Nanny Trial: Exact causes of death still unclear

British experts supported the prosecution's case, saying Matthew Eappen's injuries could not have been inflicted accidentally and gone unnoticed for three weeks. A neurosurgeon at Great Ormond Street Hospital said that to diagnose shaken-baby syndrome a number of things must be looked for: other injuries, fractures, other evidence of abuse and bleeding at the back of the eye.

The prosecution made the case that the injury that killed Matthew had been sustained on 4 February, when he was alone with Woodward and could not have been an accident, such was the force required - equivalent to falling 15 feet. Evidence from brain scans and of retinal bleeding proved this, they said.

Robert Barnes, a neuroradiologist, said the injuries were caused "within minutes or hours" of brain scans he did shortly after Matthew's admission to hospital. Lois Smith, an ophthalmologist said she had seen signs of recent retinal bleeding which showed extreme force had been used on him - equalling nine on a scale of severity from one to 10. The type of bleeding found in nine-month-old Matthew's eyes, she added, was "very, very rarely" seen in accident cases.

Brain scans can help date injuries, said the Ormond Street neurosurgeon, because the blood found between the brain and brain lining - the dura - changes colour with the length of the injury. Recent injuries would show up red, whereas older ones would change colour to grey and clear.

During the trial, Joseph Madsen, a paediatric neurosurgeon at Boston Children's Hospital, said that when he started to operate first a clearish, yellowish liquid, followed by blood, spurted from Matthew's skull. He agreed it had the appearance of serum - which the defence claimed is indicative of an old injury.

Jan Leestma, a neuro-pathology specialist, also said his examination showed evidence that a blood clot existed between the brain and the dura - before 4 February and there was evidence of re-bleeding in the area. The defence also called America's leading expert on shaken-baby syndrome, Lawrence Thibault, who raised the fact that there appeared to be no other injuries. Any violent shaking, he said, would have produced "injury to the neck, the structure surrounding the spinal cord and the cord itself". None of these was found.

But experts in Britain said the injuries were so severe they could not have been caused accidentally. Brian Neville, a paediatric neurologist at the Institute of Child Health/Great Ormond Street said: "The impact would have to be pretty immediate. There could well have been a previous injury that was partially resolved but the idea of building up pressure over three weeks - they wouldn't really do this. It would just make the baby ill. Something occurred of some severity."

A Home Office pathologist, Paula Lannas, agreed: "From my assessment of the evidence this is a tragedy for this little girl but in my opinion I believe her to be guilty. This sort of case arises all the time but she was unlucky because she got captured by the system. I couldn't accept that the injuries were three weeks old, as the parents were both doctors, and a child sustaining that extent of injury could not have appeared to be normal for three weeks leading up to his death. It would be apparent that something was wrong."