A Canadian startup has made the process of getting a COVID-19 vaccine much easier with a robot that injects a shot directly into the muscle without the use of a needle. Developers of the Cobi robot, designed at the University of Waterloo in Ontario, say the droid has successfully performed the first autonomous robotic intramuscular injection. Cobi relies on a high-pressure jet of serum that passes through an opening in the skin the width of a human hair. Using LiDAR sensors, the same technology employed by autonomous vehicles to map the road, Cobi makes a model of the patient's body and AI-based software determines the optimal site for injection. 'We outfitted Cobi to use a needle-free injection technology and to demonstrate that patients could receive intramuscular injections, such as vaccines, without needles and no involvement from a healthcare professional,' said Tim Lasswell, co-founder and CEO of Cobionix in a statement.

Thirty years ago this month, Europe's most famous mummy was discovered lying face-down in the ice, on the edge of a lake nearly two miles high in the Ötztal Alps bordering Austria and Italy. Naturally preserved by more than 5,000 years of sun, wind, and freezing temperatures, the leathery remains of Ötzi the Iceman quickly became a global sensation, the subject of countless books and documentaries and even a feature film reconstructing his life in Neolithic Europe and his violent death. Today, Ötzi is carefully tended to by researchers at the South Tyrol Museum of Archaeology in Bolzano, Italy, where his wizened body is kept in a custom cold chamber maintained at a constant temperature of –21.2 degrees Fahrenheit. Four or five times a year, his remains are sprayed with sterile water to create an icy, protective exoskeleton that ensures he stays a "wet mummy" (one naturally preserved in a wet rather than dry environment). Go behind the scenes of Ötzi's 2010 autopsy.

Theoretical analyses for graph learning methods often assume a complete observation of the input graph. Such an assumption might not be useful for handling any-size graphs due to the scalability issues in practice. In this work, we develop a theoretical framework for graph classification problems in the partial observation setting (i.e., subgraph samplings). Equipped with insights from graph limit theory, we propose a new graph classification model that works on a randomly sampled subgraph and a novel topology to characterize the representability of the model. Our theoretical framework contributes a theoretical validation of mini-batch learning on graphs and leads to new learning-theoretic results on generalization bounds as well as size-generalizability without assumptions on the input.

An automated feature selection pipeline was developed using several state-of-the-art feature selection techniques to select optimal features for Differentiating Patterns of Care (DPOC). The pipeline included three types of feature selection techniques; Filters, Wrappers and Embedded methods to select the top K features. Five different datasets with binary dependent variables were used and their different top K optimal features selected. The selected features were tested in the existing multi-dimensional subset scanning (MDSS) where the most anomalous subpopulations, most anomalous subsets, propensity scores, and effect of measures were recorded to test their performance. This performance was compared with four similar metrics gained after using all covariates in the dataset in the MDSS pipeline. We found out that despite the different feature selection techniques used, the data distribution is key to note when determining the technique to use.

However, when Dazed tested it using country names, it described the UK and US as "good", France as "nice", and Russia as "a great place to visit", but said Nigeria, Mexico, and Iraq were "dangerous", while Iran was "bad". Clearly, the software – like much artificial intelligence – has a problem with racism. Its creators have addressed this in a post-launch Q&A, writing: "Today's society is unequal and biased. This is a common issue with AI systems, as many scholars have argued, because AI systems are trained on historical or present data and have no way of shaping the future of society, only humans can. What AI systems like Delphi can do, however, is learn about what is currently wrong, socially unacceptable, or biased, and be used in conjunction with other, more problematic, AI systems (to) help avoid that problematic content."

Products developed to manage artificial intelligence data are still largely fragmented, solving one problem at a time for developers, but not the entire life cycle. Enter Sama, a company providing high-quality training data that powers AI technology applications. CEO Wendy Gonzalez said the company is developing the first end-to-end AI tool for training data through machine learning. To do this, the company secured an oversubscribed $70 million in Series B financing led by Caisse de dépôt et placement du Québec (CDPQ), with participation from First Ascent Ventures, Salesforce Ventures, Vistara Growth and all existing investors. The new capital infusion comes two years after the company raised $14.8 million in a Series A round.

Use of artificial intelligence (AI) by hospitals is now hitting its stride after a long ramp-up as more systems and facilities sink serious dollars and effort into connected healthcare. Funded in part by a $20 million grant from the National Institute on Aging (NIA), MassAITC's stated mission is to advance in-home connected care as "90% of older Americans would prefer to stay in their homes as they age," but with Alzheimer's presenting daunting challenges. "While at-home health care technologies hold significant promise, they have not been specifically developed for older adults or Alzheimer's patients, caregivers and their clinicians. Further, many current treatment and intervention regimes are limited in terms of their ability to be remotely delivered, managed and adapted to patient needs and caregiver abilities." Adapting to consumers' developing digital health preferences is becoming a prime differentiator for providers.

In popular usage, Data Gravity refers to the ability of a body of data to attract applications, services and other data. In this work we introduce a broader concept, "Digital Gravity" which includes not just data, but other elements of the AI/ML workflow. This concept is born out of our recent experiences in developing and deploying an AI-based decision support platform intended for use in a public health context. In addition to data, examples of additional considerations are compute (infrastructure and software), DevSecOps (personnel and practices), algorithms/programs, control planes, middleware (considered separately from programs), and even companies/service providers. We discuss the impact of Digital Gravity on the pathway to adoption and suggest preliminary approaches to conceptualize and mitigate the friction caused by it.

Empirically observed time series in physics, biology, or medicine, are commonly generated by some underlying dynamical system (DS) which is the target of scientific interest. There is an increasing interest to harvest machine learning methods to reconstruct this latent DS in a completely data-driven, unsupervised way. In many areas of science it is common to sample time series observations from many data modalities simultaneously, e.g. electrophysiological and behavioral time series in a typical neuroscience experiment. However, current machine learning tools for reconstructing DSs usually focus on just one data modality. Here we propose a general framework for multi-modal data integration for the purpose of nonlinear DS identification and cross-modal prediction. This framework is based on dynamically interpretable recurrent neural networks as general approximators of nonlinear DSs, coupled to sets of modality-specific decoder models from the class of generalized linear models. Both an expectation-maximization and a variational inference algorithm for model training are advanced and compared. We show on nonlinear DS benchmarks that our algorithms can efficiently compensate for too noisy or missing information in one data channel by exploiting other channels, and demonstrate on experimental neuroscience data how the algorithm learns to link different data domains to the underlying dynamics

Maternal and child mortality is a public health problem that disproportionately affects low- and middle-income countries. Every day, 800 women and 6,700 newborns die from complications related to pregnancy or childbirth. And for every maternal death, about 20 women suffer serious birth injuries. However, nearly all of these deaths and negative health outcomes are preventable. Midwives are key to revert this situation, and thus it is essential to strengthen their capacities and the quality of their education. This is the aim of the Safe Delivery App, a digital job aid and learning tool to enhance the knowledge, confidence and skills of health practitioners. Here, we use the behavioral logs of the App to implement a recommendation system that presents each midwife with suitable contents to continue gaining expertise. We focus on predicting the click-through rate, the probability that a given user will click on a recommended content. We evaluate four deep learning models and show that all of them produce highly accurate predictions.