About

Kim Butts Pauly is a Professor of Radiology (Radiological Sciences Lab) and, by courtesy, of Electrical Engineering at Stanford University. Her research focuses on investigating, developing, and applying focused ultrasound technology for neuromodulation, blood-brain barrier opening, and tissue ablation for neuro and body applications. Her work includes studying the mechanisms of focused ultrasound neuromodulation of deep brain structures, monitoring and ensuring accurate delivery of ultrasound intensities to the brain, and using microbubbles to open the blood-brain barrier to enhance drug delivery for brain cancers. Additionally, she explores how ultrasound modulates the immune system and how this can be directed toward cancer therapy. Her engineering efforts involve finding in situ pressure and tissue strain from MRI-measured radiation force displacement, as well as using focused ultrasound for thermal ablation in movement disorders and cancers. She has contributed to the development of robust volumetric MR thermometry techniques in the presence of motion and transducers. Dr. Pauly is also a member of several professional organizations, including Bio-X, Stanford Cancer Institute, and Wu Tsai Neurosciences Institute, and has held leadership roles such as Director of the Center for Biomedical Imaging at Stanford and Vice Chair for Research in the Department of Radiology.

Research topics

• Computer Science
• Medicine
• Psychology
• Neuroscience
• Acoustics
• Geology
• Radiology
• Artificial Intelligence
• Physics
• Audiology
• Engineering
• Paleontology
• Aerospace engineering
• Telecommunications
• Speech recognition
• Geodesy

Selected publications

• Parameter optimisation for mitigating somatosensory confounds during transcranial ultrasonic stimulation

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-03-19 · 2 citations

  preprintOpen access

  Transcranial ultrasonic stimulation (TUS) redefines what is possible with non-invasive neuromodulation by oaering unparalleled spatial precision and flexible targeting capabilities. However, peripheral confounds pose a significant challenge to reliably implementing this technology. While auditory confounds during TUS have been studied extensively, the somatosensory confound has been overlooked thus far. It will become increasingly vital to quantify and manage this confound as the field shifts towards higher doses, more compact stimulation devices, and more frequent stimulation through the temple where co-stimulation is more pronounced. Here, we provide a systematic characterisation of somatosensory co-stimulation during TUS. We also identify the conditions under which this confound can be mitigated most eaectively by mapping the confound-parameter space. Specifically, we investigate dose-response eaects, pulse shaping characteristics, and transducer-specific parameters. We demonstrate that somatosensory confounds can be mitigated by avoiding near-field intensity peaks in the scalp, spreading energy across a greater area of the scalp, ramping the pulse envelope, and delivering equivalent doses via longer, lower-intensity pulses rather than shorter, higher-intensity pulses. Additionally, higher pulse repetition frequencies and fundamental frequencies reduce somatosensory eaects. Through our systematic mapping of the parameter space, we also find preliminary evidence that particle displacement (strain) may be a primary biophysical driving force behind peripheral somatosensory co-stimulation. This study provides actionable strategies to minimise somatosensory confounds, which will support the thorough experimental control required to unlock the full potential of TUS for scientific research and clinical interventions.

  PublisherOA PDFDOI

• Parameter optimisation for mitigating somatosensory confounds during transcranial ultrasonic stimulation

  Brain stimulation · 2025-06-11 · 8 citations

  articleOpen access

  BACKGROUND: Transcranial ultrasonic stimulation (TUS) redefines what is possible with non-invasive neuromodulation by offering unparalleled spatial precision and flexible targeting capabilities. However, peripheral confounds pose a significant challenge to reliably implementing this technology. While auditory confounds during TUS have been studied extensively, the somatosensory confound has been overlooked thus far. It will become increasingly vital to quantify and manage this confound as the field shifts towards higher doses, more compact stimulation devices, and more frequent stimulation through the temples where co-stimulation is more pronounced. METHODS: Here, we provide a systematic characterisation of somatosensory co-stimulation during TUS. We also identify the conditions under which this confound can be mitigated most effectively by mapping the confound-parameter space. Specifically, we investigate dose-response effects, pulse shaping characteristics, and transducer-specific parameters. RESULTS: We demonstrate that somatosensory confounds can be mitigated by avoiding near-field intensity peaks in the scalp, spreading energy across a greater area of the scalp, ramping the pulse envelope, and delivering equivalent doses via longer, lower-intensity pulses rather than shorter, higher-intensity pulses. Additionally, higher pulse repetition frequencies and fundamental frequencies reduce somatosensory effects. Through our systematic mapping of the parameter space, we also find preliminary evidence that particle displacement (strain) may be a primary biophysical driving force behind peripheral somatosensory co-stimulation. CONCLUSION: This study provides actionable strategies to minimise somatosensory confounds, which will support the thorough experimental control required to unlock the full potential of TUS for scientific research and clinical interventions.

