SJC Upholds Abusive Head Trauma Testimony in Infant Death Case: Commonwealth v. Shu Feng Hsu (SJC-13855, September 9, 2026)

The Supreme Judicial Court affirmed the conviction of Shu Feng Hsu for involuntary manslaughter in the death of her eleven-month-old niece, Chloe Chen. The court held that the evidence was sufficient to support the conviction and that the trial judge did not abuse his discretion by admitting expert testimony on abusive head trauma, so long as experts were limited to stating that the victim’s injuries were consistent with the diagnosis rather than asserting it as the cause of death.

The Facts

In February 2018, Chloe lived on the second floor of a Quincy home with her parents and grandparents. Hsu and her husband, Chloe’s uncle, lived on the first floor with their two-year-old daughter. Several days per week, Hsu cared for both children while the other adults worked.

On February 15, 2018, Chloe’s grandmother and father cared for her in the morning. Surveillance cameras installed in the home showed Chloe as a healthy, active infant. She ate, watched television, cruised in her walker, and babbled. Around 2 or 2:30 P.M., Chloe’s father left for work and Hsu took over as caregiver. Hsu took both children to her bedroom for a nap.

At approximately 4:20 P.M., a living room surveillance camera captured the sound of a baby crying intermittently for about five minutes. At around 4:25 P.M., the camera recorded eleven thuds occurring over approximately one minute. The thuds were interspersed with the baby’s cries. After the eleventh thud, the crying slowly stopped and was replaced by coughing and gagging sounds.

At 4:29 P.M., Hsu called 911. She told the operator that Chloe seemed soft and was having difficulty breathing. A Quincy police officer arrived six minutes later and noticed that Chloe’s breathing was shallow and that she appeared to be seizing. Paramedics observed a small bruise or abrasion on Chloe’s forehead.

Chloe was taken to Boston Medical Center. Dr. James Holsapple, the neurosurgeon who treated her, observed that she was unconscious and seizing with nonreactive pupils. A CT scan showed an acute subdural hematoma covering most of the left hemisphere of her brain, an eight-millimeter midline shift, and swelling. Holsapple performed emergency brain surgery approximately one hour and fifteen minutes after Chloe’s arrival. He found a large volume of high-pressure, acute blood beneath her skull. Despite the surgery, Chloe became unstable and died on February 17, 2018.

An autopsy revealed multiple injuries not attributable to medical intervention, including bruises on Chloe’s head, left forearm, and torso. The autopsy also found multiple brain and retinal hemorrhages, as well as evidence of one brain hemorrhage that appeared to be weeks or months old.

The Investigation

On February 16, 2018, Hsu participated in a police interview. She said that Chloe cried a lot but denied that the crying frustrated her. She told officers that Chloe had been sleeping and suddenly clenched in a way that suggested a seizure. Hsu denied touching, dropping, or shaking Chloe, noting that if someone had struck her, there would have been a bruise or blood.

Police seized memory cards from surveillance cameras in and around the home. Although there was a camera in Hsu’s bedroom, there was no memory card in it. Hsu told police she had sold the card before February 15, though she could not remember to whom. The living room camera’s memory card was hidden inside a small egg. Hsu said she had removed it at her husband’s request and that her two-year-old daughter grabbed it and hid it.

Police arrested Hsu at her home on November 20, 2018. After searching for twenty minutes, they found her hiding in a bedroom closet.

The Daubert-Lanigan Hearing

Before trial, Hsu filed motions in limine to exclude expert testimony regarding abusive head trauma and requesting a Daubert-Lanigan hearing. She argued that the theory was no longer generally accepted in the relevant scientific community.

At the hearing, Hsu’s expert, Dr. John Galaznik, testified that there was no biomechanical, animal, or human experimental data confirming that shaking would produce primary brain injury, subdural bleeding, or retinal hemorrhaging in a twenty-pound infant. He also testified that research showed a short fall of as little as two to three feet could be capable of inducing subdural bleeding and the findings in this case.

The Commonwealth’s expert, Dr. Kimberly Schwartz, explained that abusive head trauma is a medical diagnosis identified through a differential diagnosis procedure. She described a 2018 study finding that the short fall mortality rate for children under five years old is 0.48 per one million annually.

