Quentir Medicine Monitor
Evidence-based insights for quantum medicine. Published by Quentir Systems LLC · November 2, 2026.
An eight-year-old boy in Shanghai had a seizure focus in the part of the left frontal lobe that textbooks call Broca's area, the classical seat of spoken language. His team at the Children's Hospital of Fudan University located both the focus and his language function with a 128-channel helmet of optically pumped magnetometers, then destroyed most of the focus with a laser. Six months later his seizures were controlled and his speech was intact.
The case report, by Yuanfeng Zhou, Yanjiong Zhang and colleagues in the hospital's neurology, neurosurgery and radiology departments, with coauthors from Beihang University and Fosun MedTech, appeared in Acta Epileptologica on 3 May 2026. It describes one child with drug-resistant epilepsy and one presurgical workup. Its authors present it as a feasibility demonstration, and they call for clinical trials to test the diagnostic value of the method in children.
The clinical problem is familiar to every epilepsy surgery program. Removing or burning tissue in a language region can cost a child the ability to speak, so surgeons want a map of where language actually sits before they operate. The usual tools for presurgical language mapping are functional MRI and the Wada test, an injection of an anesthetic drug into one carotid artery to silence one hemisphere at a time. This boy could use neither.
What the Fudan team found in the boy's left inferior frontal gyrus with MRI, PET and video EEG
His seizures began at five years and nine months of age, with clapping of both hands, repeated blinking and stiffening of the arms. Five antiseizure drugs, oxcarbazepine, levetiracetam, lacosamide, clonazepam and perampanel, brought four months free of seizures at best. At six years and five months the seizures returned in clusters of one to more than ten a day, and he was diagnosed with drug-resistant epilepsy.
Three tests pointed to the same spot. MRI showed abnormal signal in the opercular part of the left inferior frontal gyrus, PET showed reduced metabolism there, and video EEG recorded focal seizures starting in the left frontal region. Trio whole-exome sequencing of the child and both parents found no explanatory genetic variant, and his full-scale IQ on the Wechsler scale was 119. The hospital's refractory epilepsy team decided the focus should be removed or thermally coagulated. The difficulty was its location: the opercular part of the inferior frontal gyrus corresponds to Brodmann area 44, which with the neighboring triangular part, area 45, makes up the classical Broca's area.
According to the report, both standard tests were refused: the Wada test because it is invasive, and functional MRI because the boy tolerated the scanner's loud noise poorly. That left OPM-MEG as the only source of quantitative data on where his language function lay.
How the 128-channel OPM-MEG helmet recorded resting discharges and a 150-trial verb task
Magnetoencephalography, or MEG, measures the faint magnetic fields produced by electrical activity in the brain. The Fudan system, the Marvel MEG 128 from Beijing X-Magtech Medical Technologies, fixes 128 sensors to a child-size helmet. Each sensor is an optically pumped magnetometer: a laser prepares the quantum spin states of atoms in a small glass vapor cell, and a magnetic field changes how those atoms absorb light. The system sits inside a magnetic shielding enclosure that avoids building a full shielded room, and the authors report sensor noise below 15 femtotesla per square-root hertz between 10 and 100 hertz.
The boy had two sessions. The first was a 60-minute resting recording, eyes closed and lying still, to capture epileptic discharges. Three MEG technicians reviewed the recording and marked the spikes, and 20 fitted dipoles placed their sources mainly in the middle frontal gyrus, the inferior frontal gyrus and the insula. Those locations agreed with the Broca's area outline, the averaged spike maps and the low-metabolism zone on PET.
The second session tested language. In each of 150 trials, a Chinese noun appeared for three seconds after a two-second rest, and the boy silently thought of a matching verb, so that "football" called up "kick." The team measured the drop in beta-band brain rhythm during this task and projected it onto his own MRI with a beamformer, a standard method for locating the sources of MEG signals. The strongest activity sat in the left hemisphere, in the orbitofrontal cortex, the insula and the superior temporal gyrus. The report describes minimal overlap with the epileptic focus, and at a display threshold of 60 percent of the maximum it found no overlap with the classical Broca's area on the brain surface.
