I could be very well wrong, but those look a lot like Newton's rings, which can be caused by two layers of dissimilar materials that are normally in close contact expanding or contracting differently, making one of them buckle slightly. And by slightly, I mean that the spacing between them varies by fractions of a wavelength of light. (Another, similar possibility is differential stress polarization effects, again with a thermal basis. It's really hard to tell in B&W, though.) The "origin" of the pattern probably indicates a relative "hot spot". If the camera is shot cold, or allowed a little while to warm up completely, you probably wouldn't see a problem, but if it's been on just long enough to partially warm up, there might be enough difference in temperature across the sandwich that is the sensor, its lenses and filters, to cause a visible pattern.