Phase Plate data at focus

Dear all,
I have recoreded phase plate data at 0 defocus (I know, I should have used at least a small decfocus, but I didnt), which will obviously go any CTF estimation crazy. However, if the images are really at focus, I dont need any CTF estimation / correction at all.
The problem is, not all images are perfect, some do have defocus, some astigmatism. Is there an easier way then manual curation?
My idea: I use CTFFIND (which is quick), followed by “manually curate exposures” to reject images that show any thon rings (good and bad, and the ones with strong signals from the ice). These images I would then process with CTF correction turned off.
But what about the mics that had thon rings with fits up to 3A? Can I use them as well, basically having a mixture of corrected and uncorrected mics?
Could I reimport the accepted mics (that are AT focus) with a constant CTF output and use them together with the mics at defocus and use CTF correction in later processing?

Best
Dario

At focus, accept, will be used w/o CTF correction

At defocus, reject but would like to keep w/ CTF correction (similar mics have more particles, less film)

At focus, astigmatism, reject

At defocus, astigmatism, reject

Ice, reject