Two years ago, Robert J. Wilkes of Missoula was a single father who had just won custody of his 3-month-old son, Gabriel.
Today he stands convicted, despite his denials, of knowingly killing his own child.
"The extent of the damage and the custody of the child - there was no other possibility," juror Allan Oines said after the trial.
Wilkes cried at trial when authorities played a 9-1-1 recording of the desperate call to report Gabriel's grievous injuries. An earlier Missoulian report quoted his insistent denials.
"It makes me angry that my son is dead," he testified. "It makes me angry that I'm the one accused of his death. I don't know who did it. But I had no reason to get mad at my son. He was 3 months old."
Multiple doctors pointed to bleeding in the brain and eyes visible by medical imaging. Their diagnosis: abusive head trauma, more commonly known as shaken baby syndrome. This conclusion and his own testimony led authorities to charge Wilkes - and not another suspect - as the child's killer.
Wilkes' story of what happened Oct. 4, 2008, never changed.
He spent the day moving between apartments in the same complex, leaving Gabriel with a neighbor. He chatted with her when he returned and fed his son a bottle of formula. At home, he laid Gabriel on the floor to rest, grabbed a drink from the kitchen and turned on the television.
Wilkes said when he checked back a few minutes later, he found Gabriel vomiting from his nose and mouth. Unable to find his cell phone, he rushed next door and cried for the sitter to call 9-1-1.
She saw him holding an unresponsive, limp baby who was struggling to breathe. She followed the dispatcher's directions and performed CPR as they waited for an ambulance.
ER physicians at Missoula's Community Medical Center treated the child initially, but soon evacuated him to a Spokane hospital specializing in pediatric medicine and better equipped to care for Gabriel's severe neurological injuries.
But nothing worked. Doctors feared the child would remain forever in a vegetative state, so Gabriel was eventually removed from life support.
Wilkes was sentenced Wednesday to 40 years in state prison for deliberate homicide, with 10 years conditionally deferred.
At first, Deputy Missoula County Attorney Suzy Boylan wasn't sure she could build a case against Wilkes. With no witness, no motive, no history of abuse and no discernable guilty behavior, the case would hinge on a jury's ability to deduce guilt or innocence from indirect evidence.
"Child abuse cases are by nature circumstantial," Boylan said. "In 11 years, we've only had two, maybe three, shaken baby cases that could be charged."
The Spokane physicians who treated Gabriel and the medical examiner who conducted his autopsy found bleeding in his brain and all four quadrants of his eyes, Boylan said. The sudden signs of brain damage were the result of "non-accidental trauma," court records said.
But the medical testimony alone wasn't quite enough. It would be difficult to prove - beyond a reasonable doubt - that the injuries occurred while in the father's custody and not the baby sitter's care, since she supervised Gabriel just an hour earlier.
Under questioning by Missoula Police Detective Dean Chrestenson, Wilkes continued to assert his innocence. Gabriel's attending physicians would testify, but Boylan needed a pediatric specialist to fit the pieces together in a way jurors could understand.
Montana's only pediatric neurologist had recently retired, so Boylan turned to Dr. Rich Kaplan, a certified child abuse pediatrician from Minnesota. Kaplan was a familiar name, having trained Montana lawyers, investigators and child protection workers at workshops and conferences.
Kaplan's analysis of the medical records, combined with one crucial detail from Chrestenson's investigation, gave Boylan the key to her case.
"They said it's all about the bottle," she said.
Both Wilkes and the baby sitter confirmed that Gabriel fed from a bottle just before his father took him home.
It was "almost inconceivable," Kaplan eventually testified, that the child could have fed effectively after sustaining such severe injuries. He said the nerves that tell the body when to eat, breathe and sleep had been violently sheared and it had happened under the father's care.
With that, Boylan had her case.
"Brains just don't go kablooey without a cause," she said.
Wilkes was charged in late March 2009, nearly six months after the chilling 9-1-1 phone call.
Since pediatric radiologist Dr. John Caffey first described a "triad" of brain and eye injuries as shaken baby syndrome in 1974, investigators and prosecutors have seized the opportunity to catch previously elusive child abusers and killers.
But while medical evidence of violent trauma has helped convict thousands, it also can lead to wrongful convictions.
