🚨NEW PREPRINT🚨 What if we could… 🏠 collect quality neural-behavioral data at home 🧠 identify naturalistic + personalized brain signatures of movement state 🚶🏽‍♂️use these biomarkers to drive real-time adaptive DBS for gait disorders Our new work @DorisWangLab shows how! (1/8)

Truly honored to have been part of this exciting work alongside an amazing team led by Dr. Alex Haddad, @ManishKAghi, and @LabRaleigh. Read the full paper here! (4/4) @UCSF @UCSFCancer #neurosurgery #neurooncology #meningioma https://t.co/z2GEHwqiXv

Embolized meningiomas showed⬆️pathways linked to cellular differentiation/hypoxia, and ⬇️pathways involved with mitosis/cell cycle progression, suggesting that improved long-term outcomes may occur through inhibition of the cell cycle in these tumors. (3/4)

After propensity matching, preop embo was independently associated with longer recurrence free survival (HR 0.55), particularly in those with subtotal resection (HR 0.32). (2/4)

🔎🧠Does preop embolization improve oncological outcomes in meningiomas? Our team @UCSFNeurosurg analyzed 319 patients, finding that embo improves local control and modulates gene expression in atypical WHO grade 2 meningiomas! Read it now @NeuroOnc - link in thread (1/4)

Congratulations to the 6 finalists of the DBS Think Tank Rising Stars Challenge! Tune in Aug 29 at 1pm EST to listen to each finalist present their cutting edge research! https://t.co/emfIbDc4PG

AI can help tailor deep brain stimulation to improve walking performance in people w/ Parkinson’s disease, a study finds. It’s “about finding the right settings designed to everyone’s brain activity & walking patterns,” @NeuroSurgUCSF’s @DorisWangLab says.

In new work led by @DorisWangLab, researchers developed a systematic way to quantify key aspects of gait in patients with #Parkinsons & identified personalized deep brain stimulation (DBS) settings to improve walking using machine learning: https://t.co/SRSz9BvSeW @UCSFHospitals

Can we reboot walking after Parkinson's disease DBS? Can AI teach deep brain stimulation devices to help people 'walk the walk?' A new study by Wang and colleagues in Nature Parkinson’s Disease says MAYBE. Using personalized data-driven models, researchers at UCSF have optimized

New in @Nature_NPJ, @DorisWangLab presents an AI approach for identifying personalized deep brain stimulation (DBS) settings to improve walking in people with #Parkinsons. Their findings lay the groundwork for adaptive DBS systems that adjust in real-time: https://t.co/azrsASmKGS

This project was made possible by amazing postdocs Hamid and Ken, star coords Jannine and Jacob, and of course our patients! Check out our work! (8/8) #Parkinsons #neurophysiology #neuroscience #dbs https://t.co/k9ajvBJYji

Our results show some of the first in-human classification of a specific movement type using chronic at-home data, piloting a novel pipeline for circuit discovery in the real world, which may be applicable to a wide variety of disorders treatable with neuromodulation. (7/8)

Excitingly, in silico simulation of real-time decoding on-board the DBS device showed above-chance classification across all subjects, supporting the use of the identified neural signatures to control aDBS. (6/8)

We found highly sensitive and specific biomarkers enabling above-chance decoding performance in all hemispheres. The most important features varied by subject, though pallidal features (delta, theta, beta) were generally the most important. (5/8)

Comparing neural activity between gait states, we found ⬇️ M1 alpha and beta power and ⬆️ GP delta and theta activity during periods of walking compared to rest, which may indicate antikinetic and prokinetic roles respectively for these frequency ranges. (4/8)

Here, we: 1) recorded 80+ hours of at-home neural + kinematic data from 4 people with PD using a bidirectional neurostimulator and wearable sensors 2) identified cortical-pallidal signatures of gait state 3) tested the ability of these biomarkers to drive closed-loop aDBS (3/8)

Human movement decoding from neural activity has mostly relied on external computers trained with recordings during short artificial in-lab tasks, limiting our understanding of real-world behavior and preventing the study of complex motor actions such as gait. (2/8)

Read our work on this important and underexplored intersection:

Thrilled to share our paper in @NeurosurgeryCNS characterizing 146 intracranial epidermoid cysts! We discuss high recurrence esp after STR, common postop complications, frequent long-term symptom resolution, and a unique case of malignant transformation: https://t.co/Vgtn4jZpro

#ONSNew Our work highlights the use of ICG for preservation of the SHA and adequate optic nerve and stalk perfusion during resection of a large tuberculum sellae meningioma. @NeurosurgeryCNS @NeurosurgUCSF @RithvikRamesh6