Tag Archives: Health

Enabling large-scale health studies for the research community

Posted by Chintan Ghate, Software Engineer, and Diana Mincu, Research Engineer, Google Research

As consumer technologies like fitness trackers and mobile phones become more widely used for health-related data collection, so does the opportunity to leverage these data pathways to study and advance our understanding of medical conditions. We have previously touched upon how our work explores the use of this technology within the context of chronic diseases, in particular multiple sclerosis (MS). This effort leverages the FDA MyStudies platform, an open-source platform used to create clinical study apps, that makes it easier for anyone to run their own studies and collect good quality healthcare data, in a trusted and safe way.

Today, we describe the setup that we developed by expanding the FDA MyStudies platform and demonstrate how it can be used to set up a digital health study. We also present our exploratory research study created through this platform, called MS Signals, which consists of a symptom tracking app for MS patients. The goal for this app is twofold: 1) to ensure that the enhancements to the FDA MyStudies platform made for a more streamlined study creation experience; and 2) to understand how new data collection mechanisms can be used to revolutionize patients’ chronic disease management and tracking. We have open sourced our extension to the FDA MyStudies platform under the Apache 2.0 license to provide a resource for the community to build their own studies.


Extending the FDA MyStudies platform

The original FDA MyStudies platform allowed people to configure their own study apps, manage participants, and create separate iOS and Android apps. To simplify the study creation process and ensure increased study engagement, we made a number of accessibility changes. Some of the main improvements include: cross-platform (iOS and Android) app generation through the use of Flutter, an open source framework by Google for building multi-platform applications from a single codebase; a simplified setup, so that users can prototype their study quickly (under a day in most cases); and, most importantly, an emphasis on accessibility so that diverse patient’s voices are heard. The accessibility enhancements include changes to the underlying features of the platform and to the particular study design of the MS Signals study app.


Multi-platform support with rapid prototyping

We decided on the use of Flutter as it would be a single point that would generate both iOS and Android apps in one go, reducing the work required to support multiple platforms. Flutter also provides hot-reloading, which allows developers to build & preview features quickly. The design-system in the app takes advantage of this feature to provide a central point from which the branding & theme of the app can be changed to match the tone of a new study and previewed instantly. The demo environment in the app also utilizes this feature to allow developers to mock and preview questionnaires locally on their machines. In our experience this has been a huge time-saver in A/B testing the UX and the format and wording of questions live with clinicians.


System accessibility enhancements

To improve the accessibility of the platform for more users, we made several usability enhancements:

  1. Light & dark theme support
  2. Bold text & variable font-sizes
  3. High-contrast mode
  4. Improving user awareness of accessibility settings

Extended exposure to bright light themes can strain the eyes, so supporting dark theme features was necessary to make it easier to use the study app frequently. Some small or light text-elements are illegible to users with vision impairments, so we added 1) bold-text and support for larger font-sizes and 2) high-contrast color-schemes. To ensure that accessibility settings are easy to find, we placed an introductory one-time screen that was presented during the app’s first launch, which would directly take users to their system accessibility settings.


Study accessibility enhancements

To make the study itself easier to interact with and reduce cognitive overload, we made the following changes:

  1. Clarified the onboarding process
  2. Improved design for questionnaires

First, we clarified the on-boarding process by presenting users with a list of required steps when they first open the app in order to reduce confusion and participant drop-off.

The original questionnaire design in the app presented each question in a card format, which utilizes part of the screen for shadows and depth effects of the card. In many situations, this is a pleasant aesthetic, but in apps where accessibility is priority, these visual elements restrict the space available on the screen. Thus, when more accessible, larger font-sizes are used there are more frequent word breaks, which reduces readability. We fixed this simply by removing the card design elements and instead using the entire screen, allowing for better visuals with larger font-sizes.


The MS Signals prototype study

To test the usability of these changes, we used our redesigned platform to create a prototype study app called MS Signals, which uses surveys to gather information about a participant’s MS-related symptoms.

MS Signals app screenshots.

MS Studies app design

As a first step, before entering any study information, participants are asked to complete an eligibility and study comprehension questionnaire to ensure that they have read through the potentially lengthy terms of study participation. This might include, for example, questions like "In what country is the study available?" or “Can you withdraw from the study?" A section like this is common in most health studies, and it tends to be the first drop-off point for participants.

