Train personalized machine learning models without collecting sensitive user data

DynamoFL | Vaikkunth Mugunthan


Poster presentation about DynamoFL


DynamoFL enables machine learning teams to build personalized ML solutions without collecting sensitive user data. Our users leverage DynamoFL to train models using federated learning. DynamoFL slashes data transfer costs and boosts performance all while preserving user privacy.

Learn more about DynamoFL | View Poster PDF | Watch Video

ViTrack: Continious Hemodynamic Monitor

Dynocardia | Greg Boverman, Gokul Prasath Rajamanickam, & José Wong

Poster presentation by Dynocardia


Dynocardia: Dynocardia (the Company), the developer of ViTrackTM, a proprietary platform technology, seeks to validate, manufacture, license, and market wearable solutions for applications across patient and wellness monitoring markets globally. The Company has received $6.5 million in funding, which includes grants from the National Heart, Lung, and Blood Institute, the National Institute of Biomedical Imaging and Bioengineering, and the National Science Foundation. In addition, the Company has received funding and accelerator benefits from two prestigious accelerators, MassTech and Luminate, sponsored by the States of Massachusetts and New York, respectively. The Company is laying the groundwork for an initial market entry into the hospital critical care market sub-segment. It will eventually deploy the technology into various markets, including the consumer market. Over time, by integrating data across care settings, Dynocardia will produce unique predictive insights benefitting the entire healthcare ecosystem and creating monumental scale for the company.

Urgent, unmet need for accurate, continuous, and ubiquitous BP measurements: The current standard of arm cuff-based BP measurements is inaccurate and provides only single-point BP measurements, which leads to morbidity and mortality in hospitals. Because of this, 50% of the critically ill patients in high acuity centers receive invasive intra-arterial pressure (IAP) monitoring. Apart from hospital settings, accurate and continuous measurements at home and other real- world settings will significantly improve outcomes in the 1.5 billion people globally with high BP or hypertension, a leading cause of stroke and heart attacks. Other conditions that will benefit from good BP management are sleep apnea, heart and renal failure, and dementia.

The ViTrackTM Technology: To address this clinical need, in 2015, the National Institutes of Health called for a fundamentally new approach for BP measurement and selected and funded the joint Massachusetts Institute of Technology and Tufts University School of Medicine founding team to develop ViTrackTM. In April 2020, Dynocardia executed a license with Tufts University and MIT for exclusive worldwide rights to the technology. ViTrack will address a >100-year-old challenge and an unmet need for non-intrusive, accurate, continuous BP measurements. This platform technology, powered by proprietary optomechanical sensor technology and method, also provides continuous and accurate heart rate, respiratory rate, and other cardio-hemodynamic parameters. Preclinical and clinical studies have demonstrated ViTrack’s ability to measure BP accurately and continuously as per FDA standards. The Company is currently conducting clinical studies for product validation in world-renowned hospitals, including Massachusetts General Hospital and UMASS Memorial, and with planned expansions to Cleveland Clinic, NYC Health + Hospitals, Baylor Medical Center, and others.

Learn more about Dynocardia | View Video | View Poster PDF | Watch Video

Leela AI: Creating resilient intelligences

Leela AI | Steve Kommrusch

LeelaAi poster presentation


Leela AI’s proprietary technology is based on research done at the MIT AI Lab. It combines self-motivated knowledge acquisition with deep learning to deliver causal understanding, creating resilient AI. Built on Leela AI’s technology, is a uniquely reconfigurable tool. It can digitize highly variable motion and activity, connecting cause-and- effect to model custom events.

Learn more about Leela AI | View Poster PDF | Watch Video 

BLTN for Public Safety - Save Time, Collaborate, Solve Crimes

Multitude Insights | Matt White


At Multitude Insights, we use advanced natural language processing techniques and AI-enabled search to augment detectives and crime analysts’ workflows by saving them critical time and energy, enhancing inter-city collaboration, and easing criminal information sharing.

Key Features:

  • Increase case closure by intelligently detecting collaboration opportunities among neighboring cities’ police departments.
  • Save 30 minutes to 2 hours a day using our sematic-search technology that rapidly retrieves unstructured data to aid in investigations.
  • Expand the aperture – work with a wide variety of public safety stakeholders like schools, utili- ties, and corporate security teams on our collaborative intelligence network.

Factor models are routinely used in the financial industry to identify and quantify sources of systematic risk to manage the risk of a portfolio of securities and hedge investment positions, or in valuation contexts to estimate the cost of capital of an asset. In this paper, researchers construct latent factor models out of 150 well known factors from the asset pricing literature using autoencoders to explain away most of the anomalies in the cross section of asset prices. They show that their models outperform the classical Fama-French and Hou-Xue-Zhang factor models on a large set of test assets.

Learn more about Multitude Insights | Watch Video