Search
Menu

Mobile Device Uses Holographic Analysis, Deep Learning to Detect Parasites

Facebook X LinkedIn Email
Researchers at the UCLA Henry Samueli School of Engineering have developed an inexpensive, portable platform that can detect motile parasites in bodily fluids rapidly. Using the new platform, more than 3 milliliters (ml) of a bodily fluid sample can be imaged and analyzed within 20 minutes, providing a throughput that is orders of magnitude better than traditional optical microscopy-based examination.

High-throughput bodily fluid screening device, which screens and analyzes ~3.2 mL of fluid sample within ~20 min. UCLA Samueli School of Engineering.
High-throughput bodily fluid screening device, which screens and analyzes ~3.2 ml of fluid sample within ~20 min. Courtesy of Y. Zhang et al.
Light: Science & Applications, DOI:10.1038/s41377-018-0110-1.

This label-free computational imaging platform automatically detects and counts motile microorganisms using holographic speckle analysis. It records high-frame-rate videos of the holographic patterns of the samples illuminated with laser light. Then, a motion analysis algorithm analyzes these captured videos at the microscale and converts the locomotion of the target parasites within the sample into a signal spot, which is detected and counted using artificial intelligence.

“Our platform can be considered as a motion detector in the microscopic world, which locks onto any moving objects within the sample,” said researcher Yibo Zhang. “Locomotion is used as both a biomarker and a contrast mechanism to distinguish parasites from normal cells.”

The research was led by Aydogan Ozcan, Chancellor’s Professor of Electrical and Computer Engineering at UCLA and the associate director of the California NanoSystems Institute at UCLA, along with professor Kent Hill.


“Although motility is a common feature of various parasites and other disease-causing microorganisms, its use as a fingerprint for diagnosis is highly underexplored,” Ozcan said.

Proof of concept was demonstrated using Trypanosoma parasites, which cause sleeping sickness and Chagas disease. Using a holographic speckle analysis algorithm combined with deep learning-based classification, the researchers demonstrated label-free detection of trypanosomes within spiked whole blood and artificial cerebrospinal fluid (CSF) samples, achieving a limit of detection of 10 trypanosomes per ml of whole blood (about five times better than current state-of-the-art parasitological methods) and three trypanosomes per ml of CSF.

This improved detection limit could enable earlier detection of sleeping sickness and Chagas disease, improving cure rates and reducing prevalence of these diseases. The researchers also demonstrated the use of their device to detect Trichomonas vaginalis, demonstrating that their technique is applicable to various parasites and motile microorganisms.

The device is compact and lightweight, and the cost of the prototype is less than $1850, which could be reduced to less than $800 when manufactured in larger volumes, the researchers said. “Thanks to its high sensitivity, ease-of-use, reduced cost, and portability, we believe our technique can improve parasite screening efforts, especially in resource-poor areas and endemic regions,” said researcher Hatice Ceylan Koydemir. 


The research was published in Light: Science & Applications (https://doi.org/10.1038/s41377-018-0110-1).  


Published: December 2018
Glossary
holography
Holography is a technique used to capture and reconstruct three-dimensional images using the principles of interference and diffraction of light. Unlike conventional photography, which records only the intensity of light, holography records both the intensity and phase information of light waves scattered from an object. This allows the faithful reproduction of the object's three-dimensional structure, including its depth, shape, and texture. The process of holography typically involves the...
deep learning
Deep learning is a subset of machine learning that involves the use of artificial neural networks to model and solve complex problems. The term "deep" in deep learning refers to the use of deep neural networks, which are neural networks with multiple layers (deep architectures). These networks, often called deep neural networks or deep neural architectures, have the ability to automatically learn hierarchical representations of data. Key concepts and components of deep learning include: ...
Research & TechnologyeducationAmericasUCLAHenry Samueli School of EngineeringAydogan OzcanMicroscopyOpticsSensors & Detectorsoptical sensorsmedicalBiophotonicsportable devicesholographycomputational imagingdeep learningholographic analysisTech Pulse

We use cookies to improve user experience and analyze our website traffic as stated in our Privacy Policy. By using this website, you agree to the use of cookies unless you have disabled them.