  PublisherOA PDFDOI

• Optimized ultrasound neuromodulation for non-invasive control of behavior and physiology

  Neuron · 2024-07-29 · 87 citations

  articleOpen access
  PublisherOA PDFDOI

• Development and validation of a computational method to predict unintended auditory brainstem response during transcranial ultrasound neuromodulation in mice

  Brain stimulation · 2023 · 9 citations

  Senior authorCorresponding
  • Computer Science
  • Acoustics
  • Audiology

  BACKGROUND: Transcranial ultrasound stimulation (TUS) is a promising noninvasive neuromodulation modality. The inadvertent and unpredictable activation of the auditory system in response to TUS obfuscates the interpretation of non-auditory neuromodulatory responses. OBJECTIVE: The objective was to develop and validate a computational metric to quantify the susceptibility to unintended auditory brainstem response (ABR) in mice premised on time frequency analyses of TUS signals and auditory sensitivity. METHODS: Ultrasound pulses with varying amplitudes, pulse repetition frequencies (PRFs), envelope smoothing profiles, and sinusoidal modulation frequencies were selected. Each pulse's time-varying frequency spectrum was differentiated across time, weighted by the mouse hearing sensitivity, then summed across frequencies. The resulting time-varying function, computationally predicting the ABR, was validated against experimental ABR in mice during TUS with the corresponding pulse. RESULTS: = 0.97). CONCLUSIONS: To reduce ABR in mice during in vivo TUS studies, 1) reduce the amplitude of a rectangular continuous wave envelope, 2) increase the rise/fall times of a smoothed continuous wave envelope, and/or 3) change the PRF and/or duty cycle of a rectangular or sinusoidal pulsed wave to reduce the gap between pulses and increase the rise/fall time of the overall envelope. This metric can aid researchers performing in vivo mouse studies in selecting TUS signal parameters that minimize unintended ABR. The methods for developing this metric can be adapted to other animal models.

  PublisherOA PDFDOI

• Synthetic Skull CT Generation with Generative Adversarial Networks to Train Deep Learning Models for Clinical Transcranial Ultrasound

  arXiv (Cornell University) · 2023 · 2 citations

  Senior authorCorresponding
  • Computer Science
  • Artificial Intelligence
  • Computer Science

  Deep learning offers potential for various healthcare applications, yet requires extensive datasets of curated medical images where data privacy, cost, and distribution mismatch across various acquisition centers could become major problems. To overcome these challenges, we propose a generative adversarial network (SkullGAN) to create large datasets of synthetic skull CT slices, geared towards training models for transcranial ultrasound. With wide ranging applications in treatment of essential tremor, Parkinson's, and Alzheimer's disease, transcranial ultrasound clinical pipelines can be significantly optimized via integration of deep learning. The main roadblock is the lack of sufficient skull CT slices for the purposes of training, which SkullGAN aims to address. Actual CT slices of 38 healthy subjects were used for training. The generated synthetic skull images were then evaluated based on skull density ratio, mean thickness, and mean intensity. Their fidelity was further analyzed using t-distributed stochastic neighbor embedding (t-SNE), Fréchet inception distance (FID) score, and visual Turing test (VTT) taken by four staff clinical radiologists. SkullGAN-generated images demonstrated similar quantitative radiological features to real skulls. t-SNE failed to separate real and synthetic samples from one another, and the FID score was 49. Expert radiologists achieved a 60\% mean accuracy on the VTT. SkullGAN makes it possible for researchers to generate large numbers of synthetic skull CT segments, necessary for training neural networks for medical applications involving the human skull, such as transcranial focused ultrasound, mitigating challenges with access, privacy, capital, time, and the need for domain expertise.

  PublisherOA PDFDOI

• Focused ultrasound-induced inhibition of peripheral nerve fibers in an animal model of acute pain

