USC and Caltech move speckle-based cerebral blood flow test closer to clinics

22 Oct 2025

New study tackles long-standing issue of scalp's interference with measurements.

Speckle contrast is a promising approach to assessing blood flow, developed to capture images of laser light scattered from blood vessels with affordable optical sensors and components.

A project at Caltech and USC has now tackled one long-standing question regarding the practical use of the technique, and published the results in APL Bioengineering.

The study builds on the previous development of a speckle contrast optical spectroscopy (SCOS) platform by the same institutions, a headset unit designed to assist the assessment of stroke risk in patients.

The clinical promise of that device is clear, but the SCOS technique is potentially affected by the need to measure through the scalp of the subject, with blood flow in the scalp complicating the measurement of brain blood flow at greater depths.

"This question has long plagued researchers who use light-based technology to visualize the brain," commented USC. "The team sought to confirm that SCOS is truly measuring blood flow in the brain rather than in the scalp, which also contains many blood vessels."

Previous approaches to the scalp issue have investigated whether scalp compression, time-domain spectroscopy or statistical simulations might be the answer, but these can be uncomfortable for the patient or limit the spectroscopic data being gathered. A more repeatable and safer technique would be preferable.

The team's solution was to temporarily occlude a patient's temporal artery by gentle application of pressure near the ear bone, isolating scalp blood flow while preserving cerebral circulation and causing minimal discomfort. A seven-channel SCOS system positioned over the patient's temple region then collects speckle contrast data from the resulting blood flow, with different detecting channels for sensing the scalp, the skull and layers of the brain.

Measuring exactly what's intended

In 20 participants, the researchers temporarily stopped blood flow to the scalp, then collected a series of SCOS readings. By gradually moving the detector further from the head, they captured signals reaching progressively deeper towards the brain and found that positioning the detector 2.3 centimeters from the head measured brain blood flow while minimizing interference from the scalp.

This experimental data quantified the sensitivity of scalp and brain blood flows in SCOS as a function of source–detector distance, alongside a direct comparison of blood flow and blood volume sensitivities. This data should start to establish an experimental framework for estimating scalp and brain sensitivities in optical systems, a pivotal contribution according to the project.

With the technique already being used by some of the team's clinical collaborators to help diagnose stroke and traumatic brain injury, the next steps will involve refining the technology and software to improve the resolution of images and the quality of data extracted from readings.

But the study already helps confirm the clinical potential of SCOS for detecting and responding to stroke, brain injury and dementia, commented USC's Charles Liu, and because all of the team's research has been done with humans, the tool is poised for rapid translation from the lab to the clinic.

"We look directly at humans in essentially the same way the tool will be applied, so there’s nothing lost in translation," Liu said. "We are never more than one step away from the problem we’re trying to solve. With the knowledge that we’re now measuring exactly what we intend to measure, we’re also going to expand our testing of this technique with patients in clinical settings."