I have always maintained that seismic interpretation entails the derivation of a plausible geological model, obviously limited to what a seismic image can provi

I have always maintained that seismic interpretation entails the derivation of a plausible geological model, obviously limited to what a seismic image can provide. On the other hand, insight in the local geology and knowledge about the physical properties of ambient rocks and fluids are key in narrowing down to "plausible" solutions. Similarities with interpretation of radiology seem obvious. In the last years, a fair number of seismic interpretation articles have appeared, reporting on the use of AI-like techniques, more focused on 2 & 3D pattern recognition rather than on interpretation in low signal-to-noise environments, as described below. Evidently, only success stories are reported, but it is safe to conclude that AI-like techniques definitely are an addition tool in the evaluation of the subsurface. Hence I would presume that something similar eventually will take place in the development of interpretation of radiology data. After all, geophysical techniques and physical methods for medical purposes have both benefited from each other since many decades