Note: The "control loop" is a fundamental concept in "control systems engineering". It includes a controller which controls a process of a system and a sensor which senses the system output. When we provide the output back to system (feed back) to inform its function, i.e., by defining the controller's action, we have a closed-loop, where the system output is used to control the system.
"These algorithms spot mood changes before you do, and could someday tell a stimulator to zap your brain to treat disorders"
https://spectrum.ieee.org/algorithms-play-doctor-in-brain-stimulation
IEEE Spectrum, 2019-05-17
Featured in DARPA tweet (SUBNETS program): https://twitter.com/DARPA/status/1130466005036281858
"A closed loop system... employs algorithms to constantly monitor the patient’s brain activity, and order stimulation the moment the patient needs it, with parameters specific to the patient’s neural activity."
https://link.springer.com/article/10.1134/S0022093023050113
References
Bhattacharya A, Mrudula K, Sreepada SS, et al. An Overview of Noninvasive Brain Stimulation: Basic Principles and Clinical Applications. Canadian Journal of Neurological Sciences / Journal Canadien des Sciences Neurologiques. 2022;49(4):479-492. doi:10.1017/cjn.2021.158
https://pmc.ncbi.nlm.nih.gov/articles/PMC5717031/
It is noted that is cites "common neurofeedback paradigms", such as:
🔹 Neurofeedback on "slow cortical potentials" (SCP’s), which entrains very slow oscillations in a range of 0.3–1.5 Hz,
🔹 Infra-low neurofeedback (ILN), which entrains even slower brain oscillations, ranging from 0.001 to 1.5 Hz.
Brief comment (AI-assisted) on the announced Parkinson's TMS therapy on the SCAN network — Reference: https://medicine.washu.edu/news/brain-network-responsible-for-parkinsons-disease-identified/
In Parkinson’s, the somato‑cognitive action network — the SCAN, a circuit linking movement, body awareness, and the cognitive processes that prepare and guide actions — becomes abnormally synchronized, like a city where all the traffic lights switch in unison and create gridlock.🚦🕸️
Scientists say that these areas become functionally hyperconnected — their activity increases and decreases in lockstep, rising and settling at the same time instead of adjusting independently. 🔗
TMS sends brief magnetic pulses that induce tiny electrical currents in the cortex — an electromagnetic nudge that alters the timing of brain activity. 🧲⚡
By disrupting this excessive synchrony, it breaks the rigid lockstep created by hyperconnectivity and restores the staggered coordination that allows the traffic system — and movement — to flow again. ➡️
"Scientists discover brain network that may cause Parkinson’s disease" — Featured by Scientific American (general audience source): "Scientists have found a key brain network that’s disrupted by Parkinson’s disease, according to a study published today in Nature. The results change doctors’ understanding of what causes Parkinson’s symptoms and may unlock more effective and precise treatments". https://www.facebook.com/ScientificAmerican/posts/pfbid02ubBhuGh13qoAmwv5naocV7ageB3r2VkLrvHVwCJodnxNdTomecYQDBoBQcHzTTbcl
Nature:
In February 2025, the FDA granted approval for the first adaptive (closed-loop) deep-brain stimulation approach in Parkinson’s disease, based on evidence including the ADAPT-PD trial. ADAPT-PD stands for "Adaptive DBS Algorithm for Personalized Therapy in Parkinson’s Disease"
Refrence: https://www.apdaparkinson.org/article/adaptive-deep-brain-stimulation-dbs/
The approved system, BrainSense™ aDBS, adjusts stimulation automatically, in real-time, based on the patient's neural activity detected directly from the deep-brain structure Globus Pallidus internus (GPi). It reportedly addresses excessive synchrony in the alpha–beta band — a rhythm closely linked to motor symptoms.
Video: https://www.medtronic.com/content/dam/medtronic-wide/public/united-states/products/neurological/brainsense-intro-video.mp4 Reference: https://www.medtronic.com/en-us/healthcare-professionals/products/neurological/deep-brain-stimulation/dbs-technologies/brainsense-technology.html
Another DBS system, though not adaptive (cf. open-loop), was approved by the FDA for Parkinson’s disease (PD) in 2002.
Closed-loop neurostimulation itself, had been granted approval earlier by the FDA in 2013 for epilepsy. The approved system, the Responsive Neurostimulator — which also senses brain activity and delivers stimulation automatically — targets cortical or hippocampal seizure-onset areas (rather than deep brain nuclei).
In Parkinson's disease, movement difficulties in particular, are associated with an abnormally strong slow beta pattern (13-20 Hz).
Deep brain stimulation (electrical stimulation) which suppresses or shifts this pathological frequency can alleviate these difficulties, when targeting deep brain areas such as the subthalamic nucleus (SNT) and the GP (globus pallidus).
The study showed that when SNT and frontal regions share the same timing pattern in the fast beta range (20-35 Hz) — that is the same frequency and phase (with synchronized crests and troughs)— movement difficulties diminish or disappear.
This is indicative of effective communication of the two areas, i.e., of high functional connectivity, meaning that they are acting as part of the same functional network.
As indicated by an AI assistant: The two areas share a common timing pattern — a shared beat.
“We show that Parkinson’s disease can best be treated if we stimulate a very precisely defined network. This network operates synchronized within a specific frequency band, and offers an explanation for how well patients respond to deep brain stimulation.”