A team of researchers at UCLA has developed a revolutionary wearable device that allows individuals to speak without relying on their vocal folds. This breakthrough technology holds immense potential for assisting people with voice disorders resulting from various vocal fold conditions or postoperative recovery from laryngeal cancer surgeries.

The self-powered wearable sensing-actuation system is based on soft magnetoelasticity. Weighing only 7.2 grams and boasting a skin-like modulus, this lightweight device is stable against skin perspiration and offers a maximum stretchability of 164%. Its sensing component effectively captures the movement of extrinsic laryngeal muscles and converts them into high-fidelity electrical signals.

With the assistance of machine learning algorithms, the electrical signals are then translated into speech signals with an impressive accuracy rate of 94.68%. The wearable actuation component further converts these signals into voice signals, facilitating speech expression without relying on vocal fold vibrations.

Voice disorders can have a significant impact on an individual’s quality of life, hindering their ability to communicate effectively. Traditional solutions, such as handheld electrolarynx devices or invasive medical procedures, come with their own set of limitations. However, this wearable device offers a non-invasive and convenient alternative, significantly improving the daily lives of patients with dysfunctional vocal folds.

The device’s design incorporates a thin, flexible, and adhesive structure, providing optimal comfort for users. It consists of a polydimethylsiloxane (PDMS) layer, a magnetic induction (MI) layer made of serpentine copper coil, and a magnetomechanical coupling (MC) layer composed of magnetoelastic materials. The MC layer, featuring a kirigami structure, enhances the device’s sensitivity and stretchability.

In addition to its remarkable functionality, the wearable sensing-actuation system is intrinsically waterproof, ensuring durability even in the presence of heavy perspiration. It has been successfully tested for daily language transmissions, delivering clear and accurate output of voice signals.

The implications of this innovative technology extend beyond restoring normal voice function. By significantly reducing recovery time and eliminating the need for postoperative periods of absolute voice rest, the wearable device has the potential to enhance the overall quality of life for individuals with voice disorders.

With the successful development of this wearable sensing-actuation system, the researchers envision a game-changing solution for voice disorders that not only revolutionizes communication for patients but also paves the way for future advancements in wearable bioelectronics.

Further research and development are underway to refine and optimize this technology, with the goal of making it readily available to those who could benefit from it. As this device continues to evolve, it holds tremendous promise for transforming the lives of individuals with voice disorders, offering them a new means of communication and a renewed sense of empowerment.

WORCESTER, Mass.–(BUSINESS WIRE)–UMass Memorial Health, a leading healthcare system committed to delivering world-class patient care and innovation, today announced it is working with Google Cloud to leverage artificial intelligence (AI), machine learning (ML), and advanced data analytics to bring personalized care to cardiometabolic patients. This collaboration marks a significant step towards advancing patient care in the field of cardiometabolic health.

Cardiometabolic diseases, which include conditions like heart disease, diabetes, kidney disease and obesity, affect millions of people worldwide. In the United States, heart disease alone is the leading cause of death for men and women; in 2021, it accounted for 1 in every 5 deaths, according to the Centers for Disease Control. Identifying and treating patients for advanced cardiometabolic therapies can improve healthcare outcomes and reduce the burden of these chronic conditions on patients.

This partnership will leverage the power of data and predictive analytics to transform the way healthcare providers deliver personalized care. Specifically, UMass Memorial Health will use Google Cloud’s cutting-edge AI and ML capabilities, including BigQuery and Vertex AI, Healthcare Natural Language API, and Looker, to build tools for enhanced patient care and to advance its research. Key highlights include:

• Advanced Predictive Analytics: UMass Memorial Health is developing advanced predictive analytics models to enable it to analyze vast datasets, including patient health records, clinical data, and biometric information, to identify individuals who are suitable candidates for advanced cardiometabolic therapies.
• Enhanced Patient Care: The collaboration will empower healthcare professionals at UMass Memorial Health with powerful tools to make data-driven decisions and deliver personalized care plans. By identifying eligible patients more accurately and efficiently, the goal is to improve patient outcomes and administrative processes throughout the healthcare system.
• Research and Innovation: This partnership will facilitate ongoing research and development efforts to enhance the understanding of cardiometabolic diseases and treatment options. It will support the creation of new therapeutic protocols that can positively impact the lives of patients in the long term.
• Community Impact: By improving the accuracy of identifying patients for advanced therapies, this partnership has the potential to make a significant difference in the lives of individuals dealing with cardiometabolic conditions. It will contribute to better health outcomes, cost-effectiveness, and a higher quality of life for patients in the communities UMass Memorial Health serves.