  Regional Anesthesia & Pain Medicine · 2023-02-23 · 10 citations

  articleOpen access

  BACKGROUND: Moderate-to-severe acute pain is prevalent in many healthcare settings and associated with adverse outcomes. Peripheral nerve blockade using traditional needle-based and local anesthetic-based techniques improves pain outcomes for some patient populations but has shortcomings limiting use. These limitations include its invasiveness, potential for local anesthetic systemic toxicity, risk of infection with an indwelling catheter, and relatively short duration of blockade compared with the period of pain after major injuries. Focused ultrasound is capable of inhibiting the peripheral nervous system and has potential as a pain management tool. However, investigations of its effect on peripheral nerve nociceptive fibers in animal models of acute pain are lacking. In an in vivo acute pain model, we investigated focused ultrasound's effects on behavior and peripheral nerve structure. METHODS: Focused ultrasound was applied directly to the sciatic nerve of rats just prior to a hindpaw incision; three control groups (focused ultrasound sham only, hindpaw incision only, focused ultrasound sham+hindpaw incision) were also included. For all four groups (intervention and controls), behavioral testing (thermal and mechanical hyperalgesia, hindpaw extension and flexion) took place for 4 weeks. Structural changes to peripheral nerves of non-focused ultrasound controls and after focused ultrasound application were assessed on days 0 and 14 using light microscopy and transmission electron microscopy. RESULTS: Compared with controls, after focused ultrasound application, animals had (1) increased mechanical nociceptive thresholds for 2 weeks; (2) sustained increase in thermal nociceptive thresholds for ≥4 weeks; (3) a decrease in hindpaw motor response for 0.5 weeks; and (4) a decrease in hindpaw plantar sensation for 2 weeks. At 14 days after focused ultrasound application, alterations to myelin sheaths and nerve fiber ultrastructure were observed both by light and electron microscopy. DISCUSSION: Focused ultrasound, using a distinct parameter set, reversibly inhibits A-delta peripheral nerve nociceptive, motor, and non-nociceptive sensory fiber-mediated behaviors, has a prolonged effect on C nociceptive fiber-mediated behavior, and alters nerve structure. Focused ultrasound may have potential as a peripheral nerve blockade technique for acute pain management. However, further investigation is required to determine C fiber inhibition duration and the significance of nerve structural changes.

  PublisherOA PDFDOI

• Transcranial ultrasound neuromodulation of the thalamic visual pathway in a large animal model and the dose-response relationship with MR-ARFI

  Scientific Reports · 2022 · 31 citations

  Senior authorCorresponding
  • Neuroscience
  • Medicine
  • Psychology

  Neuromodulation of deep brain structures via transcranial ultrasound stimulation (TUS) is a promising, but still elusive approach to non-invasive treatment of brain disorders. The purpose of this study was to confirm that MR-guided TUS of the lateral geniculate nucleus (LGN) can modulate visual evoked potentials (VEPs) in the intact large animal; and to study the impact on cortical brain oscillations. The LGN on one side was identified with T2-weighted MRI in sheep (all male, n = 9). MR acoustic radiation force imaging (MR-ARFI) was used to confirm localization of the targeted area in the brain. Electroencephalographic (EEG) signals were recorded, and the visual evoked potential (VEP) peak-to-peak amplitude (N70 and P100) was calculated for each trial. Time-frequency spectral analysis was performed to elucidate the effect of TUS on cortical brain dynamics. The VEP peak-to-peak amplitude was reversibly suppressed relative to baseline during TUS. Dynamic spectral analysis demonstrated a change in cortical oscillations when TUS is paired with visual sensory input. Sonication-associated microscopic displacements, as measured by MR-ARFI, correlated with the TUS-mediated suppression of visual evoked activity. TUS non-invasively delivered to LGN can neuromodulate visual activity and oscillatory dynamics in large mammalian brains.

  PublisherOA PDFDOI

• Benchmark problems for transcranial ultrasound simulation: Intercomparison of compressional wave models

  The Journal of the Acoustical Society of America · 2022 · 111 citations

  • Computer Science
  • Computer Science
  • Acoustics

  Computational models of acoustic wave propagation are frequently used in transcranial ultrasound therapy, for example, to calculate the intracranial pressure field or to calculate phase delays to correct for skull distortions. To allow intercomparison between the different modeling tools and techniques used by the community, an international working group was convened to formulate a set of numerical benchmarks. Here, these benchmarks are presented, along with intercomparison results. Nine different benchmarks of increasing geometric complexity are defined. These include a single-layer planar bone immersed in water, a multi-layer bone, and a whole skull. Two transducer configurations are considered (a focused bowl and a plane piston operating at 500 kHz), giving a total of 18 permutations of the benchmarks. Eleven different modeling tools are used to compute the benchmark results. The models span a wide range of numerical techniques, including the finite-difference time-domain method, angular spectrum method, pseudospectral method, boundary-element method, and spectral-element method. Good agreement is found between the models, particularly for the position, size, and magnitude of the acoustic focus within the skull. When comparing results for each model with every other model in a cross-comparison, the median values for each benchmark for the difference in focal pressure and position are less than 10% and 1 mm, respectively. The benchmark definitions, model results, and intercomparison codes are freely available to facilitate further comparisons.