Quantum pillar: sensing. Technology readiness: TRL 6 of 9. This is Quentir's assessment for the specific use of pediatric language mapping: a complete wearable scanner has been used in a children's hospital to plan one real operation, but this is a single case without a comparison group, and the diagnostic accuracy still has to be shown in a clinical trial.
Why cryogenic SQUID MEG fits children poorly, and what a 2022 Radiology study in Brussels measured
MEG is not new to epilepsy surgery. Conventional systems use superconducting quantum interference devices, or SQUIDs, cooled with liquid helium inside a rigid helmet sized for adults. A small child's head sits loosely in that helmet, the sensors end up farther from the brain, and the child must hold still. Optically pumped sensors need no cryogenic cooling, so they can sit on the scalp in a helmet that fits.
The clearest comparison so far comes from Brussels. In a prospective study in Radiology published on 3 May 2022, Odile Feys, Xavier De Tiège and colleagues at the Université libre de Bruxelles, with physicists from the University of Nottingham, examined five children aged five to eleven with both cryogenic MEG and 32 on-scalp optically pumped sensors. Epileptic discharges were 2.3 to 4.6 times larger in amplitude on the on-scalp system, and the signal-to-noise ratio was 27 to 60 percent higher in all but one child, whose head movement spoiled the recording. Source locations from the two systems lay about 5 millimeters apart in three of the five children. Quentir followed the British side of this work in August, in a piece on the wearable helmet that brings brain mapping closer to childhood; the Fudan report adds a language map that informed a surgical plan.
The Fudan authors also cite the American Academy of Neurology's 2017 practice guideline on functional MRI before epilepsy surgery, which gives functional MRI in place of the Wada test for lateralizing language a level C recommendation. That guideline concerns which hemisphere carries language; the Fudan case asked the narrower question of where inside the left frontal lobe it sits.
What one Acta Epileptologica case report cannot settle for hospitals considering OPM-MEG
The outcome supports the map. After MRI-guided laser interstitial thermal therapy, follow-up MRI showed that most of the focus had been destroyed, and the boy had no problems understanding or producing language. At six months his seizures were controlled. The authors read this as confirmation that his language lay outside the classical Broca's area.
The limits are just as concrete. This is one patient with no Wada test, no functional MRI and no intracranial stimulation mapping to compare against, and the follow-up runs only six months. A good outcome in one child cannot give a sensitivity or specificity for the language map. The insula appears both among the spike sources and in the language map, and the report does not discuss that shared region. One of the authors, Yi Wang, sits on the journal's editorial board and, according to the report, took no part in its review. The work was funded by the Children's Hospital of Fudan University, the Shanghai Municipal Science and Technology Major Project and the Shanghai Municipal Hospital Development Center, and the ethics committee approval is number 2024-173.
For a hospital buyer, the practical facts are the ones the report gives about the system: 128 channels, a child-size helmet and a shielding enclosure that spares the cost of a shielded room. The report does not state the system's regulatory status, price or throughput.
How Quentir Reads It
The Fudan report moves pediatric OPM-MEG from detecting epileptic spikes to guiding where a surgeon may operate near speech. The next convincing step would be a prospective series in which children who can complete functional MRI or stimulation mapping also undergo OPM-MEG language mapping, so that agreement and error rates can be counted. Until that comparison exists, the evidence reviewed here for this specific purpose, pediatric language mapping next to a seizure focus, comprises one case report supported by a physiological rationale.
Sources
Primary source: the Acta Epileptologica case report of 3 May 2026 by Yuanfeng Zhou, Yanjiong Zhang, Luo Tian, Tianshuang Wang, Min Wang, Renqing Zhu, Hao Li, Zhongwei Qiao, Zhigang Cao, Yuming Peng, Ming Ding and Yi Wang. Also drawn on: the 2022 Radiology study by Feys, De Tiège and colleagues, the 2017 American Academy of Neurology guideline by Szaflarski and colleagues, and earlier Quentir coverage.