Google Cloud and UMass Memorial Health are committed to ensuring patient privacy and data security. Google Cloud’s customers retain control over their data. Access and use of patient data are protected through the implementation of Google Cloud’s reliable infrastructure and secure data storage that support HIPAA compliance, and UMass Memorial Health’s security, privacy controls, and processes.

“Our mission is to provide the best possible care to our patients, and this partnership with Google Cloud is a significant step forward,” said Michael Hyder, M.D., MPH, Executive Director of the UMass Memorial Center for Digital Health Solutions and Associate Professor of Cardiovascular Medicine. “By using data-driven insights to identify patients who would benefit from advanced cardiometabolic therapies, we aim to elevate the high standard of care we provide.”

“We are just scratching the surface in terms of the opportunities to apply AI and advanced analytics to complex medical challenges,” said Aashima Gupta, global director of Healthcare Strategy and Solutions, Google Cloud. “This partnership exemplifies the potential of technology and healthcare institutions working together to address complex medical challenges. We are honored to partner with UMass Memorial Health to support their critical work in the field of cardiometabolic health, and we expect this to serve as a model for healthcare innovation globally.”

About UMass Memorial Health

UMass Memorial Health is the largest not-for-profit health care system in Central Massachusetts with 17,000 employees and 2,100 physicians. Our comprehensive system includes UMass Memorial Medical Center, UMass Memorial Health – Harrington, UMass Memorial Health – HealthAlliance-Clinton Hospital, UMass Memorial Health – Marlborough Hospital, UMass Memorial Health – Community Healthlink, and UMass Memorial Medical Group. We are the clinical partner of UMass Chan Medical School. For more information, visit www.ummhealth.org.

Contacts

Shelly Hazlett, UMass Memorial Health, Shelly.Hazlett@umassmemorial.org

Digital Clock and Recall (DCR

Superiority of the DCR Assessment

The study, titled “Digital Clock and Recall is Superior to the Mini-Mental State Examination (MMSE) for the detection of mild cognitive impairment and mild dementia,” adds to the growing body of evidence supporting the effectiveness of Linus Health’s DCR assessment. Over 20 other studies published in peer-reviewed scientific journals have also validated the DCR assessment within the Core Cognitive Evaluation (CCE) solution. This latest research confirms that the DCR assessment surpasses the widely-used Mini-Mental State Examination (MMSE) in detecting early signs of cognitive impairment with greater accuracy and less biased outcomes.

Advantages and Applications of the DCR

The DCR assessment, a quick and accurate test, allows medical professionals to assess a patient’s cognitive functioning across various domains. It provides essential data that is used by Linus Health’s machine-learning algorithm to determine a patient’s level of cognitive functioning and predict their future risk of developing dementia. By facilitating early screenings with immediate results, the DCR assessment enables proactive interventions and personalized care for patients.

Addressing the Need for Early Detection

According to the Alzheimer’s Association, more than 80% of Americans are unfamiliar with mild cognitive impairment, which can be an early stage of Alzheimer’s disease. The symptoms of cognitive impairment are often not noticeable, leading to delayed screenings and an increased risk to patients’ brain health. The DCR assessment aims to bridge this gap by offering a quick, accessible, and comprehensive tool that helps healthcare providers identify cognitive impairment at its early stages, providing opportunities for timely interventions and improved patient outcomes.

Study Design and Results

The study involved a clinical research team from Linus Health who compared the cognitive classifications of the DCR assessment with the MMSE test, which is commonly performed using pen and paper. The research included a population of 706 participants, with an average age of 71 years, including 59% women and 85% white individuals. Participants underwent various neuropsychological tests and were evaluated by clinicians. Results showed that the DCR assessment had superior sensitivity and accuracy compared to the MMSE in detecting and classifying cognitive impairment. The DCR assessment successfully detected cognitive impairment in over 80% of the patients who were misclassified by the MMSE, providing valuable insights into the potential of digital assessments for more accurate and equitable healthcare.