  PublisherOA PDFDOI

• Dose-dependent effects of high intensity focused ultrasound on compound action potentials in an ex vivo rodent peripheral nerve model: comparison to local anesthetics

  Regional Anesthesia & Pain Medicine · 2022-02-03 · 6 citations

  article

  BACKGROUND: In animal models, focused ultrasound can reversibly or permanently inhibit nerve conduction, suggesting a potential role in managing pain. We hypothesized focused ultrasound's effects on action potential parameters may be similar to those of local anesthetics. METHODS: In an ex vivo rat sciatic nerve model, action potential amplitude, area under the curve, latency to 10% peak, latency to 100% peak, rate of rise, and half peak width changes were assessed after separately applying increasing focused ultrasound pressures or concentrations of bupivacaine and ropivacaine. Focused ultrasound's effects on nerve structure were examined histologically. RESULTS: Increasing focused ultrasound pressures decreased action potential amplitude, area under the curve, and rate of rise, increased latency to 10% peak, and did not change latency to 100% peak or half peak width. Increasing local anesthetic concentrations decreased action potential amplitude, area under the curve, and rate of rise and increased latency to 10% peak, latency to 100% peak, and half peak width. At the highest focused ultrasound pressures, nerve architecture was altered compared with controls. DISCUSSION: While some action potential parameters were altered comparably by focused ultrasound and local anesthetics, there were small but notable differences. It is not evident if these differences may lead to differences in clinical pain effects when focused ultrasound is applied in vivo or if focused ultrasound pressures that result in clinically relevant changes damage nerve structures. Given the potential advantages of a non-invasive technique for managing pain conditions, further investigation may be warranted in an in vivo pain model.

  PublisherDOI

• Improving in situ acoustic intensity estimates using <scp>MR</scp> acoustic radiation force imaging in combination with multifrequency <scp>MR</scp> elastography

  Magnetic Resonance in Medicine · 2022-06-28 · 16 citations

  articleOpen accessSenior author

  PURPOSE: Magnetic resonance acoustic radiation force imaging (MR-ARFI) enables focal spot localization during nonablative transcranial ultrasound therapies. As the acoustic radiation force is proportional to the applied acoustic intensity, measured MR-ARFI displacements could potentially be used to estimate the acoustic intensity at the target. However, variable brain stiffness is an obstacle. The goal of this study was to develop and assess a method to accurately estimate the acoustic intensity at the focus using MR-ARFI displacements in combination with viscoelastic properties obtained with multifrequency MR elastography (MRE). METHODS: Phantoms with a range of viscoelastic properties were fabricated, and MR-ARFI displacements were acquired within each phantom using multiple acoustic intensities. Voigt model parameters were estimated for each phantom based on storage and loss moduli measured using multifrequency MRE, and these were used to predict the relationship between acoustic intensity and measured displacement. RESULTS: Using assumed viscoelastic properties, MR-ARFI displacements alone could not accurately estimate acoustic intensity across phantoms. For example, acoustic intensities were underestimated in phantoms stiffer than the assumed stiffness and overestimated in phantoms softer than the assumed stiffness. This error was greatly reduced using individualized viscoelasticity measurements obtained from MRE. CONCLUSION: We demonstrated that viscoelasticity information from MRE could be used in combination with MR-ARFI displacements to obtain more accurate estimates of acoustic intensity. Additionally, Voigt model viscosity parameters were found to be predictive of the relaxation rate of each phantom's time-varying displacement response, which could be used to optimize patient-specific MR-ARFI pulse sequences.

  PublisherOA PDFDOI

Recent grants

• NIH Grant R21EB003170

  NIH · $372k · 2007

• NIH Grant R01CA077677

  NIH · $1.4M · 2008

• MR-guided Focused Ultrasound Neuromodulation of Deep Brain Structures

  NIH · $1.6M · 2016–2020

• NIH Grant P01CA159992

  NIH · $14.7M · 2017

• Neurostimulation by Ultrasound: Physical, Biophysical and Neural Mechanisms

  NIH · $2.8M · 2014–2018

Frequent coauthors

• Ari Partanen

  Profound Medical (Canada)

  153 shared

• Vera A. Khokhlova

  Lomonosov Moscow State University

  147 shared

• Nathan McDannold

  Brigham and Women's Hospital

  142 shared

• Cyril Lafon

  Inserm

  131 shared

• Pejman Ghanouni

  Stanford University

  130 shared

• Mickaël Tanter

  Inserm

  125 shared

• Kullervo Hynynen

  Health Sciences Centre

  123 shared

• Tobias Preußer

  Fraunhofer Institute for Digital Medicine

  113 shared

Labs

• Radiological Sciences LabPI

Education

• PhD, Biophysical Sciences

  Mayo Graduate School

  1993

• BS, Physics

  Duke University

  1989

Awards & honors

• Fellow, ISMRM (2009)
• Distinguished Investigator, The Academy of Radiology Researc…
• Fellow, American Institute For Medical and Biological Engine…