“I hear you are sick” is about to take on a whole new meaning. AI is adding a new level of functionality to bioacoustics, the combination of biology and acoustics, to gather health insights from sounds produced by humans.
“From cough to speech and even breath, the sounds our bodies make are filled with information about our health,” Google Research Director Of Engineering Shravya Shetty said in an August 19 blog posting. “Subtle clues hidden within these bioacoustic sounds hold the potential to revolutionize how we screen, diagnose, monitor and manage a wide range of health conditions like tuberculosis (TB) or chronic obstructive pulmonary disease (COPD).”
Earlier this year Google introduced Health Acoustic Representations, or HeAR, a bioacoustic foundation model designed to help researchers build models that can listen to human sounds and flag early signs of disease. The Google Research team trained HeAR on 300 million pieces of audio data curated from a diverse and de-identified dataset, and trained the cough model in particular, using roughly 100 million cough sounds.
In August Google announced that it is working with India’s Salcit Technologies, a Hyderabad-based healthcare company that uses AI to analyze cough sounds and assess lung health. The Indian company is exploring how HeAR can help expand the capabilities of its bioacoustic AI models. To start, it is using HeAR to help research and enhance their early detection of TB based on cough sounds.
Indian health care providers like Apollo Hospitals and the nonprofit Healing Fields Foundation are using Salciet’s Swaasa application to screen people, including in remote areas. Salcit has India’s medical device regulator’s approval, a first for a software tool to be deployed as a medical device. In its mobile app, Swaasa allows users to upload a 10-second cough sample by coughing near their mobile phone and test for diseases with a 94% accuracy, Salcit’s co-founder Manmohan Jain told Bloomberg. The cough sound is the equivalent of giving a blood sample. The advantage is the audio sample is processed on the Cloud rather than in a laboratory, significantly reducing the cost and ensuring access to more people.
Google said it is also seeing support for HeAR from The StopTB Partnership, a United Nations-hosted organization that brings together TB experts and affected communities with the goal of ending TB by 2030.
On average, HeAR ranks higher than other models on a wide range of tasks and for generalizing across microphones, demonstrating its superior ability to capture meaningful patterns in health-related acoustic data, according to Google. Models trained using HeAR also achieved high performance with less training data, it said.
In another bioacoustics venture, Google is researching a model based on ultrasound for early breast cancer detection at the Chang Gung Memorial Hospital in Taiwan. The AI assists in lesion detection and Google aims to roll out it out globally, offering free breast cancer screening for populations that can’t afford costly mammograms.
Neither of Google’s models is yet near commercialization. But sound-based generative AI systems could democratize early disease detection, making screening accessible, affordable and scalable.
Montreal-based Ubenwa has built a foundation model for infant cries to help with the early detection of medical conditions in newborns. “The long-term vision is to integrate our technology with monitoring tools at hospitals, clinics and home once we get our FDA approval from the Ministry of Health in Canada,” Khaled Said, Ubenwa’s vice-president for product told the Montreal City News in May.
Others are working on AI tools that can detect autism from voice patterns. A 2023 article in Nature magazine showed that voice acoustics classified autism diagnosis with an overall accuracy of 91%.
A June article in the National Library of Medicine, which is run by the NIH, a part of the U.S. Department of Health and Human Services, cites 27 startups with a focus on using AI for developing biomarkers of health from the human voice.
IN OTHER NEWS IN AUGUST AND EARLY SEPTEMBER:
ARTIFICIAL INTELLIGENCE
U.S., Britain and Brussels Sign Agreement On AI Standards
Three major western jurisdictions building technologies for artificial intelligence have signed the first international treaty on the use of AI that is legally binding. The U.S., EU and UK signed the Council of Europe’s convention on AI on September 5, which emphasizes human rights and democratic values in its approach to the regulation of public and private-sector systems. Other countries are expected to sign. The convention was drafted over two years by more than 50 countries. It requires signatories to be accountable for any harmful and discriminatory outcomes of AI systems. It also requires that outputs of such systems respect equality and privacy rights, and that victims of AI-related rights violations have legal recourse.
Voice : Generative AI’s Next Wave ?
