Researchers say artificial intelligence shows that continued use of face masks by half of the population will help prevent large-scale coronavirus infections in Tokyo for the time being.
A group led by Nagoya Institute of Technology Professor Hirata Akimasa used AI to calculate possible infection trends for Japan's capital.
It made projections assuming that no new highly infectious variant will appear and that crowding will gradually return to levels before the pandemic.
The study comes just before Japan's government eases mask guidelines on Monday and downgrades the classification of COVID-19 on May 8.
It shows that if half of the people continue to wear masks, the number of daily new infections in Tokyo will reach roughly 5,500 in mid-May.
The group says the number will likely increase after the spring holiday season from late April through early May, but no major spike is expected.
But the study shows that the daily number will soar to about 8,600 in mid-May if 80 percent of people stop wearing masks.
The group says that in both cases, the number will rise to the same levels after the "bon" holiday in mid-August.
It says without additional vaccinations, the daily number will surge toward the year-end and top 10,000 a day as people's immunity levels fall.
Hirata says his team's study shows that the infections will unlikely spread on a large scale for the time being if a certain proportion of people continue to wear masks.
He calls on people to be considerate of those at high risk, such as senior citizens.