On the day before trial, the judge issued his order. He determined that abusive head trauma was generally accepted as a reliable medical diagnosis and that expert testimony regarding it was admissible. He limited experts to stating that Chloe’s injuries were “consistent with” abusive head trauma, not that abusive head trauma “was the cause” of her injuries. He also prohibited experts from testifying “that short falls cannot cause [abusive head trauma].”

The Trial

At trial, the Commonwealth presented expert testimony from three physicians. Dr. Anand Shah, a medical examiner, directed the jury’s attention to autopsy photographs showing bruises on Chloe’s head, face, and left forearm. He testified that Chloe’s death resulted from multiple significant blunt force impacts to the head that had been recently inflicted. Shah noted that the injuries were not consistent with bumping her head on a crib or a short fall of less than three feet.

Dr. Christina Stanley, a neuropathologist, opined that Chloe’s subdural hematoma was the result of significant trauma. She testified that it was inconsistent with a rebleed of the older hemorrhage because the older hemorrhage was too thin to have caused the volume of bleeding that occurred.

Schwartz, a child abuse pediatrician, reviewed Chloe’s medical records and conducted a visual examination before completing a differential diagnosis. She concluded that Chloe’s injuries were “consistent with abusive head trauma.” She opined that the subdural hematoma was inflicted shortly before Chloe presented to the hospital and that the level of force necessary to inflict it was force not from normal infant care.

Hsu presented two expert witnesses. Dr. Janice Ophoven, a pediatric forensic pathologist, testified that Chloe “suffered a traumatic brain injury due to blunt-force trauma associated with chronic and acute subdural bleeding.” She explained that children with chronic subdural fluid can suffer an immediate collapse from falling and banging their head on a crib or floor. Ophoven also testified that the significant amount of blood documented during the autopsy was due to surgical intervention rather than blunt force impact.

Galaznik testified that Chloe’s injuries indicated preexisting subdural bleeding alongside some new bleeding. He opined that the injuries could have been the result of rebleeding of a prior injury after a trivial impact or bump to the head, and that the bump could have occurred one to seven days before Chloe was taken to the hospital.

At the close of the Commonwealth’s case, Hsu moved for a required finding of not guilty. The judge allowed the motion with respect to murder in the first degree on a theory of deliberate premeditation but otherwise denied it. The jury found Hsu guilty of involuntary manslaughter on theories of battery and wanton or reckless conduct. The judge sentenced her to five to seven years in state prison.

What the Court Held

The court first addressed whether the evidence was sufficient to convict Hsu of involuntary manslaughter. The standard requires viewing the evidence in the light most favorable to the Commonwealth and determining whether any rational finder of fact could have found each element beyond a reasonable doubt.

The court found the evidence sufficient. Chloe was healthy on the morning of February 15, 2018, and Hsu was her sole caretaker that afternoon. The surveillance recording captured a baby crying, eleven distinct thuds interspersed with crying, and then coughing and gagging sounds, followed immediately by Hsu’s 911 call. The difference in Chloe’s health before and after the thuds, combined with their repetitive nature, supported a reasonable inference that Hsu injured Chloe.

Expert testimony further supported this inference. Shah testified that Chloe died from blunt force head injuries caused by significant, repeated impact. Stanley testified that the subdural hematoma resulted from significant trauma, not a rebleed of an older injury. Schwartz testified that the injuries were consistent with abusive head trauma inflicted shortly before Chloe arrived at the hospital.

While Hsu’s experts offered alternative theories, the weight of competing expert testimony was a matter for the jury. The jury could have reasonably inferred that Chloe’s injuries resulted from Hsu inflicting blunt force that constituted a battery endangering her life or wanton or reckless conduct creating a high likelihood of substantial harm.

The Abusive Head Trauma Testimony

The court next addressed whether the trial judge abused his discretion by admitting expert testimony on abusive head trauma. The judge had concluded that abusive head trauma is generally accepted by physicians and the medical community as a diagnosis reached through differential diagnosis. He relied on guidance from various organizations and peer-reviewed consensus statements designating it as a valid medical diagnosis.

