A team of scientists has developed a method for quickly finding COVID-19 in blood samples, using a deep learning system and a 3D-printed low-cost digital holographic microscope (DHM).
"We propose the use of a deep learning cell classification system using the reconstructed phase profiles output from a compact and field-portable shearing digital holographic microscope in the rapid screening of RBCs for COVID-19," the researchers explain of the work behind their letter.
"Features are extracted from the phase profile of segmented RBCs at each time frame of the reconstructed video data, then input into a bi-directional long short-term memory (Bi-LSTM) network to classify the cells based on their spatiotemporal behavior. The system is capable of successfully classifying between diseased and healthy samples without chemical processing and with fast turnaround times."
The 3D-printed prototype microscope is based around an off-the-shelf red laser diode, a one-dimensional translation stage, a 40x-magntification objective lens, a glass plate, and a CMOS image sensor. The laser illuminates the blood sample and a magnified image is projected onto the glass plate, forming an interference pattern captured by the image sensor as a video and sent to a computer for analysis.
The resulting video is fed into a neural network running on a workstation with an NVIDIA Quadro P4000 graphics processor, which proved capable of classifying a single patient's red blood cells in under 10 seconds — allowing it to quickly detect patients with COVID-19 infections with 80 per cent sensitivity and nearly 93 per cent specificity.
"The advantages of the presented system include cost, accessibility, and time to results," the researchers note, "but further investigations with larger patient pools including mild or asymptomatic infections are needed to validate the system performance more rigorously."
The team's work has been published in the journal Optics Letters under open-access terms, with more information available on IEEE Spectrum.