Iron Oxide has been blamed for not being MRI Safe.
This has led, mostly for marketing purposes, of claims that "organic" pigments are safest!
However, there is evidence that iron-oxide free pigments can cause first degree burns of the eyelid during MRI. Read this article for a case report on a patient in Germany. This article was report on PubMed site.
November 2006, Volume 187, Number 5
First-Degree Burns on MRI due to Nonferrous TattoosTobias Franiel, Sein Schmidt and Randolf Klingebiel
We report the occurrence of a patient receiving first-degree burns of the eyelids in the context of a cranial MRI investigation. The patient, a 25-year-old woman, had a narrow, permanent cosmetic tattoo on the bilateral upper eyelids (so-called permanent eyelid make-up) that was free of ferric oxide pigment. She was admitted to the ward in the context of the diagnostic assessment of a chronic inflammatory CNS disorder. During the 1.5-T cranial MRI data acquisition of a HASTE sequence (i.e., an early acquisition of scout images using the HASTE sequence), the patient indicated that she felt a burning sensation of the upper eyelids. On immediate interruption of the investigation, a brief clinical examination was initiated. A laterally accentuated erythema was found on the bilateral upper eyelids. The patient was transported to the ward, at which point there was evidence of upper eyelid edema.
The aforementioned changes were confirmed independently by the technical MRI assistant, the neuroradiologist, the ward physician, and the head nurse. The patient did not consent to photographic documentation.
On dismissal, 3 days after MRI, both edema and erythema were in spontaneous remission, documenting structural and functional restitutio ad integrum and a first-degree burn.
The manufacturer's certificate of analysis of the chemical characteristics of the tattoo components identified various heavy metals (e.g., lead, copper, zinc, chrome, arsenic, cadmium, barium, and mercury) but no ferric oxides. The manufacturer denies previous MRIassociated incidents.
This case underscores the fact that in singular cases of diagnostic MRI data acquisition a burn can occur due not only to ferric oxide pigments [1-6] but also to tattoos free of ferric oxides. The national metrology institute PTB (Physikalisch-Technische-Bundesanstalt, Berlin, Germany) identified the cause as originating in a locally induced electric current. This can, generically speaking, be induced by any metal (not just ferromagnetic metals) moving through a high-frequency magnetic field . Material warming is a function of multiple parameters: the high-frequency impulse needed for the resonance, the specific heat coefficient of the material, the specific weight, the conductivity of the metal, the blood circulation, the ambient temperature, and the ventilation .
Although Tope and Shellock  reported that only 1.5% of patients with a permanent cosmetic tattoo (n = 135) experienced dysesthesia, the case we describe here shows that in singular cases first-degree burns can develop.
Because the ferric acid pigments are not necessarily the sole cause of noxious side effects as generally supposed, we suggest that the patient must be explicitly informed about the danger of such burns, even if the analysis certifies that the tattoo components are free of ferrous content. Furthermore, we suggest that a corresponding item be added to standardized MRI informed consent forms. Finally, in the case of insufficient diagnostic alternatives to MRI, sedative medication should be avoided and continuous visual and audio monitoring are mandatory to timely recognize hazardous side effects of cosmetic tattoos.