Blood’s Uncertain Arc

Richard Jonathan O. Taduran, Ph.D.  |  18 June 2026

Popular culture—especially the TV series Dexter—has taught audiences to see bloodstain pattern analysis as almost infallible: one glance at a spray of blood, and the analyst can supposedly reconstruct the weapon, movement, and sequence of violence. Reality is far less certain. Blood obeys physics, but interpreting the marks it leaves behind still depends heavily on human judgment. In one large study by Hicklin and colleagues, analysts examining patterns whose causes were already known reached erroneous conclusions 11.2 percent of the time.

The consequences can be devastating. In the controversial Julie Rea case, prosecutors argued that blood on her nightshirt was cast-off from the knife used to kill her son. Defense experts saw transfer instead. Rea was convicted, later acquitted, and ultimately exonerated.

A new study by Savita Brickman-Maxwell and Eugene Liscio tests whether HemoVision can make cast-off analysis more measurable. The system reconstructs a blood-bearing object’s path through three-dimensional space.

Following the Swing

A cast-off pattern forms when blood is released from a moving object, such as a weapon being swung. Rather than treating the event as a single point of origin, the researchers asked HemoVision to estimate the curved path through which the object travelled.

They produced 36 controlled patterns using sheep blood and a mechanical rig with a fixed 75-centimetre circular arm. Ten patterns came from downward vertical swings, ten from upward swings, ten from downward diagonal swings, and six from single-blind trials. The blood landed on smooth white paper attached to a vertical wall.

Investigators placed reference markers around each pattern, photographed it from several angles, and uploaded the images into HemoVision. The analyst selected suitable stains, fitted ellipses to their shapes to determine each droplet’s impact angle, and allowed the software to calculate trajectories. From these trajectories, HemoVision generated a tubular swing path envelope, or TSPE: a three-dimensional corridor representing where the bloody object probably moved.

That is the study’s most useful innovation. Cast-off is movement, not a stationary event, and the TSPE visualizes that movement while displaying uncertainty. The horizontal position was estimated most accurately, while height was least accurate. Average three-dimensional errors ranged from approximately 12 to 27 centimetres among trial groups. The known swing path was fully contained within the TSPE in 18 trials and partially overlapped it in the other 18. Most envelopes were about 10 centimetres wide, although diagonal trials showed greater variation.

These results do not provide an exact replay of a strike. They offer something more scientifically honest: a measurable region of probability that investigators can compare with bodies, objects, statements, and the surrounding scene. The workflow also preserves photographs and measurements, allowing later review after scene release and reducing reliance on physical stringing at crime scenes by investigators and defense experts.

Possibility, Not Proof

But promising is not the same as casework-ready. This was a preliminary experiment under highly controlled conditions. A wooden arm followed a fixed circular path and deposited blood on a smooth, white, vertical surface. Human violence is rarely so orderly. Real strikes involve the hips, shoulders, elbows, and wrists, while the victim, assailant, and weapon may all change position.

Subjectivity also remains inside the automated workflow. The analyst decides which stains to select, how to fit ellipses around them, and which unusual trajectories to remove. Circular, distorted, incomplete, or overlapping stains can make those judgments difficult. The blind trials were limited: although absent during pattern creation, the analyst knew the rig’s fixed radius—a luxury unavailable at a real crime scene—and had already examined the other 30 patterns.

There are physical limits. HemoVision projects straight trajectories, although gravity bends droplets in flight. That likely contributed to larger errors in estimated height. More importantly, the study examined only the arc of controlled cast-off. It did not test overlapping blows, textured surfaces, corners, clothing, moving participants, or mixed scenes.

Could it work in the Philippines? Potentially. A photographic system may be more accessible than costly laser scanners. But imported software is not automatically validated evidence. Philippine laboratories would need trained analysts, calibrated photography, quality assurance, proficiency testing, and local studies using tile, concrete, wood, fabric, poor lighting, humidity, and cramped rooms.

These findings also cannot be extended to every blood pattern. Drips, transfers, wipes, swipes, arterial stains, expirated blood, impact spatter, and mixed patterns arise through different mechanisms. HemoVision’s cast-off feature should be understood exactly as tested: a promising tool for one pattern type, under narrow conditions, requiring independent validation.

Independent teams must determine whether different analysts examining the same stains produce comparable envelopes, measurements, exclusions, and conclusions before the method enters casework.

Standards Over Spectacle

Forensic science earns authority not from impressive technology, but from standards. A three-dimensional reconstruction may look persuasive in court, yet visual polish cannot substitute for known error rates, repeatable procedures, independent validation, proficiency testing, and transparent documentation of every analyst decision.

HemoVision’s tubular swing path envelope is valuable because it does not pretend to identify one perfect trajectory. It presents a region of uncertainty. That restraint should guide how investigators report the result and how lawyers, judges, and juries understand it. The software can support reconstruction, test competing accounts, and direct attention within a scene. It cannot determine guilt, identify the person holding the object, or explain the entire violent event by itself.

Every forensic method must be tested under the conditions in which it will actually be used. A computer may draw the arc of a possible swing, but scientific standards determine whether that arc deserves to become evidence.