To The Editor: I was dismayed at the cover of the January 1999 issue of Pediatric Research, which showed three images of an infant brain obtained with medical optical imaging, CT, and MRI. The MRI and CT images are mislabeled. This was nothing compared with the disappointment I experienced when I actually read the related article (Pediatr Res 45:54-59, 1999).

The authors evaluated the ability of the new technique of medical optical imaging (MOI) to detect intracranial hemorrhage in newborns and compared their findings to standard imaging modalities. Since there is no radiologist listed as a co-investigator, I can only assume that the comparison was made retrospectively with written reports and not the actual images. This is clearly the case when one looks at the images presented. Figure 3A is labeled as showing a left germinal matrix hemorrhage. To the best of my ability, I cannot identify such a hemorrhage. Figure 4B is labeled as showing a large hemorrhage on the right side. At best, a small germinal matrix hemorrhage is suggested on this image. Figure 5A purportedly shows large bilateral hemorrhages, confirmed by Figures 5B and 5C. Again, at best, these reveal small intraventricular clots. One can argue that the hemorrhages progressed in the 3-d interval between studies. Yet, if this is the case, it suggests that these comparisons are not quite valid. The authors state in the discussion that in CT and MRI, intravenous contrast materials are frequently required for these studies. Although the authors are correct in the most general of terms, intravenous contrast is almost never used in the evaluation of newborns with intracranial hemorrhage. They further suggest that indocyanine green is nontoxic, unlike radiation-absorbing x-ray or CT contrast agents. Currently used non-ionic contrast agents have a similar long track record of safety, especially in the pediatric age group.

In a final statement, the authors write that currently used imaging techniques such as MRI, CT, and Doppler are inadvisable in critically ill patients because they are not portable, are difficult, involve ionizing radiation, and are static. Neither MRI nor Doppler ultrasound use ionizing radiation, Doppler ultrasound is both portable and provides dynamic information, and none of the techniques are particularly difficult to perform or interpret.

As a member of the Society for Pediatric Research, I was embarrassed by this article. It overstates the potential advantages of medical optical imaging and grossly overstates the risks and disadvantages of current imaging techniques. As imaging tools become more prevalent in pediatric research studies, rigorous evaluation of their strengths and weaknesses becomes key to the credibility of a journal such as this.

George A. Taylor, M.D.

Department of Radiology; Children's Hospital; Boston, MA 02115