It maybe possible to estimate the gaps using Schlumberger modelling package Petrel. If the logs have gpas find a couple of neighbour los complete and ask Petrel to generate them in the loction of the incomplete log. That is a bit complex process because requires loading logs.

An alternative is to use AI and ask it to generate the gaps by training it with te existing logs. A simple shape generation based on a known pattern.

The step by step I leave it for you to find out