Can I avoid the sequential scan?

I run the following select (Online 5.07.UC1) : select * from dates_table where date(date_field) between"050596" and "050596" date_field is defined as datetime year to second and is indexed. I get a sequential scan, which kills performance. I'm sure this is because of the "date" function in the select. Can I achieve the same result with a different select, this time encouraging the optimiser to use the index?