A team including researchers from UC Berkeley hopes to spur contributions to Alzheimer’s disease research by developing an online game that enlists citizens as scientists.
In the game, Stall Catchers, players are tasked with scanning moving images of capillaries in a mouse brain to find the passages where blood flow has been “stalled.” Players’ annotations are then aggregated and cross-checked by experts, saving stretched researchers time.
“If we have a trained scientist who takes the data, it takes them over a week to analyze data that it takes less than two hours to acquire,” said Chris Schaffer, an associate professor of biomedical engineering at Cornell University. “If we had the help of the public in that analysis step, scientists could focus on collecting data and how it changed the number of stalls rather than spending lots of time on the analysis.”
Alzheimer’s patients suffer from reduced blood flow to the brain, which causes cognitive and memory problems that ultimately cannot be cured. Schaffer said that in earlier experiments, the team genetically engineered mice to have the disease and began recording interactions within the circulatory system.
The results of the game will help scientists gain more insight into the crucial 2 percent of stalled capillaries that account for a 30 percent reduction in blood flow.
“In animal models of the disease, through intervention and cognitive studies, they can reverse memory loss and depression associated with Alzheimer’s dementia,” said Pietro Michelucci, the project’s principal investigator and director of the Human Computation Institute. “This is the first time ever that anyone’s been able to do this in the lab.”
According to Michelucci, his team is packaging the analysis as a game on desktop and mobile platforms to make it more accessible to the general public. Players receive points for correctly identified stalls, which are recorded to their accounts and ranked on a global leaderboard.
While the game is currently in development from the testing phase, the most successful player has accumulated upward of 1.3 million points so far.
“If we can get … 10,000 people all playing this game, we can analyze this data much faster and we can get the result almost in the same amount of time as to collect data,” said Michelucci.
The scientists are hoping for the same level of success for Stall Catchers as was achieved by Stardust@home, another online game developed in 2006 by Andrew Westphal, a fellow at the UC Berkeley Space Sciences Laboratory. Stardust@home collected data from more than 30,000 players to help identify interstellar dust from 100 million searches — ultimately leading to the discovery of seven particles that may have originated in interstellar space.
“We found that this approach was very successful in identifying candidates (for) the first solid materials outside the solar system,” Westphal said. “So we are very optimistic that this will accelerate research in Alzheimer’s disease as well.”
Additionally, the researchers said human ingenuity is advantageous over computer algorithms in reading and processing data such as that of Alzheimer’s disease in animals.
“There are some things that humans do better than machines, and the challenge in science is to figure out which parts do machines do best and which do humans do best,” Michelucci said. “Machine-based classifiers can’t get the level of accuracy that (Cornell) needed, and you’ll see why using humans is a good idea.”
Contact Kimberly Nielsen and Justin Sidhu at firstname.lastname@example.org.