Dear Editor‐in‐Chief,I would like to comment on the article by Jino Bak, Jin Hwa Choi, Jae‐Sung Kim, and Suk Won Park, “Modified dose difference method for comparing dose distributions,” J. Appl. Clin. Med. Phys. 2012; 13(2): 73–80.Something almost identical to their method was in fact first described by Bakai et al.(
) who called their dimensionless index χ. While it is true that Bak et al. refer to this work in Reference 13, the authors do so in a list of variations on the γ index. However, γ indices involve a search over dose and spatial coordinates, while χ is a gradient‐dependent function like theirs, and I do not think that they properly acknowledge the similarity. The difference between their “modified dose” index and χ lies in the dimensionless parameter β, described in their Eq. (1) as:
where and are distance‐to‐agreement and dose difference criteria (my symbols), and ▽D is the gradient of the reference dose distribution. Where the authors define a modified dose difference,Bakai et al. would define the index χ as
for an actual dose difference δD at a point (x,y,z), when their Eq. (6) is cast in the above terminology. In addition, Bakai et al. provided a convincing theoretical basis for the factor, rather than (), in terms of the minimum distance between two surfaces in (x,y,z,D) space when measured in appropriate units.To be fair to the authors, Bakai et al. did not actually define a modified dose difference , nor the weighting factor explicitly, so the similarity could easily have been missed. I realize it is a bit unseemly to reference one's own work, but I did make that definition myself in connection with evaluating 3D dose differences.(
)Sincerely, Will Ansbacher