Unlike the MMSE, the DCR assessment is not influenced by ethnicity, thereby addressing the issue of health disparities among ethnic minorities at higher risk of dementia. Additionally, the DCR assessment can be administered in multiple languages, ensuring increased accessibility for a diverse patient population. By analyzing over 500 metrics captured during the three-minute test, the DCR assessment can detect subtle signals related to various cognitive domains.

Linus Health’s Digital Clock and Recall (DCR) assessment has emerged as a superior alternative to traditional paper-based tests for early detection of cognitive impairment and dementia. With its enhanced sensitivity, accuracy, and reduced ethnic bias, the DCR assessment offers the potential to revolutionize the screening and management of cognitive health. By facilitating early intervention and personalized care, this digital assessment opens doors to a future where individuals can receive timely and proactive support to maintain optimal brain health.

Upon thorough examination of 41 articles, the review discovered that overall perceptions of ML-based risk prediction models were positive. However, the findings also revealed areas of concern and reservations among healthcare professionals and patients.

One of the main gaps in knowledge identified in the study was the need for further research to determine optimal methods for model explanation and alerts. Ensuring transparency and explainability of ML models is crucial to gaining trust and acceptance from both healthcare professionals and patients. Developing robust techniques for explaining model predictions can help alleviate concerns and increase the adoption of ML-based risk prediction models in healthcare.

Additionally, the review highlighted the importance of understanding patients’ perceptions of ML-based predictive models as they play a crucial role in accepting and utilizing these models in their healthcare decisions. Therefore, it is essential to also consider their perspectives, preferences, and understanding of the technology to ensure effective communication and successful implementation.

The study pointed out that as the field of personalized medicine continues to advance, there is an increasing demand for complex risk prediction tools. ML methods, such as polygenic risk scores, are being explored to provide more accurate assessments of individual risk profiles. However, the current use of these tools is limited within healthcare settings, further emphasizing the need for research and validation to pave the way for wider adoption.

In summary, while ML-based risk prediction models show promise in healthcare, the systematic review revealed the importance of addressing concerns and reservations among healthcare professionals and patients. By bridging knowledge gaps, developing transparent explanations, and understanding patient perceptions, the field can work towards maximizing the potential benefits of ML in risk prediction and personalized medicine. Further research and collaboration between researchers, healthcare professionals, and patients are key to achieving this goal.

Led by Amitava Banerjee and colleagues, the study analyzed extensive electronic health record data from a cohort of 313,062 patients sourced from The Health Improvement Network and Clinical Practice Research Datalink databases. To enhance the robustness of their findings, the researchers cross-referenced the data with the Hospital Episode Statistics, the UK death registry, and the UK Biobank. By employing unsupervised and supervised machine learning methods, the team successfully identified five distinct clusters of heart failure patients: early onset, late onset, atrial fibrillation-related, metabolic, and cardiometabolic.

Furthermore, the researchers delved into the underlying biological mechanisms associated with each heart failure subtype by examining polygenic risk scores and single-nucleotide polymorphisms (SNPs). The analysis revealed a specific SNP linked to the atrial fibrillation-related subtype, while polygenic risk scores for hypertension, myocardial infarction, and obesity were associated with the late-onset and cardiometabolic subtypes. Leveraging supervised machine learning, the team developed a prediction model, complete with an online risk calculator accessible to both patients and clinicians.

The study’s comprehensive approach and utilization of large-scale data warrant recognition. While previous research has employed machine learning to subtype heart failure patients, this study distinguishes itself by incorporating a significantly larger study population and reinforcing its findings through external validation. Additionally, the genetic validation sheds light on how big data associations can be correlated with underlying biological processes.

Although the study is a remarkable advancement, it does have certain limitations, which the authors acknowledge. One limitation is the lack of data on ejection fraction and imaging study results, which could have strengthened the results. The prediction model exhibited only moderate discriminatory ability regarding 1-year mortality and could benefit from focusing on individual heart failure subtypes to reduce data heterogeneity. Furthermore, more advanced artificial intelligence algorithms, such as convolutional neural networks and eXtreme gradient boosting, could enhance risk prediction.

The researchers’ findings hold promising implications for the future of heart failure treatment. By identifying novel subtypes, this research may facilitate the development of targeted therapies that can benefit patients. Additionally, further investigation into the potential of AI-powered pattern recognition techniques in echocardiography and cardiac MRI could advance the field, enabling the subtyping of heart failure, outcome prediction, and treatment response assessment.

