Can you elaborate for the sake of the OP, though? I mean, this post is about as non-committal as can be.
Yeah, that was on purpose. I've found that many non-radiology physicians have a very poor understanding of what radiologists do. People love their preconceived notions, and I didn't really feel like challenging those while this was posted in the general forum.
Most of the time when we see patients, it's because of procedures, but you can see patients really whenever you want. I try, although I often come up short, to introduce myself to several patients while they were still on the scanner. It's good for the specialty, and it gives one an opportunity to get even the smallest amount of additional history, which can be hugely helpful.
It's easiest for me to breakdown radiology patient contact by modality or subspecialty. Here's a list off the top of my head:
Neuro: diagnostic angiograms (angiograms with treatment is usually a neuro IR gig), lumbar punctures, spine biopsies, rarely discograms.
Pediatrics: fluoro (mostly UGIs, VCUGs, and enemas), ultrasound (most pediatric radiology that I know want to scan as opposed to trusting the technologists).
GI/GU: fluoro (esophagrams/UGIs, SBFTs, BEs, HSGs, occasionally VCUGs), CT-guided biopsies.
US: most people will trust their technologist on routine studies, but if I'm brought something out of the ordinary, then I definitely scan; saline sonohysterograms; US-guided biopsies (usually kidney, liver, or thyroid).
MSK: fluoro (joint injections, some places do facet injections, arthrograms), occasionally a CT-guided bone biopsy.
Chest: fluoro (sniff tests), CT-guided biopsies.
Nucs: lymphoscintigraphy, therapeutic radioiodine administration, bone scans (if there's a stress fracture, then I always tell the patient before they leave).
Mammo: a ton of biopsies (stereo, US-guided, MRI-guided), I also speak with every woman that comes in for a diagnostic examination; also, I do a focused breast examination prior to doing any breast ultrasound, and my findings of that examination as well as any relevant history or symptoms the patient reports go into my dictation. For all of these reasons, breast imaging is the closest thing that diagnoistc radiology has to being a clinic.
IR: should be obvious; I consider this basically to be minimally-invasive surgery.
There's not a single day of my radiology career where I haven't done at least one of the above. I'm not trying to delude anyone into thinking radiologists have a lot of patient contact. Most radiologists, myself included, don't want a lot of patient contact, so we're very happy with that aspect of our jobs. However, if that's your thing, then the opportunities are there. As you can tell, these are only opportunities for very short-lived contact dealing with a focused issue. When the woman with the breast lump starts talking about her blood pressure medication, I tell her to go see her PCP.