Imaging mass spectrometry (IMS) is quickly becoming a technique of reference to visualize the lipid distribution in tissue sections. Still, many questions remain open, and data analysis has to be optimized to avoid interpretation pitfalls. Here we analyze how the variation on the [Na+]/[K+] relative abundance affects the detection of lipids between sections of spinal cord of (uninjured) control rats and of models of spinal cord demyelination and traumatic contusion injury. The [M + Na]+/[M + K]+ adducts ratio remained approximately constant along transversal and longitudinal sections of spinal cord from control animals, but it strongly changed depending on the type of lesion. A substantial increase in the abundance of [M + Na]+ adducts was observed in samples from spinal cord with demyelination, while the intensity of the [M + K]+ adducts was stronger in those sections from mechanically injured spinal cords. Such changes masked the modifications in the lipid profile due to the injury and only after summing the signal intensity of all adducts and corresponding monoprotonated molecular ions of each detected lipid in a single variable, it was possible to unveil the real changes in the lipid profile due to the lesion. Such lipids included glycerophospholipids (both diacyl and aryl-acyl), sphingolipids, and nonpolar lipids (diacyl and triacylglycerols), which are the main lipid classes detected in positive-ion mode. Furthermore, the results demonstrate the sensitivity of the technique toward modification in tissue homeostasis and that the [M + Na]+/[M + K]+ ratio may be used to detect alterations in such homeostasis.