To minimize study drop-off at this early stage, we kept the eligibility test brief and reflected correct answers for the comprehension test back to the participants. This helps minimize the number of times a user may need to go through the initial eligibility questionnaire and ensures that the important aspects of the study protocol are made clear to them.

After successful enrollment, participants are taken to the main app view, which consists of three pages:

  • Activities:
    This page lists the questionnaires available to the participant and is where the majority of their time is spent. The questionnaires vary in frequency — some are one-time surveys created to gather medical history, while others are repeated daily, weekly or monthly, depending on the symptom or area they are exploring. For the one-time survey we provide a counter above each question to signal to users how far they have come and how many questions are left, similar to the questionnaire during the eligibility and comprehension step.
  • Dashboard:
    To ensure that participants get something back in return for the information they enter during a study, the Dashboard area presents a summary of their responses in graph or pie chart form. Participants could potentially show this data to their care provider as a summary of their condition over the last 6 months, an improvement over the traditional pen and paper methods that many employ today.
  • Resources:
    A set of useful links, help articles and common questions related to MS.

Questionnaire design

Since needing to frequently input data can lead to cognitive overload, participant drop off, and bad data quality, we reduced the burden in two ways:

  1. We break down large questionnaires into smaller ones, resulting in 6 daily surveys, containing 3–5 questions each, where each question is multiple choice and related to a single symptom. This way we cover a total of 20 major symptoms, and present them in a similar way to how a clinician would ask these questions in an in-clinic setting.
  2. We ensure previously entered information is readily available in the app, along with the time of the entry.

In designing the survey content, we collaborated closely with experienced clinicians and researchers to finalize the wording and layout. While studies in this field typically use the Likert scale to gather symptom information, we defined a more intuitive verbose scale to provide better experience for participants tracking their disease and the clinicians or researchers viewing the disease history. For example, in the case of vision issues, rather than asking participants to rate their symptoms on a scale from 1 to 10, we instead present a multiple choice question where we detail common vision problems that they may be experiencing.

This verbose scale helps patients track their symptoms more accurately by including context that helps them more clearly define their symptoms. This approach also allows researchers to answer questions that go beyond symptom correlation. For example, for vision issues, data collected using the verbose scale would reveal to researchers whether nystagmus is more prominent in patients with MS compared to double vision.

Side-by-side comparison with a Likert scale on the left, and a Verbose scale on the right.

Focusing on accessibility

Mobile-based studies can often present additional challenges for participants with chronic conditions: the text can be hard to read, the color contrast could make it difficult to see certain bits of information, or it may be challenging to scroll through pages. This may result in participant drop off, which, in turn, could yield a biased dataset if the people who are experiencing more advanced forms of a disease are unable to provide data.

In order to prevent such issues, we include the following accessibility features:

  • Throughout, we employ color blind accessible color schemes. This includes improving the contrast between crucial text and important additional information, which might otherwise be presented in a smaller font and a faded text color.
  • We reduced the amount of movement required to access crucial controls by placing all buttons close to the bottom of the page and ensuring that pop-ups are controllable from the bottom part of the screen.

To test the accessibility of MS Signals, we collaborated with the National MS Society to recruit participants for a user experience study. For this, a call for participation was sent out by the Society to their members, and 9 respondents were asked to test out the various app flows. The majority indicated that they would like a better way than their current method to track their symptom data, that they considered MS Signals to be a unique and valuable tool that would enhance the accuracy of their symptom tracking, and that they would want to share the dashboard view with their healthcare providers.


Next steps

We want to encourage everyone to make use of the open source platform to start setting up and running their own studies. We are working on creating a set of standard study templates, which would incorporate what we learned from above, and we hope to release those soon. For any issues, comments or questions please check out our resource page.

Source: Google AI Blog


Responsible AI at Google Research: Context in AI Research (CAIR)

Posted by Katherine Heller, Research Scientist, Google Research, on behalf of the CAIR Team

Artificial intelligence (AI) and related machine learning (ML) technologies are increasingly influential in the world around us, making it imperative that we consider the potential impacts on society and individuals in all aspects of the technology that we create. To these ends, the Context in AI Research (CAIR) team develops novel AI methods in the context of the entire AI pipeline: from data to end-user feedback. The pipeline for building an AI system typically starts with data collection, followed by designing a model to run on that data, deployment of the model in the real world, and lastly, compiling and incorporation of human feedback. Originating in the health space, and now expanded to additional areas, the work of the CAIR team impacts every aspect of this pipeline. While specializing in model building, we have a particular focus on building systems with responsibility in mind, including fairness, robustness, transparency, and inclusion.


