A four-month-old baby allegedly shaken to death showed "absolutely no evidence of trauma in the brain or other areas of the body", a court has been told.
The four appeals stem from the Angela Cannings case
Consultant neuropathologist Dr Waney Squier said Patrick McGuire had a "difficult" birth, but there had been no evidence of "imposed force" later.
Lorraine Harris, now freed, was jailed in 2000 for manslaughter of her son.
She is one of four people appealing against convictions for killing or harming babies by shaking them.
The test cases stem from a review of 297 cases ordered after Wiltshire woman Angela Cannings' convictions for murdering her two baby sons were overturned in 2003.
She was cleared after judges ruled no-one should be prosecuted solely on the basis of medical opinion that was disputed between experts.
If the four current appellants are cleared, dozens of other cases of "shaken baby syndrome" could be challenged.
FOUR TEST CASES
Raymond Rock, of Great Yarmouth, Norfolk, serving life for murdering his girlfriend's 13-month-old daughter, Heidi Davis, in 1998. He insisted she wriggled out of his arms and fell to the floor
Lorraine Harris, 36, of Long Eaton, Derbyshire, jailed for the manslaughter of her four-month-old son Patrick McGuire in 2000. She said the baby became ill and stopped breathing after a vaccination
Alan Cherry, convicted in 1995 of the manslaughter of his girlfriend's 22-month-old daughter Sarah Eburne-Day. He claimed she fell off a stool
Michael Faulder, 34, of Gateshead, jailed for two-and-a-half years in 1999 for causing grievous bodily harm to a boy aged seven weeks. He said he accidentally dropped the baby while trying to put him in his pushchair
Such a ruling could also affect hundreds of cases where a parent - usually the father - has been denied access to a child on the basis of allegedly violent treatment.
The Court of Appeal is hearing fresh evidence that "shaken baby" symptoms - swelling of the brain, bleeding behind the eyes and between the brain and skull - may not be caused by abuse.
They could be caused by a "far lesser threshold of force", such as short falls, vaccinations, difficult births or genetic conditions, it is being claimed.
Neuropathologist Dr Jennian Geddes told the court on Friday that when she began recent research into the syndrome she was surprised that brain tissue she examined "showed no evidence of trauma whatsoever under the microscope".
Michael Mansfield QC said on Thursday that far from being child killers or abusers, the four appellants had been "loving and supportive".
In each case jurors had been asked to rely solely on now-discredited expert evidence alleging that the children had been violently mistreated, he said.
The question the court faced was whether the convictions were safe when there were now alternative hypotheses to those relied on by the experts, he said.
Crown counsel Richard Horwell told the judges that Mr Mansfield's arguments were flawed.