Hidden Veins in the Skull Could Transform How Surgeons Treat Hydrocephalus

Deep inside every human skull, sandwiched between the dense plates of bone that protect the brain, lies a network of veins that scientists have largely ignored for centuries. These diploic veins run through the spongy middle layer of the calvaria, the diploë, forming valveless channels that connect the coverings of the brain with the venous sinuses that drain blood from the head. In some people, these channels balloon into strikingly enlarged structures known as diploic venous lakes. A new study from Stanford University now maps these lakes in unprecedented detail, and the researchers believe the findings could point toward a radical redesign of the devices used to drain excess cerebrospinal fluid from the brains of patients with hydrocephalus.

The research, published in the open-access journal Heliyon, was led by Christopher J. Shin and Tarik F. Massoud of Stanford University Medical Center. The team set out to answer a deceptively simple question: if engineers ever build a shunt that diverts cerebrospinal fluid directly into the skull bone instead of threading a tube all the way to the abdomen, where exactly inside the skull should such a device be placed? Their answer, drawn from high-resolution magnetic resonance imaging of forty adults, is that the largest diploic venous lakes cluster predictably near well-known surface landmarks of the skull, most notably the pterion and the asterion, and could therefore serve as anatomically reliable targets for a new generation of intraosseous drainage devices.

To understand why this matters, it helps to consider the shortcomings of the current gold standard. Long-term ventriculoperitoneal shunting, the standard treatment for communicating hydrocephalus, including idiopathic normal pressure hydrocephalus, carries a well-documented burden of complications. Shunts can become infected or obstructed, abdominal pseudocysts can form around the distal catheter, bowel perforation can occur, and overdrainage can precipitate subdural hemorrhages. For decades, neurosurgeons have sought alternatives that would keep cerebrospinal fluid within the body while avoiding the fragile plumbing of a peritoneal catheter. One idea, explored by a small group of researchers in Alberta over the past two decades, is to use the skull itself as a receptacle, allowing diverted fluid to percolate through the cancellous bone of the diploë and into the diploic veins, which ultimately deliver it to the dural venous sinuses.

That concept, however, has run into a fundamental problem. Computational fluid dynamics modeling of skull bone specimens has shown that the porosity and permeability of the diploë vary enormously from one region to another and from one patient to the next. Fluid pushed into the bone may dissipate at unpredictable rates, making it impossible to guarantee consistent drainage. The Stanford team’s hypothesis was that bypassing the spongy bone altogether and cannulating, or placing a catheter tip directly into or beside, a markedly enlarged diploic venous lake might sidestep this variability. Rather than relying on the bone as a passive reservoir, the device would tap directly into the venous highway that already connects the diploic space to the dural sinuses.

Testing that idea required a detailed anatomical map, and diploic veins are notoriously difficult to study. They cannot be dissected cleanly, they are invisible on conventional angiography, and they have been called the least studied of all the vascular networks of the head and neck. The Stanford investigators turned to a volumetric, fat-suppressed, post-contrast spoiled gradient recalled echo sequence acquired on 3 Tesla MRI scanners, with isotropic slices just 0.9 millimeters thick. Gadolinium contrast enhancement made the venous channels within the bone stand out vividly, allowing the researchers to trace their branching patterns in three orthogonal planes. They retrospectively sampled brain MRI examinations performed between May 2014 and September 2020, randomly selecting studies reported as normal or near-normal and excluding any patient whose intracranial or calvarial pathology might have altered venous hemodynamics.

The final cohort comprised forty adults, twenty-seven women and thirteen men, with a mean age of 48.4 years, drawn from a broad range of clinical indications including headache, multiple sclerosis, seizures, and cognitive symptoms. A neuroradiologist and a senior trainee reviewed every scan together on a picture archiving system, reaching consensus on each measurement. The team catalogued eighty bilateral diploic venous channels and classified their configurations into six patterns previously described by the anthropologist Israel Hershkovitz and colleagues: coronal, hybrid, bonsai, serpentine, spider, and thousand lakes. The coronal pattern was by far the most common, appearing in thirty-nine of the eighty channels, followed by hybrid and bonsai configurations. The parietal bones hosted the majority of these structures, and twenty-four subjects showed bilaterally symmetrical arrangements.