As the prevalence of heart failure continues to rise, the integration of machine learning and big data analysis may offer new avenues for understanding and managing this complex condition. The ability to tailor treatment strategies to individual heart failure subtypes holds tremendous potential for improving patient outcomes and quality of life.

The Koneksa-sponsored clinical validation study, “A basket observational study to determine usability, analytical validity, clinical vaLidity, and biomarker discovery for at-home EEG, weArable, and mobile device collected objective measuRemeNt of disturbed Sleep and neurologic disorders (LEARNS),” is a syndicated, observational study to discover biomarkers and determine the usability and validity of measures of sleep and neurologic disorders collected using at-home EEG, wearables, and smartphones.

The LEARNS study will investigate the Koneksa neuroscience solution, as well as the Beacon Biosignals platform, applied to at-home EEG, compared with in-clinic polysomnography, in neurological and sleep disorders. Current cohorts in the LEARNS study include Parkinson’s disease, narcolepsy, obstructive sleep apnea, Alzheimer’s disease, Huntington’s disease, and mild cognitive impairment.

“The utilization of evidence-based digital biomarkers is the key to transforming healthcare,” said Chris Benko, CEO, Koneksa. “We’re on a mission to revolutionize drug development in clinical trials by harnessing the power of our digital solutions with real-time, data-driven insights. Our collaboration with Beacon Biosignals will enable sponsors to integrate our digital solutions into clinical trials to identify treatment signals earlier and faster than with traditional measures.”

“The LEARNS study seeks to generate evidence to extend the Koneksa solution to new digital modalities and diseases,” said John Wagner, MD, PhD, Chief Medical Officer, Koneksa. “Beacon’s mobile EEG will join more than a dozen digital health technologies that are already integrated into the Koneksa solution. Our goal is to provide the widest possible range of validated, scalable digital measures to sponsors in one platform.”

“Partnering with Koneksa is a testament to our dedication to pioneering innovation,” said Jacob Donoghue, MD, PhD, Co-Founder and CEO, Beacon Biosignals. “With our shared science-first ethos, we are united in our relentless pursuit to develop cutting-edge tools that revolutionize drug development and accelerate the delivery of life-changing therapies to patients.”

“We are excited to partner with Koneksa on the LEARNS trial. This opportunity will expand the reach of Beacon’s EEG-based precision neurodiagnostics to more sponsors, and will provide critical validation of methodology that we believe could be transformational to patients,” said Christine Vietz, PhD, Beacon’s Chief Research and Development Officer.

The LEARNS study will measure the validity of smartphone-based motor, speech, and cognitive assessments versus gold-standard tools. The study will also evaluate whether assessment combinations can identify diagnostic features at the cohort diagnostic boundary. LEARNS has the potential to expand not only the Koneksa platform, but also the entire field’s understanding of digital biomarkers in neuroscience R&D.

Koneksa’s extensive commitment to research includes the potential for industry participation in studies. Sponsors who participate in the LEARNS Access Program will have input on final protocol and cohort design, as well as early access to raw data, analytical tools, and results.

About Koneksa

Koneksa is a healthcare technology company pioneering evidence-based digital biomarkers to accelerate clinical research and guide decision-making in drug development and market strategy. Our evidence-based biomarker solutions enable efficient clinical trial designs to help innovative therapies reach patients faster. Koneksa aims to revolutionize treatment effect detection in clinical research and improve patient outcomes. Learn more at koneksahealth.com.

About Beacon Biosignals

Beacon’s machine learning platform for EEG enables and accelerates new treatments that transform the lives of patients with neurological, psychiatric or sleep disorders. Through novel machine learning algorithms, large clinical datasets, and advances in software engineering, Beacon Biosignals empowers biopharma companies with unparalleled tools for efficacy monitoring, patient stratification, and clinical trial endpoints from brain data. For more information, visit https://beacon.bio/. For partnership inquiries, visit https://beacon.bio/contact. Follow us on Twitter (@Biosignals) or LinkedIn.

