were empirically demonstrated.
• The system's scalability and flexibility present new avenues for blood treatment research.