For morphometric analysis, the researchers focused on the dominant channel in each subject, defining diploic veins as channels under 3 millimeters in maximal diameter and diploic lakes as focally dilated channels measuring 3 millimeters or more. The 3-millimeter cutoff was chosen partly because it exceeds the 2.1-millimeter outer diameter of the largest 14-gauge intraosseous needle currently available clinically, meaning any lake above this threshold could in principle accommodate a cannula. Of the forty dominant channels, seventeen, or 42.5 percent, qualified as diploic lakes. Their dimensions varied dramatically: mean maximal diameter was 8.3 millimeters, but the median was only 4.2 millimeters and the largest lake reached an extraordinary 46.8 millimeters, with cross-sectional areas ranging from 4.5 to 183.8 square millimeters. This right-skewed distribution underscores how much individual anatomy can differ from person to person.

The most striking result, however, concerned location. More than eighty percent of the large-caliber channels lay within 20 millimeters of the nearest established external skull landmark, and 84.6 percent lay within 30 millimeters. The pterion, the junction where the frontal, parietal, temporal, and sphenoid bones meet, was the most frequent site of proximity, with twenty-six channels found at a mean distance of just 10.9 millimeters. The asterion, the junction of the parietal, temporal, and occipital bones near the transverse sinus, hosted only four channels, but those were almost contiguous with the landmark, at a mean distance of 0.5 millimeters. Crucially, every channel associated with the pterion or asterion lay within 20 millimeters of its landmark. By contrast, channels near the lambda, at the back of the skull, sat farthest away, averaging 64.3 millimeters. This clustering aligns with known drainage pathways: the anterior temporal diploic veins converge toward the sphenoparietal sinus near the pterion, while the posterior temporal veins connect to the transverse and sigmoid sinuses near the asterion.

Equally important for any future device design was what the study did not find. There were no statistically significant differences in vessel size between men and women, and correlations between age and vessel dimensions were negligible. This demographic stability, consistent with earlier osteological evidence, suggests that surface-based targeting strategies developed from these measurements could generalize across patients without needing to adjust for sex or age. The findings also reinforce a growing appreciation of the diploic system’s physiological significance. Beyond serving as a potential thermoregulatory radiator for the brain, the diploic veins appear to participate in cerebrospinal fluid drainage, with arachnoid granulations and lateral lacunae providing pathways by which fluid filters into the venous channels. Some anthropologists have even proposed that the evolution of the diploic system played a role in the expansion of the human brain, with similarities noted between modern humans and Neanderthals.

The authors are careful to frame their work as a foundational anatomical study rather than a clinical validation. The analysis did not assess flow dynamics, pressure gradients, or cannulation feasibility, and vessel size alone does not establish that a diploic lake can sustain cerebrospinal fluid drainage. Measurements were obtained by consensus review rather than by independent blinded readers, and the retrospective clinical cohort may not perfectly represent a healthy population. Earlier experimental work offers encouraging context, however: cadaveric infusion studies have shown that fluid injected into the diploic space rapidly reaches the superior sagittal sinus and lateral lacunae, confirming direct venous continuity, and animal models have demonstrated similarly low-resistance connections. Whether a catheter tip parked within or beside a diploic lake would outperform the surrounding diploë as a drainage receptacle remains to be proven, likely requiring bioengineering studies in large animal models, and possibly even surgically constructed venous lakes, since the human diploic system may have no perfect animal counterpart. Still, by converting a century of qualitative observation into quantitative morphometrics anchored to palpable skull landmarks, the Stanford team has given device designers, for the first time, a reproducible map of where the skull’s hidden veins are waiting.

Subject of Research: MRI-based morphometry of calvarial diploic venous lakes for designing intraosseous cerebrospinal fluid diversion devices

Article Title: MRI-derived spatial morphometrics of calvarial diploic venous lakes for designing next-generation cerebrospinal fluid diversionary devices

Article References: Shin, C. J., Bates, N. S., Decker, J. H., & Massoud, T. F. (2026). MRI-derived spatial morphometrics of calvarial diploic venous lakes for designing next-generation cerebrospinal fluid diversionary devices. Heliyon, 12(15), Article e45529. https://doi.org/10.1016/j.heliyon.2026.e45529

Image Credits: AI Generated

DOI: 10.1016/j.heliyon.2026.e45529

Keywords: diploic veins, diploic venous lakes, hydrocephalus, cerebrospinal fluid shunt, intraosseous infusion, skull anatomy, MRI, neurosurgery, pterion, asterion, ventriculoperitoneal shunt, dural venous sinuses