In addition to @Buffy's good answer, I would note that an important part of this decision will be based on the degree to which material from the prior survey article is included, since the person is already established as a co-author of the prior article.
- If significant portions of the prior article are effectively incorporated in this work, then the default position is for all prior authors to be co-authors, since they are co-authors of a subset of the work. In this case, the person should be a co-author.
- If the new work does not include significant portions of the prior work, then the default position is for authorship to be determined independently. In this case, it is your judgement call whether their contributions rise to the level of authorship or should be an acknowledgement.
In both cases, the person can no longer consent so, as @Buffy points out, consent will pass to their heirs. Usually, though, this won't be a hard decision for the heirs, as a good scientific publication is a lasting mark on the world that honors the memory of the deceased. I have experience with this myself (though I wasn't the one who negotiated with the heirs), and family of the deceased not only happily supported the publication, but went on to establish a standing scholarship in the area in the person's name.