Data

The CAIR team focuses on understanding the data on which ML systems are built. Improving the standards for the transparency of ML datasets is instrumental in our work. First, we employ documentation frameworks to elucidate dataset and model characteristics as guidance in the development of data and model documentation techniques — Datasheets for Datasets and Model Cards for Model Reporting.

For example, health datasets are highly sensitive and yet can have high impact. For this reason, we developed Healthsheets, a health-contextualized adaptation of a Datasheet. Our motivation for developing a health-specific sheet lies in the limitations of existing regulatory frameworks for AI and health. Recent research suggests that data privacy regulation and standards (e.g., HIPAA, GDPR, California Consumer Privacy Act) do not ensure ethical collection, documentation, and use of data. Healthsheets aim to fill this gap in ethical dataset analysis. The development of Healthsheets was done in collaboration with many stakeholders in relevant job roles, including clinical, legal and regulatory, bioethics, privacy, and product.

Further, we studied how Datasheets and Healthsheets could serve as diagnostic tools that surface the limitations and strengths of datasets. Our aim was to start a conversation in the community and tailor Healthsheets to dynamic healthcare scenarios over time.

To facilitate this effort, we joined the STANDING Together initiative, a consortium that aims to develop international, consensus-based standards for documentation of diversity and representation within health datasets and to provide guidance on how to mitigate risk of bias translating to harm and health inequalities. Being part of this international, interdisciplinary partnership that spans academic, clinical, regulatory, policy, industry, patient, and charitable organizations worldwide enables us to engage in the conversation about responsibility in AI for healthcare internationally. Over 250 stakeholders from across 32 countries have contributed to refining the standards.

Healthsheets and STANDING Together: towards health data documentation and standards.

Model

When ML systems are deployed in the real world, they may fail to behave in expected ways, making poor predictions in new contexts. Such failures can occur for a myriad of reasons and can carry negative consequences, especially within the context of healthcare. Our work aims to identify situations where unexpected model behavior may be discovered, before it becomes a substantial problem, and to mitigate the unexpected and undesired consequences.

Much of the CAIR team’s modeling work focuses on identifying and mitigating when models are underspecified. We show that models that perform well on held-out data drawn from a training domain are not equally robust or fair under distribution shift because the models vary in the extent to which they rely on spurious correlations. This poses a risk to users and practitioners because it can be difficult to anticipate model instability using standard model evaluation practices. We have demonstrated that this concern arises in several domains, including computer vision, natural language processing, medical imaging, and prediction from electronic health records.

We have also shown how to use knowledge of causal mechanisms to diagnose and mitigate fairness and robustness issues in new contexts. Knowledge of causal structure allows practitioners to anticipate the generalizability of fairness properties under distribution shift in real-world medical settings. Further, investigating the capability for specific causal pathways, or “shortcuts”, to introduce bias in ML systems, we demonstrate how to identify cases where shortcut learning leads to predictions in ML systems that are unintentionally dependent on sensitive attributes (e.g., age, sex, race). We have shown how to use causal directed acyclic graphs to adapt ML systems to changing environments under complex forms of distribution shift. Our team is currently investigating how a causal interpretation of different forms of bias, including selection bias, label bias, and measurement error, motivates the design of techniques to mitigate bias during model development and evaluation.

Shortcut Learning: For some models, age may act as a shortcut in classification when using medical images.

The CAIR team focuses on developing methodology to build more inclusive models broadly. For example, we also have work on the design of participatory systems, which allows individuals to choose whether to disclose sensitive attributes, such as race, when an ML system makes predictions. We hope that our methodological research positively impacts the societal understanding of inclusivity in AI method development.


Deployment

The CAIR team aims to build technology that improves the lives of all people through the use of mobile device technology. We aim to reduce suffering from health conditions, address systemic inequality, and enable transparent device-based data collection. As consumer technology, such as fitness trackers and mobile phones, become central in data collection for health, we explored the use of these technologies within the context of chronic disease, in particular, for multiple sclerosis (MS). We developed new data collection mechanisms and predictions that we hope will eventually revolutionize patient’s chronic disease management, clinical trials, medical reversals and drug development.

