A study published in APL Bioengineering has introduced a refined method to differentiate scalp and brain blood flow signals using speckle contrast optical spectroscopy (SCOS). The work — led by researchers based at the University of Toronto, Canada — tackles one of neuroimaging’s long-standing challenges: how to reliably isolate cerebral hemodynamics from superficial artifacts.
As functional near-infrared spectroscopy (fNIRS) and laser speckle methods increasingly appear in ophthalmic and neuro-ophthalmic research, understanding their depth sensitivity and physiological confounders is becoming ever more relevant. Optical imaging methods such as fNIRS and SCOS detect dynamic speckle patterns generated by coherent light scattering off moving red blood cells. The challenge is that light traverses both scalp and skull before reaching the brain — and the returning photons carry a mixed signature of both layers’ blood flow.
This superficial contamination limits clinical translation: while brain-level perfusion or oxygenation changes are of prime interest, scalp vascular responses can obscure the neural signal.
The researchers applied temporary occlusion of the superficial temporal artery (STA) in healthy volunteers while monitoring speckle contrast signals at multiple source–detector separations. By gently compressing the STA — a major supplier of scalp perfusion, but which doesn't supply blood to the brain — they selectively reduced superficial blood flow without altering cerebral circulation. This biological phantom approach allowed the team to directly measure how much of the optical signal at each channel originated from the scalp versus the brain.
During STA occlusion, the short-separation detectors showed large reductions in blood-flow-related contrast, confirming high scalp sensitivity. In contrast, long-separation channels exhibited much smaller changes, consistent with greater cerebral contribution. By integrating these experimental data into photon-transport and sensitivity models, the study authors could estimate the relative weighting of scalp and brain signals across optical paths. The model predicted that roughly 60–70 % of signal variance at common fNIRS separations still arises from superficial layers — quantifying a problem long suspected but rarely measured in vivo.
The study provides a benchmark for interpreting optical hemodynamic data in both research and clinical contexts. For neuro-ophthalmic investigators using optical technologies to study cortical visual responses, autonomic reflexes, or optic nerve head circulation, these results highlight the need to account for — and ideally correct — superficial confounds.
Beyond neuroscience, the same principles apply to ocular imaging. The eye shares similar multilayered scattering characteristics, and improved separation of superficial (e.g., conjunctival or episcleral) and deep (retinal or choroidal) flow components could refine perfusion metrics obtained by optical coherence or speckle-based angiography.
As functional near-infrared spectroscopy (fNIRS) and laser speckle methods increasingly appear in ophthalmic and neuro-ophthalmic research, understanding their depth sensitivity and physiological confounders is becoming ever more relevant. Optical imaging methods such as fNIRS and SCOS detect dynamic speckle patterns generated by coherent light scattering off moving red blood cells. The challenge is that light traverses both scalp and skull before reaching the brain — and the returning photons carry a mixed signature of both layers’ blood flow.
This superficial contamination limits clinical translation: while brain-level perfusion or oxygenation changes are of prime interest, scalp vascular responses can obscure the neural signal.
The researchers applied temporary occlusion of the superficial temporal artery (STA) in healthy volunteers while monitoring speckle contrast signals at multiple source–detector separations. By gently compressing the STA — a major supplier of scalp perfusion, but which doesn't supply blood to the brain — they selectively reduced superficial blood flow without altering cerebral circulation. This biological phantom approach allowed the team to directly measure how much of the optical signal at each channel originated from the scalp versus the brain.
During STA occlusion, the short-separation detectors showed large reductions in blood-flow-related contrast, confirming high scalp sensitivity. In contrast, long-separation channels exhibited much smaller changes, consistent with greater cerebral contribution. By integrating these experimental data into photon-transport and sensitivity models, the study authors could estimate the relative weighting of scalp and brain signals across optical paths. The model predicted that roughly 60–70 % of signal variance at common fNIRS separations still arises from superficial layers — quantifying a problem long suspected but rarely measured in vivo.
The study provides a benchmark for interpreting optical hemodynamic data in both research and clinical contexts. For neuro-ophthalmic investigators using optical technologies to study cortical visual responses, autonomic reflexes, or optic nerve head circulation, these results highlight the need to account for — and ideally correct — superficial confounds.
Beyond neuroscience, the same principles apply to ocular imaging. The eye shares similar multilayered scattering characteristics, and improved separation of superficial (e.g., conjunctival or episcleral) and deep (retinal or choroidal) flow components could refine perfusion metrics obtained by optical coherence or speckle-based angiography.
