Welcome
The Laboratory for Cognition and Neural Stimulation (LCNS) is a team of scientists and clinicians within the Neurology Department at the University of Pennsylvania committed to advancing groundbreaking basic, translational, and clinical research involving neuromodulation. The central objectives of the LCNS are to employ noninvasive neuromodulation techniques in order to
- further elucidate the neural basis of cognition,
- develop and implement novel stimulation-based therapies for neurologic disorders, particularly those that affect cognition and sensorimotor function throughout the lifespan,
- reveal critical and modifiable behaviorally-relevant circuit and network properties of the human brain.
In order to accomplish these objectives, LCNS investigators combine noninvasive brain stimulation with complimentary approaches, including but not limited to behavioral measures, patient lesion studies, structural and functional neuroimaging, and measures of cortical physiology such as electroencephalography (EEG) and magnetoencephalography (MEG). The LCNS actively aims to cultivate partnerships both in the University of Pennsylvania community and beyond, and seeks to share its intellectual expertise, research infrastructure, and brain stimulation resources in order to promote far-reaching multidisciplinary research. In pursuing our objectives, the LCNS aspires to make high-impact discoveries that advance the well-informed use of noninvasive neuromodulation in the fields of cognitive neuroscience, neurology, neurorehabilitation, and beyond.
News
-
Leveraging AI to help stroke survivors recover speech abilities
Friday, February 13, 2026
Stroke survivors with aphasia often face long, highly variable paths to regaining speech and language abilities. In a recent study, Shreya Parchure, a researcher in the LCNS Lab (Laboratory for Cognition and Neural Stimulation), explored how explainable artificial intelligence (AI) can help personalize speech therapy for individuals recovering from stroke.
Rather than relying on standardized treatment approaches, Parchure and her team developed AI models that integrate clinical characteristics—such as stroke severity and patient background—with linguistic features like word difficulty. The system can predict how a patient may respond to specific therapy targets and suggest more individualized treatment strategies. Importantly, the AI is designed to be transparent, allowing clinicians to understand how recommendations are generated and increasing trust in its use in clinical settings.
The research also includes the development of an AI-powered tool intended to support therapists in designing more precise, data-informed rehabilitation plans.