@article{oai:tokyo-metro-u.repo.nii.ac.jp:00002537,
 author = {Yamazaki, Junpei and Yoshida, Makoto and Mizunuma, Hiroshi},
 issue = {8},
 journal = {Injury},
 month = {},
 note = {Introduction: We explored several modes of violent shaking using a dummy doll with an eyeball model to
reproduce abusive events that lead to retinal haemorrhages (RH) seen in shaken baby syndrome or abusive
head trauma (SBS / AHT).
Materials and Methods: A dummy doll equipped with an eyeball model was prepared. The eyeball model
was  lled with a model of vitreous body, i.e. agar gel or water, and was with a pressure sensor to measure
normal stress.
Results: The modes of shaking were classi ed into three patterns, i.e. fast shaking with the fore arms,
fast shaking with the whole arms and synchronized shaking with the whole arms. The frequency of the
cyclic acceleration-deceleration history experienced by the head of the dummy doll was 5.0, 4.0 and 2.2 Hz,
respectively, with the maximum acceleration of 20, 20 and 60 m/s2, respectively. We considered the last of
these three modes of shaking as possibly corresponding to the worst case of violent shaking. This mode of
shaking could be instructed to volunteers who acted as imitate perpetrators, and resulted in both increased
peak intensities of the acceleration experienced by the head of the dummy doll and increased stresses on the
retina at the posterior pole of the eyeball model.
Discussion: The time integral of the stress through a single cycle of shaking was 107 Pa s, much larger
than that of a single event of fall, which resulted in 60 to 73 Pa s. Taking into account that abusive shaking
is likely to include multiple cycles, the time integral of the stress due to abusive shaking can be even larger.
This clear di erence may explain why RH in SBS / AHT is frequent, while RH in accidental falls is rare.},
 pages = {1196--1206},
 title = {Experimental Analyses of the Retinal and Subretinal Haemorrhages Accompanied by Shaken Baby Syndrome / Abusive Head Trauma Using a Dummy Doll},
 volume = {45},
 year = {2014}
}