otolaryngology


These AI-Powered Digital Health Devices Debut At CES 2020

#artificialintelligence

In a 2018 Accenture Digital Health report, 75 percent of respondents said technology played an essential role in managing their health. When it comes to artificial intelligence (AI) powered digital health and wearable devices, 72 percent said they're willing to share their wearable data with their health insurance plan. The report also found that when AI and robotics consumer interest surpassed the choices available today for virtual care. At the Consumer Electronics Show (CES) 2020 in Las Vegas, January 7-10, 2020, AI-powered digital health devices will be prevalent. For people with hearing loss or who are visually impaired, machine learning in digital health devices can open new possibilities to hear conversations more clearly or see the world around them.


These AI-Powered Digital Health Devices Debut At CES 2020

#artificialintelligence

In a 2018 Accenture Digital Health report, 75 percent of respondents said technology played an essential role in managing their health. When it comes to artificial intelligence (AI) powered digital health and wearable devices, 72 percent said they're willing to share their wearable data with their health insurance plan. The report also found that when AI and robotics consumer interest surpassed the choices available today for virtual care. At the Consumer Electronics Show (CES) 2020 in Las Vegas, January 7-10, 2020, AI-powered digital health devices will be prevalent. For people with hearing loss or who are visually impaired, machine learning in digital health devices can open new possibilities to hear conversations more clearly or see the world around them.


Researchers develop AI that reads lips from video footage

#artificialintelligence

But even state-of-the-art systems struggle to overcome ambiguities in lip movements, preventing their performance from surpassing that of audio-based speech recognition. In pursuit of a more performant system, researchers at Alibaba, Zhejiang University, and the Stevens Institute of Technology devised a method dubbed Lip by Speech (LIBS), which uses features extracted from speech recognizers to serve as complementary clues. They say it manages industry-leading accuracy on two benchmarks, besting the baseline by a margin of 7.66% and 2.75% in character error rate. LIBS and other solutions like it could help those hard of hearing to follow videos that lack subtitles. It's estimated that 466 million people in the world suffer from disabling hearing loss, or about 5% of the world's population.


Making the Subjective Objective: Machine Learning and Rhinoplasty

#artificialintelligence

A total of 100 patients ultimately met full inclusion criteria. The average post-surgical follow up for this cohort was 29 weeks (median, 14 weeks; range, 12-64 weeks). Patients ranged from 16 to 72 years old (mean, 32.75 years; median, 28.00 years; standard deviation, 12.79 years). The ranking CNN algorithm on average estimated patients preoperatively to be 0.03 years older than their actual age. The correlation coefficient between actual age and predicted preoperative age was r 0.91.


Deep Compressed Pneumonia Detection for Low-Power Embedded Devices

arXiv.org Machine Learning

Deep neural networks (DNNs) have been expanded into medical fields and triggered the revolution of some medical applications by extracting complex features and achieving high accuracy and performance, etc. On the contrast, the large-scale network brings high requirements of both memory storage and computation resource, especially for portable medical devices and other embedded systems. In this work, we first train a DNN for pneumonia detection using the dataset provided by RSNA Pneumonia Detection Challenge [4]. To overcome hardware limitation for implementing large-scale networks, we develop a systematic structured weight pruning method with filter sparsity, column sparsity and combined sparsity. Experiments show that we can achieve up to 36x compression ratio compared to the original model with 106 layers, while maintaining no accuracy degradation. We evaluate the proposed methods on an embedded low-power device, Jetson TX2, and achieve low power usage and high energy efficiency. Keywords: Pneumonia detection · YOLO · structured weight pruning.


AI takes 10 seconds to diagnose pneumonia on chest x-rays

#artificialintelligence

The system--CheXpert--was developed using 188,000 chest images by researchers at Stanford University and fine-tuned by a team at Intermountain Medical Center to identify suspected pneumonia. If implemented in an emergency department, physicians may be able to treat patients sooner--vital to those suffering from pneumonia.


Livio AI: In Conversation with Achin Bhowmik

#artificialintelligence

Achin Bhowmik discusses how Starkey's Livio AI came to market and what it means for the future of amplification devices. All my life, I've been passionate about developing perceptual computing technologies, such as sensors and artificial intelligence. My focus at Intel was to use these technologies to make more intelligent machines. That was an incredible time in my career as the world is getting smarter and there is so much to explore and invent. But Starkey CEO, Mr Austin came to me and asked, "Do you want to use the same advanced technologies, but instead of focusing on making more intelligent machines, help people perceive and understand the world better?"


Machine learning improves the diagnosis of patients with head and neck cancers

#artificialintelligence

Dr. Klaus-Robert Müller, Professor for Machine Learning at TU Berlin, the research group employed artificial intelligence-based methods to render this …


Health of the future is taking shape: AI Health from Berlin

#artificialintelligence

The innovative power of the German capital in the area of AI is not only noticeable in the high-profile areas of business intelligence and process management, but is also demonstrated by the excellent work of the AI companies which deal intensively with intelligent health and represent about 10 per cent of the Berlin AI ecosystem. AI systems from Berlin are used in a variety of ways: they help in the diagnosis and data analysis of specific disease patterns, but are also used in operation planning and in supporting the internal processes of hospitals. Apps for intelligent data recording and analysis in the field of prevention are being developed in the context of fitness and health. Chatbots, i.e. systems with which people can communicate in natural language, also accompany patients during the healing process. A number of start-ups in Berlin are pushing the boundaries of traditional healthcare with innovative solutions which could also break new ground on the international stage - always at the interface between business and research.


Meniere's Disease Prognosis by Learning from Transient-Evoked Otoacoustic Emission Signals

arXiv.org Machine Learning

Accurate prognosis of Meniere disease (MD) is difficult. The aim of this study is to treat it as a machine-learning problem through the analysis of transient-evoked (TE) otoacoustic emission (OAE) data obtained from MD patients. Thirty-three patients who received treatment were recruited, and their distortion-product (DP) OAE, TEOAE, as well as pure-tone audiograms were taken longitudinally up to 6 months after being diagnosed with MD. By hindsight, the patients were separated into two groups: those whose outer hair cell (OHC) functions eventually recovered, and those that did not. TEOAE signals between 2.5-20 ms were dimension-reduced via principal component analysis, and binary classification was performed via the support vector machine. Through cross-validation, we demonstrate that the accuracy of prognosis can reach >80% based on data obtained at the first visit. Further analysis also shows that the TEOAE group delay at 1k and 2k Hz tend to be longer for the group of ears that eventually recovered their OHC functions. The group delay can further be compared between the MD-affected ear and the opposite ear. The present results suggest that TEOAE signals provide abundant information for the prognosis of MD and the information could be extracted by applying machine-learning techniques.