First, we extended the open-source FDA MyStudies platform, which is used to create clinical study apps, to make it easier for anyone to run their own studies and collect good quality data, in a trusted and safe way. Our improvements include zero-config setups, so that researchers can prototype their study in a day, cross-platform app generation through the use of Flutter and, most importantly, an emphasis on accessibility so that all patient’s voices are heard. We are excited to announce this work has now been open sourced as an extension to the original FDA-Mystudies platform. You can start setting up your own studies today!

To test this platform, we built a prototype app, which we call MS Signals, that uses surveys to interface with patients in a novel consumer setting. We collaborated with the National MS Society to recruit participants for a user experience study for the app, with the goal of reducing dropout rates and improving the platform further.

MS Signals app screenshots. Left: Study welcome screen. Right: Questionnaire.

Once data is collected, researchers could potentially use it to drive the frontier of ML research in MS. In a separate study, we established a research collaboration with the Duke Department of Neurology and demonstrated that ML models can accurately predict the incidence of high-severity symptoms within three months using continuously collected data from mobile apps. Results suggest that the trained models can be used by clinicians to evaluate the symptom trajectory of MS participants, which may inform decision making for administering interventions.

The CAIR team has been involved in the deployment of many other systems, for both internal and external use. For example, we have also partnered with Learning Ally to build a book recommendation system for children with learning disabilities, such as dyslexia. We hope that our work positively impacts future product development.


Human feedback

As ML models become ubiquitous throughout the developed world, it can be far too easy to leave voices in less developed countries behind. A priority of the CAIR team is to bridge this gap, develop deep relationships with communities, and work together to address ML-related concerns through community-driven approaches.

One of the ways we are doing this is through working with grassroots organizations for ML, such as Sisonkebiotik, an open and inclusive community of researchers, practitioners and enthusiasts at the intersection of ML and healthcare working together to build capacity and drive forward research initiatives in Africa. We worked in collaboration with the Sisonkebiotik community to detail limitations of historical top-down approaches for global health, and suggested complementary health-based methods, specifically those of grassroots participatory communities (GPCs). We jointly created a framework for ML and global health, laying out a practical roadmap towards setting up, growing and maintaining GPCs, based on common values across various GPCs such as Masakhane, Sisonkebiotik and Ro’ya.

We are engaging with open initiatives to better understand the role, perceptions and use cases of AI for health in non-western countries through human feedback, with an initial focus in Africa. Together with Ghana NLP, we have worked to detail the need to better understand algorithmic fairness and bias in health in non-western contexts. We recently launched a study to expand on this work using human feedback.

Biases along the ML pipeline and their associations with African-contextualized axes of disparities.

The CAIR team is committed to creating opportunities to hear more perspectives in AI development. We partnered with Sisonkebiotik to co-organize the Data Science for Health Workshop at Deep Learning Indaba 2023 in Ghana. Everyone’s voice is crucial to developing a better future using AI technology.


Acknowledgements

We would like to thank Negar Rostamzadeh, Stephen Pfohl, Subhrajit Roy, Diana Mincu, Chintan Ghate, Mercy Asiedu, Emily Salkey, Alexander D’Amour, Jessica Schrouff, Chirag Nagpal, Eltayeb Ahmed, Lev Proleev, Natalie Harris, Mohammad Havaei, Ben Hutchinson, Andrew Smart, Awa Dieng, Mahima Pushkarna, Sanmi Koyejo, Kerrie Kauer, Do Hee Park, Lee Hartsell, Jennifer Graves, Berk Ustun, Hailey Joren, Timnit Gebru and Margaret Mitchell for their contributions and influence, as well as our many friends and collaborators at Learning Ally, National MS Society, Duke University Hospital, STANDING Together, Sisonkebiotik, and Masakhane.

Source: Google AI Blog


Audioplethysmography for cardiac monitoring with hearable devices

Posted by Xiaoran "Van" Fan, Experimental Scientist, and Trausti Thormundsson, Director, Google

The market for true wireless stereo (TWS) active noise canceling (ANC) hearables (headphones and earbuds) has been soaring in recent years, and the global shipment volume will nearly double that of smart wristbands and watches in 2023. The on-head time for hearables has extended significantly due to the recent advances in ANC, transparency mode, and artificial intelligence. Users frequently wear hearables not just for music listening, but also for exercising, focusing, or simply mood adjustment. However, hearable health is still mostly uncharted territory for the consumer market.

