Deep Learning

New Artificial Intelligence Platform Uses Deep Learning to Diagnose Dystonia with High Accuracy in Less Than One Second

Researchers at Mass Eye and Ear have developed a unique diagnostic tool called DystoniaNet that uses artificial intelligence to detect dystonia from MRI scans in 0.36 seconds. DystoniaNet is the first technology of its kind to provide an objective diagnosis of the disorder. In a new study of 612 brain MRI scans, the platform diagnosed dystonia with 98.8 percent accuracy.

Calibrated approach to AI and deep learning models could more reliably diagnose and treat disease

In a recent preprint (available through Cornell University’s open access website arXiv), a team led by a Lawrence Livermore National Laboratory computer scientist proposes a novel deep learning approach aimed at improving the reliability of classifier models designed for predicting disease types from diagnostic images, with an additional goal of enabling interpretability by a medical expert without sacrificing accuracy. The approach uses a concept called confidence calibration, which systematically adjusts the model’s predictions to match the human expert’s expectations in the real world.

Identifying the Dark Matter of the Molecular World

Scientists have deployed artificial intelligence to identify more of the billions of metabolites that are currently unknown. The small molecules underlie and inform every aspect of our lives, including energy production, the fate of the planet, and our health. “Beast Mode” helps explain how they did it.

Computer Scientist Develops the Art of Artificial Intelligence

Dr. Kang Zhang uses artificial intelligence (AI) to teach computers to create illustrations in the style of the famous masters: Jackson Pollock and his paint splatters or Joan Miró and his curved shapes and sharp lines. The process involves feeding computers examples of colors, abstract shapes and layouts so they can learn to produce their own versions of masterpieces.

ORNL researchers develop ‘multitasking’ AI tool to extract cancer data in record time

To better leverage cancer data for research, scientists at ORNL are developing an artificial intelligence (AI)-based natural language processing tool to improve information extraction from textual pathology reports. In a first for cancer pathology reports, the team developed a multitask convolutional neural network (CNN)—a deep learning model that learns to perform tasks, such as identifying key words in a body of text, by processing language as a two-dimensional numerical dataset.

Predicting chaos using aerosols and AI

Using aerosols as ground truth, researchers at the McKelvey School of Engineering at Washington University in St. Louis have developed a deep learning method that accurately simulates chaotic trajectories — from the spread of poisonous gas to the path of foraging animals.

LLNL computer scientists explore deep learning to improve efficiency of ride-hailing and autonomous electric vehicles

Computer scientists at Lawrence Livermore National Laboratory are preparing the future of commuter traffic by applying Deep Reinforcement Learning — the same kind of goal-driven algorithms that have defeated video game experts and world champions in the strategy game Go — to determine the most efficient strategy for charging and driving electric vehicles used for ride-sharing services.

NUS deep-learning AI system puts Singapore on global map of big data analytics

⎯ A team of researchers from the National University of Singapore (NUS) has put Singapore on the global map of Artificial Intelligence (AI) and big data analytics. Their open-source project, called Apache SINGA, “graduated” from the Apache Incubator on 16 October 2019 and is now Southeast Asia’s first Top-Level Project (TLP) under the Apache Software Foundation, the world’s largest open-source software community.