Contacts

Media
Kimberly Ha
KKH Advisors
917-291-5744
kimberly.ha@kkhadvisors.com

One of the key challenges in ILD management is early detection. Longitudinal studies have shown that incidentally detected subclinical interstitial lung abnormalities (ILAs) on HRCTs are associated with an increased risk of developing idiopathic pulmonary fibrosis (IPF). However, ILAs are also common and occur in up to 9% of lung cancer patients. Not all ILAs represent clinically meaningful diseases, making it challenging to reliably stratify them.

Accurate prognostication using baseline data is also essential in ILD management. Although diagnostic guidelines for several fibrotic lung diseases are available, ILD’s natural history is highly variable among individual patients. Some patients show progressive disease behavior despite conventional therapy, and waiting for patients to show a period of progression presents a major challenge in translating this concept to clinical care.

Effective monitoring of disease response to therapy through HRCT also proves to be a challenge. Visual evaluation of the disease can have high inter-reader variability and low sensitivity to changes in disease severity over short follow-up periods.

Addressing the challenges using machine learning algorithms

To address these challenges, machine learning algorithms can be implemented to identify ILD in at-risk populations, predict the extent of lung fibrosis, and correlate radiological abnormalities with lung function decline. Machine learning can also be used as an endpoint in treatment trials. Advances in image processing and analysis provide further opportunities to use machine learning that incorporates deep-learning-based image analysis and radiomics.

QTC and ML Methods

The authors discuss two methods for analyzing lung medical images: quantitative computed tomography (QCT) and Machine Learning (ML).

QCT uses algorithms to perform lung segmentation, image preprocessing, and image registration to measure variable attenuation corresponding to structural abnormalities.
On the other hand, ML performs feature extraction on digital images to produce numerical descriptors of texture, shape, and other distinctive characteristics that can be computationally analyzed to evaluate potential associations with clinical parameters. Machine Learning can be applied to image features extracted from CTs to evaluate diffuse lung disease and ILD.
The researchers further discuss the use of Deep Learning (DL) in feature-vector-based classification or regression, which can be linked together into composite feature vectors. The two methods are complementary and can provide meaningful prognostic signals, although their use is relatively unexplored and needs further research.

While ML and DL have the potential to identify imaging biomarkers not previously seen by the human eye, they can also lead to difficulties in interpreting abstract features internal to algorithms, which could result in these features and algorithms behaving like black boxes.
Validation is essential to ensure technical performance and clinical significance and minimize bias. Integration and application of ML and DL algorithms to clinical practice should also be studied.
The researchers suggest that ML and DL could play a crucial role in addressing diagnostic, prognostic, and therapeutic clinical research gaps in ILDs. Specifically, it could aid in the stratification of ILAs and established ILD and provide more sensitive tools to identify therapeutic responses in clinical trials.

Conclusions

Machine Learning can be a valuable tool in ILD management. By using ML algorithms to develop more precise and specific digital biomarkers, it can help address ILD diagnosis, prognostication, and therapy response challenges.
However, collaboration and consistency are required to develop optimal algorithms, and candidate radiological biomarkers should be validated against appropriate predictors of disease outcomes. Clinical and imaging data, as well as biological material such as bronchoalveolar lavage, cryobiopsy, or surgical lung biopsy, are essential in evaluating ILD.

Previous studies have shown that clinical features and initial laboratory values cannot predict which infants will develop seizures in the first days after birth. Models have attempted to predict neonatal seizures by combining clinical and early EEG data but with limited sensitivity and specificity. Additionally, these models rely primarily on retrospective, manual review of charts and EEGs to extract the key data to input into the model. This is time-consuming and might not be possible in clinical practice.

McKee and colleagues harnessed the power of their electronic medical record (EMR)-embedded EEG documentation system, which has reported more than 42,000 EEGs since 2018, including those from 1,117 neonates included in their study, 150 of whom had hypoxic ischaemic encephalopathy. The authors show that compliance with this documentation system has steadily increased since its introduction to more than 98%. This large trove of standardized EEG reports provides the basis for seizure-prediction model development.

The authors began with traditional logistic regression models, which showed seizure prediction accuracy of 84% (95% CI 78–89) in the overall cohort. However, machine learning algorithms outperformed logistic regression, achieving seizure prediction accuracy of up to 90% (95% CI 83–94), with recall (sensitivity) of up to 97% (91–100) in a random forest model. The model also performed well in the subset of neonates with hypoxic ischaemic encephalopathy, with an accuracy of 97% (88–99) and recall of 100% (100–100).