In “APG: Audioplethysmography for Cardiac Monitoring in Hearables,” presented at MobiCom 2023, we introduce a novel active in-ear health sensing modality. Audioplethysmography (APG) enables ANC hearables to monitor a user's physiological signals, such as heart rate and heart rate variability, without adding extra sensors or compromising battery life. APG exhibits high resilience to motion artifacts, adheres to safety regulations with an 80 dB margin below the limit, remains unaffected by seal conditions, and is inclusive of all skin tones.

APG sends a low intensity ultrasound transmitting wave (TX wave) using an ANC headphone's speakers and collects the receiving wave (RX wave) via the on-board feedback microphones. The APG signal is a pulse-like waveform that synchronizes with heartbeat and reveals rich cardiac information, such as dicrotic notches.


Health sensing in the ear canal

The auditory canal receives its blood supply from the arteria auricularis profunda, also known as the deep ear artery. This artery forms an intricate network of smaller vessels that extensively permeate the auditory canal. Slight variations in blood vessel shape caused by the heartbeat (and blood pressure) can lead to subtle changes in the volume and pressure of the ear canals, making the ear canal an ideal location for health sensing.

Recent research has explored using hearables for health sensing by packaging together a plethora of sensors — e.g., photoplethysmograms (PPG) and electrocardiograms (ECG) — with a microcontroller to enable health applications, such as sleep monitoring, heart rate and blood pressure tracking. However, this sensor mounting paradigm inevitably adds cost, weight, power consumption, acoustic design complexity, and form factor challenges to hearables, constituting a strong barrier to its wide adoption.

Existing ANC hearables deploy feedback and feedforward microphones to navigate the ANC function. These microphones create new opportunities for various sensing applications as they can detect or record many bio-signals inside and outside the ear canal. For example, feedback microphones can be used to listen to heartbeats and feedforward microphones can hear respirations. Academic research on this passive sensing paradigm has prompted many mobile applications, including heart rate monitoring, ear disease diagnosis, respiration monitoring, and body activity recognition. However, microphones in consumer-grade ANC headphones come with built-in high-pass filters to prevent saturation from body motions or strong wind noise. The signal quality of passive listening in the ear canal also heavily relies on the earbud seal conditions. As such, it is challenging to embed health features that rely on the passive listening of low frequency signals (≤ 50 Hz) on commercial ANC headphones.


Measuring tiny physiological signals

APG bypasses the aforementioned ANC headphone hardware constraints by sending a low intensity ultrasound probing signal through an ANC headphone's speakers. This signal triggers echoes, which are received via on-board feedback microphones. We observe that the tiny ear canal skin displacement and heartbeat vibrations modulate these ultrasound echoes.

We build a cylindrical resonance model to understand APG’s underlying physics. This phenomenon happens at an extremely small scale, which makes the raw pulse signal invisible in the raw received ultrasound. We adopt coherent detection to retrieve this micro physiological modulation under the noise floor (we term this retrieved signal as mixed-down signal, see the paper for more details). The final APG waveform looks strikingly similar to a PPG waveform, but provides an improved view of cardiac activities with more pronounced dicrotic notches (i.e., pressure waveforms that provide rich insights about the central artery system, such as blood pressure).

A cylindrical model with cardiac activities ℎ(𝑡) that modulates both the phase and amplitude of the mixed-down signal. Based on the simulation from our analytical model, the amplitude 𝑅(𝑡) and phase Φ(𝑡) of the mixed-down APG signals both reflect the cardiac activities ℎ(𝑡).


APG sensing in practice

During our initial experiments, we observed that APG works robustly with bad earbuds seals and with music playing. However, we noticed the APG signal can sometimes be very noisy and could be heavily disturbed by body motion. At that point, we determined that in order to make APG useful, we had to make it more robust to compete with more than 80 years of PPG development.

While PPGs are widely used and highly advanced, they do have some limitations. For example, PPGs sensors typically use two to four diodes to send and receive light frequencies for sensing. However, due to the ultra high-frequency nature (hundreds of Terahertz) of the light, it's difficult for a single diode to send multiple colors with different frequencies. On the other hand, we can easily design a low-cost and low-power system that generates and receives more than ten audio tones (frequencies). We leverage channel diversity, a physical phenomenon that describes how wireless signals (e.g., light and audio) at different frequencies have different characters (e.g., different attenuation and reflection coefficients) when the signal propagates in a medium, to enable a higher quality APG signal and motion resilience.