The court held that the judge properly determined that abusive head trauma is generally accepted as a valid medical diagnosis. Numerous national and international medical societies and organizations recognize it. A 2016 survey of children’s hospital physicians with over 600 respondents found that ninety-three percent indicated that abusive head trauma is a valid evidence-based diagnosis. The diagnosis is made like any other medical diagnosis, by considering clinical history, physical examination, and laboratory and imaging data.

Although Hsu provided literature describing shortcomings of abusive head trauma and shaken baby syndrome diagnoses, unanimity of opinion is not essential under the general acceptance test. The judge had broad discretion to weigh the Daubert-Lanigan factors. The sources Hsu cited primarily disputed whether shaking alone can cause the injuries, but abusive head trauma encompasses multiple mechanisms of injury, including shaking, blunt impact, or a combination.

Hsu also argued that Schwartz’s testimony did not assist the jury because the medical examiner had already testified that Chloe died from blunt force impacts. The court disagreed. An expert opinion concerning the mechanism likely to have caused the injuries was relevant and helpful to a lay juror who would not understand the significance of injuries such as subdural hematomas or midline shifts. Schwartz’s testimony provided important context about timing, symptoms, and the level of force associated with the injuries.

The court rejected Hsu’s contention that the word “abusive” came close to an opinion on the ultimate issue of guilt. While an expert may not provide an opinion on innocence or guilt, there is no prohibition on testimony that touches the ultimate issue. Schwartz’s testimony was confined to the mechanism by which the injuries were suffered but did not speculate as to the circumstances or ultimate responsibility for those injuries. She never offered an opinion as to Hsu’s innocence or guilt.

The judge imposed appropriate limits to mitigate any prejudicial effect. He prohibited experts from stating that abusive head trauma was the cause of Chloe’s injuries and confined them to stating that the injuries were “consistent with” abusive head trauma. This prevented expert witnesses from encroaching on the jury’s ultimate responsibility. Hsu had multiple opportunities to challenge the testimony through cross-examination and by providing her own expert testimony. The judge also instructed the jury that, although they could weigh expert testimony in their fact finding, they were required to reach their own conclusions on guilt or innocence.

The Short Fall Testimony

Finally, the court examined the judge’s ruling on expert testimony about short falls. The judge concluded that there is no general acceptance of the theory that accidental short falls can never cause symptoms consistent with abusive head trauma. His order prohibited testimony “that short falls cannot cause [abusive head trauma].”

Hsu argued that the order was violated when Shah and Schwartz testified that Chloe’s injuries were not consistent with a short fall. The court disagreed. When asked whether Chloe’s injuries would likely be caused by a short fall less than three feet, Shah responded that they “would not be consistent with that short [of a] fall.” Schwartz opined that Chloe’s subdural injury could not be caused by a fall from a sitting position. In both instances, this testimony reflected specific analysis of Chloe’s injuries, whereas the order prohibited testimony that short falls in general cannot cause abusive head trauma. The experts’ testimony thus did not violate the order.

Why It Matters

This decision provides important guidance on the admissibility of abusive head trauma testimony in Massachusetts criminal cases. The court reaffirmed that abusive head trauma is a generally accepted medical diagnosis when properly understood as a diagnosis reached through differential diagnosis, not a diagnosis requiring a specific triad of symptoms. The decision allows prosecutors to present such testimony while requiring trial judges to impose limits that prevent experts from usurping the jury’s role.

The decision also clarifies that experts may testify that a child’s specific injuries are not consistent with a short fall, even though they may not testify that short falls can never cause abusive head trauma in general. This distinction allows fact-specific expert analysis while acknowledging ongoing scientific debate about rare cases.

The case joins recent Supreme Judicial Court decisions recognizing the need for defense access to expert testimony challenging abusive head trauma diagnoses. Here, unlike in prior cases where convictions were vacated because defendants lacked access to defense experts, Hsu presented two experts who offered alternative theories. The jury heard both sides of the scientific debate and made its own credibility determinations. The decision thus illustrates the kind of adversarial testing of expert testimony that the court has emphasized is necessary in these cases.

Read the full opinion: Commonwealth v. Shu Feng Hsu (SJC-13855, September 9, 2026).