Scientists' understanding of what exactly causes these traumatic injuries evolved the way science does: with theories, questioning and research. Over time, serious uncertainties developed about whether the combination of intracranial bleeding, retinal hemorrhages and brain swelling exclusively resulted from forceful shaking.
They learned these particular abusive injuries can be difficult to distinguish from ones caused by some blood disorders, vitamin deficiency and other nonviolent means.
"A certain percentage of babies get just a little bit of blood (between the skull and brain) just from being born," said Kaplan, referring to a report last year by British researchers evaluating newborns who had not yet left the hospital.
But Kaplan questions some of the research contradicting the diagnosis. He says a thorough analysis by a well-trained abuse expert considers more than the three variables once thought conclusive. Experts also should look for specific forms of bone fractures, evaluate the placement and contour of skull injuries, and test all potential causes.
By itself, biomechanical research is limited by the impossibility of re-creating rage and other actual conditions of abusive head trauma in a lab, Kaplan said.
"All this research shows so far is you can't shake a crash test dummy hard enough to hurt a baboon," he said.
No one claims child abuse is a myth. But scientific controversies are rarely resolved in courtrooms where jurors, judges and lawyers - who often have little scientific expertise - must determine guilt or innocence.
And each case is different.
Boylan learned about the scientific debates as a staff attorney for the National Center for the Prosecution of Child Abuse and carefully avoids using the term "shaken baby syndrome" in court.
She and Kaplan said that without careful analysis and an updated understanding of the diagnosis, the signs of abuse also could be mistaken for nonviolent ones and certainly do not limit the cause to shaking only.
The American Academy of Pediatrics encourages physicians to thoroughly investigate alternative causes before issuing their final reports because of the "enormous social, psychological and legal implications for families."
"People do get convicted on bad science and it's terrible," said Kaplan. "If I give a diagnosis on a child, it's a medical diagnosis, not a legal diagnosis."
Yet the medical diagnosis is a crucial linchpin in cases with no witnesses and little, if any, corroborating evidence.
"I call it medically diagnosed murder," DePaul University law professor Deborah Tuerkheimer said in a telephone interview.
Her research into three decades of shaken baby cases suggests some of those convicted in the past might not be convicted today, given the progression of research, scientific standards and legal understanding. And, she thinks some people wrongfully convicted could be actually innocent.
"To date our system has failed," Tuerkheimer wrote in a 2009 study for the Washington Law Review.
Some U.S. courts have conceded wrongful convictions in shaken baby cases. Defendants have been acquitted.
Boylan said she acknowledges the changed understanding and will continue to prosecute well-investigated cases like Wilkes'.
"I would have softened a sentencing recommendation if he had admitted that he had actually did it so it could be used for research," Boylan said.
She recommended the judge sentence Wilkes to life, with parole eligibility after 30 years.
Not all of the jurors in Wilkes' trial were easily convinced of his guilt.
"The holdouts were two men who just could not believe that anyone who would ask for (custody of) their child would want to do that to him," said juror Oines.
Wilkes' attorney, public defender Scott Spencer, emphasized his client's consistent denials during the trial. But he offered little challenge to the prosecution's medical testimony other than to say that, by itself, it did not warrant a conviction. No medical experts testified for the defense.
Spencer declined to discuss the case for this story, but the Missoulian quoted his succinct summary of the defense's case during the trial.
"Is it suspicious? Absolutely," Spencer told the jury. "Is it a tragedy? Absolutely. But (Wilkes) didn't do it. There's not one person who has walked into this courtroom and said, ‘I know what happened.' Something happened. Somebody did something. But you have to be convinced beyond a reasonable doubt."
The infant's injuries did not fit with Wilkes' description of events, but the defendant presented no other possibilities, Boylan said.
She asked the jurors to consider "Gabe" and weigh the evidence presented by her 11 medical witnesses. Their testimony was crucial, she said after the trial.
"I had all these incredible facts and a row of doctors," Boylan said.
Wilkes cried at his sentencing hearing and told Judge Ed McLean he misses his son.
McLean later said in his judgment that Wilkes is a threat to the community and emphasized that the defendant is not taking accountability for his crime.
"Although you may feel remorse for Gabriel's death, you do not accept responsibility," McLean said.
Jayme Fraser is a junior studying print journalism at the University of Montana who is interning at the Missoulian this summer. She can be reached at 523-5241 or at firstname.lastname@example.org.