Next, we experimentally demonstrate the effectiveness of using multiple frequencies in the APG signaling. We transmit three probing signals concurrently with their frequencies spanning evenly from 30 KHz to 32 KHz. A participant was asked to shake their head four times during the experiment to introduce interference. The figure below shows that different frequencies can be transmitted simultaneously to gather various information with coherent detection, a unique advantage to APG.

The 30 kHz phase shows the four head movements and the magnitude (amplitude) of 31 kHz shows the pulse wave signal. This observation shows that some ultrasound frequencies might be sensitive to cardiac activities while others might be sensitive to motion. Therefore, we can use the multi-tone APG as a calibration signal to find the best frequency that measures heart rate, and use only the best frequency to get high-quality pulse waveform.

The mixed-down amplitude (upper row) and phase (bottom row) for a customized multi-tone APG signal that spans from 30 kHz to 32 kHz. With channel diversity, the cardiac activities are captured in some frequencies (e.g., magnitude of 31 kHz) and head movements are captured in other frequencies (e.g., magnitude of 30 kHz, 30 kHz, and phase of 31 kHz).

After choosing the best frequency to measure heart rate, the APG pulse waveform becomes more visible with pronounced dicrotic notches , and enables accurate heart rate variability measurement.

The final APG signal used in the measurement phase (left) and chest ECG signal (right).

Multi-tone translates to multiple simultaneous observations, which enable the development of array signal processing techniques. We demonstrate the spectrogram of a running session APG experiment before and after applying blind source separation (see the paper for more details). We also show the ground truth heart rate measurement in the same running experiment using a Polar ECG chest strap. In the raw APG, we see the running cadence (around 3.3 Hz) as well as two dim lines (around 2 Hz and 4 Hz) that indicate the user’s heart rate frequency and its harmonics. The heart rate frequencies are significantly enhanced in signal to noise ratio (SNR) after the blind source separation, which align with the ground truth heart rate frequencies. We also show the calculated heart rate and running cadence from APG and ECG. We can see that APG tracks the growth of heart rate during the running session accurately.

APG tracks the heart rate accurately during the running session and also measures the running cadence.


Field study and closing thoughts

We conducted two rounds of user experience (UX) studies with 153 participants. Our results demonstrate that APG achieves consistently accurate heart rate (3.21% median error across participants in all activity scenarios) and heart rate variability (2.70% median error in inter-beat interval) measurements. Unlike PPG, which exhibits variable performance across skin tones, our study shows that APG is resilient to variation in: skin tone, sub-optimal seal conditions, and ear canal size. More detailed evaluations can be found in the paper.

APG transforms any TWS ANC headphones into smart sensing headphones with a simple software upgrade, and works robustly across various user activities. The sensing carrier signal is completely inaudible and not impacted by music playing. More importantly, APG represents new knowledge in biomedical and mobile research and unlocks new possibilities for low-cost health sensing.


Acknowledgements

APG is the result of collaboration across Google Health, product, UX and legal teams. We would like to thank David Pearl, Jesper Ramsgaard, Cody Wortham, Octavio Ponce, Patrick Amihood, Sam Sheng, Michael Pate, Leonardo Kusumo, Simon Tong, Tim Gladwin, Russ Mirov, Kason Walker, Govind Kannan, Jayvon Timmons, Dennis Rauschmayer, Chiong Lai, Shwetak Patel, Jake Garrison, Anran Wang, Shiva Rajagopal, Shelten Yuen, Seobin Jung, Yun Liu, John Hernandez, Issac Galatzer-Levy, Isaiah Fischer-Brown, Jamie Rogers, Pramod Rudrapatna, Andrew Barakat, Jason Guss, Ethan Grabau, Pol Peiffer, Bill Park, Helen O'Connor, Mia Cheng, Keiichiro Yumiba, Felix Bors, Priyanka Jantre, Luzhou Xu, Jian Wang, Jaime Lien, Gerry Pallipuram, Nicholas Gillian, Michal Matuszak, Jakub Wojciechowski, Bryan Allen, Jane Hilario, and Phil Carmack for their invaluable insights and support. Thanks to external collaborators Longfei Shangguan and Rich Howard, Rutgers University and University of Pittsburgh.

Source: Google AI Blog