In the rural parts of Uganda, lab technicians spend hours each day on thankless and seemingly unceasing work. The most common tests they run are for malaria. A technician smears a blood sample on a slide, treats it with dye, and then slowly scans it for cells that contain malaria parasites. She then uses a handheld clicker to record how many parasites she sees.
A typical test might take from 30 minutes to an hour. A health center might see dozens of patients in a day. When Manu Prakash visited on a recent trip, “There were places where the technicians couldn’t stop to talk to me, because they were busy working, which could last for eight to 10 hours per day,” he says.
An Indian-born biophysicist who works at Stanford University, Prakash is best known for creating the Foldscope—a $1 pocket microscope that magnifies objects by more than 2,000 times and can be folded from a single sheet of paper embedded with microoptics. But on this trip, while field-testing the Foldscope, Prakash realized that being cheap wasn’t enough. His devices also needed to be fast.
Rapid diagnostic tests can quickly check whether someone has malaria, but they don’t count the number of parasites. That figure is important: It reveals the severity of an infection and informs treatment choices. To count parasites, you need trained technicians and good microscopes. “There’s incredible talent, but it’s limited by their tools,” Prakash says. “I would meet health-care workers who would save their salary for a year to buy a fancier microscope.”
So Prakash and his colleague Hongquan Li built a fancier microscope—a high-speed, malaria-detecting device that they’ve called Octopi. It can automatically scan entire blood-smeared slides for malaria parasites, using a neural network trained on more than 20,000 existing images. Octopi works off a phone charger. It analyzes slides at speeds that are 120 times faster than traditional microscopy. Weighing fewer than seven pounds, it’s portable. And at a do-it-yourself cost of $250 to $500, it’s cheaper than many basic microscopes or other automated slide-analyzing devices.
Prakash has spent his career building extremely cheap medical devices that can be used in some of the poorest parts of the world. Besides the Foldscope, he developed a $10 skin patch that can detect parasitic worms. And he developed a 20-cent, hand-powered centrifuge that can spin medical samples at up to 125,000 revolutions per minute, achieving what costly, bulky, and expensive machines can do using little more than paper, string, and tape. But diagnostic speed was a new challenge.
At first, Li disassembled and reverse-engineered hundreds of DVD drives to try to build something that could scan slides quickly and efficiently. Eventually, he decided to fashion something from scratch. What he built was a fully modular microscope, with separate illuminating, scanning, and processing units that snap together magnetically. It looks quite unlike a standard microscope: There’s no eyepiece, for a start, nor a need for one. To use it, a technician prepares a slide in the usual way. “And then, you load the slide in the microscope and hit the scan button,” Li says.
The modular design makes the microscope very flexible. Technicians can switch between a low-magnification module that can efficiently find the parasites on a slide, and a high-magnification one that can more sensitively count them within those hot spots. They can also swap between different types of imaging, from the basic kind, in which white light shines through a slide from below, to more advanced techniques that look at the colors of samples treated with fluorescent dyes.
For malaria, the latter is crucial, because Li found that malaria parasites fluoresce in a slightly different color than surrounding blood cells. The distinction—roughly, teal versus blue—is hard to discern with the naked eye, but to Octopi, it’s clear. At first “we thought, That can’t be right,” Prakash says. “But the parasites do light up differently!”
The Octopi name has many origins. It’s a very loose acronym, which stands for “open configurable high-throughput platform for infectious diseases.” The microscope is very versatile, in the way that octopuses are. The “pi” ending, though the wrong plural form for octopus, is a nod to Raspberry Pi, a simple computer designed for use in developing countries. And, “my kids are of the age where they love octopuses,” Prakash says.
Prakash announced Octopi last month, to widespread acclaim online. His paper describing the project has been uploaded in advance of formal review and publication, and he is ushering the device into clinical trials in Peru, Uganda, and India. In the meantime, Elena Gómez-Díaz, a malaria researcher at IPBLN-CSIC in Spain, is impressed. “Malaria diagnosis is time-consuming,” she says. “This device automatizes the process at affordable cost. Whether it will replace the talent and hard work of so many skilled technicians working in malaria diagnosis in endemic countries, I can’t say. To me, they are heroes.”
Prakash feels the same. His goal isn’t to replace technicians but to make their lives easier—and not just when examining malaria. Octopi’s modularity means that it could be easily reconfigured to detect other diseases, too, by swapping in the right unit—an advantage it has over the many other technologies designed to spot malaria specifically. Prakash’s team has already used Octopi to look at the bacteria that cause tuberculosis and pneumonia and the parasites behind sleeping sickness and leishmaniasis. “It’s like a Swiss Army knife,” he says. “Anyone in the world can make a module.”
He hopes they do. Prakash has made Octopi’s assembly instructions and code freely available, and as he did with Foldscope, he is distributing 100 of the devices to researchers around the world, on the proviso that any data they collect are also openly available. The goal is the democratization of microscopy. Prakash envisions a future in which health workers worldwide diagnose infectious diseases with a network of cheap, automated microscopes, whose algorithms are constantly improved by the data collected by the entire community.
“If you like robotics, biology, and want to tackle these problems: join forces,” he said on Twitter. “Build DIY tools, apply them to problems you care about. Make new Octopi modules, teach someone, start [an] Octopi club. [The] microscopic world is for everyone.”