This model applies machine learning algorithms to widely available clinical data – in this case, CEEG reports. The authors show that machine learning algorithms might be well-suited to process the density of large EMR datasets. Recent studies have highlighted the potential for machine learning algorithms to detect seizures in real-time and predict seizures in infants with hypoxic ischaemic encephalopathy. McKee and colleagues expand this technology further, applying it not just to infants with hypoxic ischaemic encephalopathy or to direct analysis of digital EEG recordings. Instead, they apply it to broader groups of neonates at risk of seizures and use only summary CEEG reports from the EMR.

This work supports the promise of machine learning for neonatal seizure prediction, although much work must be done before clinical applications are possible. The authors’ institutional EMR-embedded EEG reporting system is valuable for their model but also has limitations. Because it is an institution-specific system, there are no comparable datasets from other centers to validate this model. Similarly, this limits clinicians’ ability to implement their model at other institutions without similar EMR-embedded EEG documentation systems. It is not known whether neonatal CEEG reports across other institutions have the same standardization and predictive ability as those of this single institution. In a gesture toward generalizability, the authors provide an online calculator for seizure prediction based on their model. This calculator allows clinicians to input a patient’s CEEG characteristics to view the classifier decision representing the estimated seizure risk. However, it is not shown whether this is valid for CEEG reports outside of their standardized institutional template.

We all know the progress and speed with which Artificial Intelligence (AI) is entering different sectors. Thanks especially to deep neural networks such as Machine Learning and Deep Learning, algorithms to support healthcare practice, i.e., public health, are being commercialised.

Despite many steps forward, only a few AI-based tools have actually been implemented in the various healthcare systems.

Lino Mari, Head of Technology at Healthware International, looked at the main reasons for this – transparency, data quality, and the ability of doctors and patients to rely on algorithms.

The quality of the data source and thus of the data itself is one of the main concerns of those in healthcare environments dealing with technology. Indeed, it is not always possible to establish the quality of the data and access to the algorithm source code.

In addition, there are still not enough published studies to directly support algorithms that have been tested in silico and may not reflect clinical practice.

How do you establish data quality? In the tech world it is often said: garbage in, garbage out.

Notably, the tech giants have several activities in their pipeline related to health technology innovation. Let’s discover them in details.

Amazon

AmazonDX.com has been launched to the general public and several deals have been signed between Amazon Care and several enterprises.
On the investment side, through the alexa fund they invested $4.4 million in Pixieray and launched the AWS Healthcare Accelerator.
And that’s not all. Amazon has signed a 5-year partnership with the National Safety Council to establish new research, technologies and processes to combat musculoskeletal disorders, one of the leading causes of workplace injuries.

Apple

Apple has announced several updates to its Health app (as part of its upcoming iOS 15 software update) that will allow patients to analyse personal health data trends and share their self-generated health data with doctors, family and friends. They  also launched new Fitness+ programmes for pregnant women and the elderly. Rumours are persistent about the Cupertino company’s ambition to launch a digital primary care service.
There are several established partnerships for sharing medical records data with companies such as Cerner, Allscripts, athenahealth and CPSI. A recently formed partnership with Rockley Photonics, a biosensor company, has reignited other rumours about non-invasive blood glucose and blood pressure monitoring for a future Apple Watch.

Facebook

Rumours about Facebook’s potential smartwatch suggest it could include a heart rate monitor. Facebook has also added a number of new emotional health resources to its platform’s Mental Health Resource Center.
Numerous partnerships were signed with various technology, health and academic institutions to establish the Alliance for Advancing Health Online, which aims to improve public understanding of how social media and behavioural science can be used to help community health. Facebook has also partnered with the government of India to launch a vaccine search tool through its app.

Google

Derm Assist is the first medical device launched by Google, powered by artificial intelligence that helps dermatologists in diagnosing skin conditions. Details have also emerged on Google’s progress on a health record initiative aimed at consumers.
Through the deal with HCA Healthcare, Google wants to develop algorithms that could help improve operational efficiency, monitor patients and guide doctors’ decisions.
GV (Google Ventures) has invested in Kindbody, Brightline, Pill Club, Overture Life, Dyno Therapeutics, Headway, Affinia Therapeutics, Treeline Biosciences, Adagio Therapeutics, Tend, and Ventus Therapeutics.