For most of a century the motor cortex came with a picture attached: the homunculus, a little man draped head to toe down the precentral gyrus, lips here, hand there, foot at the top. Then in April 2023 a team led by Evan Gordon and Nico Dosenbach reported in Nature that the little man is interrupted. Using precision functional MRI, they found three regions interleaved between the foot, hand and mouth areas that did not move anything in particular. They connected to each other, lit up during action planning and whole-body movement, and tied into arousal and internal physiology. They called it the somato-cognitive action network, SCAN for short: less a keyboard key than a conductor.
Parkinson's disease has always fitted awkwardly inside the keyboard model. It starts, often, with things that belong to no single limb: disturbed sleep, constipation, a lost sense of smell, a flattened voice, a feeling that the body has become heavy. On 4 February 2026 a large international group led by Jianxun Ren and Hesheng Liu, with Dosenbach among the authors, published in Nature a study arguing that the mismatch is the clue. They assembled 11 independent datasets into a multimodal, multi-intervention imaging collection of 863 participants, including a precision imaging cohort of 166 people with Parkinson's and 60 healthy controls, a deep brain stimulation sweet-spot dataset of 342 patients, and comparison cohorts with essential tremor (45) and dystonia (42). Resting-state functional connectivity showed that the substantia nigra, the dopamine-producing structure that degenerates in Parkinson's, and all three established deep brain stimulation targets (subthalamic nucleus, globus pallidus, ventral intermediate thalamus) were selectively connected to the SCAN rather than to the effector-specific motor regions beside it. Surgeons have been aiming, without knowing it, at one network.
The second finding is what gives the first its weight. In Parkinson's, the connection between the SCAN and those deep structures was not merely present, it was too strong. The paper reports this hyperconnectivity as specific: it showed up in Parkinson's and not in the comparison movement-disorder cohorts, and it was absent for the classic effector regions. The authors also report that stronger connectivity went with more severe symptoms. That is an association measured at one moment in time across groups of people, not a demonstration that the wiring causes the illness, and the paper's own design cannot settle which came first. "Our work shows that the disease is rooted in a much broader network dysfunction," Liu said in a Washington University release accompanying the paper.
Then the team went looking for the signature in things that already work. Across six cohorts undergoing deep brain stimulation, transcranial magnetic stimulation, MRI-guided focused ultrasound and levodopa, treatments that improved symptoms also pushed SCAN-to-subcortex connectivity back toward control levels. Levodopa significantly reduced it (one-tailed paired t = 3.58, P = 0.001). In 14 patients scanned before and after subthalamic stimulation surgery, motor scores improved at 1, 3, 6 and 12 months and connectivity drifted back toward the healthy pattern. In 17 patients with electrodes laid on the cortex during surgery, stimulating the motor part of the subthalamic nucleus evoked its strongest responses in SCAN territory rather than effector territory. In the focused ultrasound cohort, the closer the lesion sat to the thalamic SCAN target, the greater the motor improvement (Spearman rho = -0.68, P = 0.031). Four very different interventions, one common denominator.
The part that will travel fastest is the smallest. The team randomised 36 people with Parkinson's, mean age 65, one to one, to receive transcranial magnetic stimulation aimed either at their individually mapped SCAN regions or at standard effector-specific motor cortex. Targeting the SCAN, the abstract states, doubled the efficacy of the treatment. Thirty-six people is a pilot, 18 per arm, a comparison of two targeting strategies rather than a definitive efficacy trial, and it is reported here as preliminary because that is what it is. Nothing in this paper changes what any clinic offers this week, and TMS for Parkinson's remains a research and adjunct setting, not a replacement for the drugs and devices people are already using.
What it may change is the explanation a family gets. If the network that integrates posture, arousal, gut and gaze with the intention to act is the one running hot, then the breathiness, the sleep that breaks apart, the freezing in a doorway and the slow arithmetic of a conversation stop being a ragged list of extras attached to a movement disorder. They become symptoms of the same thing. And if the substantia nigra, the subthalamic nucleus, the pallidum and the thalamic target all report to that same conductor, the long-standing puzzle of why such different procedures produce such similar relief has a candidate answer. The next question, which this paper does not answer, is whether aiming at the conductor on purpose, in hundreds of people rather than dozens, holds up.