I'm confused how can we change register concatenation data (eg. OUTX_H&OUT_L: FD5Ch) into a 10-bit data (eg. X: -676*0.061=-41mg)? Where -676 comes from and why this number is negtive? [page 16 of AN4748 Application note]
Two's complement is used for signed number representation, please see this wiki page.
Retrieving data ...