AI’s coupling with voice was in the news in August:
*In an experiment that surpassed expectations, implants in an ALS patient’s brain were able to recognize words he tried to speak, and AI helped produce sounds that came close to matching the patient’s true voice.
*Executives from Alphabet’s Google DeepMind, Microsoft. and Meta joined tech founders in Bangalore to watch one of India’s top AI startups unveil a new product that might change how the world’s most populous country uses the technology. Bloomberg reported that Sarvam AI, often described as India’s OpenAI, introduced software for businesses that can interact with customers using spoken voice rather than just text. The technology was developed with data from 10 native Indian languages and priced at a rupee per minute to capture the market. In a video at the event, Vinod Khosla, a billionaire venture capitalist and investor in Sarvam, said, “These voice bots have the potential to reach a billion people.”By incorporating AI voice features, tech companies hope to create more dynamic, conversational services that can respond to users verbally in real time and automate certain tasks. In India, that’s already playing out across a wide range of consumer and business applications.
*Amazon said it is planning to introduce a new version of Alexa due for release in October ahead of the U.S. holiday season, which will be powered primarily by Anthropic’s Claude artificial intelligence models and will carry out more complicated requests, such as ordering food or drafting emails all from a single prompt.
*Pindrop, an Atlanta, Georgia-based startup released a new product over the summer that it says can detect AI-generated speech in both phone calls and digital media, to help combat fraud.
FOOD AND BEVERAGE
Solar Foods To Go Public As It Readies To Enter U.S. Market With Food Made From Thin Air
Finnish food tech startup Solar Foods, which makes protein from air, is expected to go public on the Nasdaq’s First North Growth market in Finland on Sept. 10. The company, which The Innovator first featured as a startup of the week five years ago, makes Solein a nutritionally rich and versatile ingredient which can replace protein virtually in any food. It can also be used as a fortifier to complement the nutritional profile of various foods: it can be a source of iron, fiber and B vitamins, and it can also bring different techno-functionalities into food products.Solein received a novel food regulatory approval that allows for the sale of food products containing Solein in Singapore in September 2022. It has since provided the novel ingredient for limited-edition food products and test marketing in Singapore, such as a Solein-powered snack bar and a Solein chocolate ice. It has secured GRAS (Generally Recognized as Safe) status in the U.S, an achievement which Solar Foods said it sees as a major step towards Solein’s commercialization and entry onto the U.S. market. The next steps in Solein’s commercialization in the U.S. are to register the Solar Foods production facility Factory 01 in Finland with the U.S. Food and Drug Administration (FDA), which requires a food safety plan that fulfills specific requirements. This step will enable Solar Foods to start food export from Factory 01 in Finland to the United States. Solar Foods estimates sales could start by the end of this year.
ENERGY
Offshore Use of Nuclear Energy Could Be An Option
Saipem, which focuses on the engineering and construction of large projects in the energy and infrastructure sectors and European nuclear energy startup newcleo have signed a collaboration agreement to identify solutions for the offshore application of newcleo’s technology. The objective is to study the application of newcleo’s “Small Modular Lead-cooled Fast Reactor” (SM-LFR) technology to provide zero-emission electricity and process heat to oil and gas offshore installations to improve their sustainability performance. The agreement also allows for the possibility of extending the use of newcleo’s technology to produce zero-emission electricity through floating nuclear units, connected to the electricity grid on land or to other users. newcleo’s technology leverages passive security systems (i.e. exploiting natural forces or phenomena without requiring active mechanisms), which it says enables much greater efficiency in the use of extracted uranium compared to other types of conventional fission reactors. The company reuses spent nuclear fuel used by other reactors, in line with the principles of the circular economy.
ANTITRUST
Nvidia and Google Are The Latest AI Tech Giants Under Fire For Alleged Antitrust Violations
Nvidia’s value was slashed by $279 billion to $2.6 trillion, marking the largest one-day drop in history for a U.S. company, after news broke that the U.S. Justice Department is escalating its investigation into whether the company has made it harder for clients to switch to other semiconductor suppliers and is penalizing buyers that refuse to exclusively use Nvidia’s AI chips. Nvidia it has not been formally subpoenaed. The U.S. government is investigating antitrust charges against multiple AI tech giants. In August Google was found to have violated antitrust law by illegally maintaining a monopoly in Internet search. Justice Department officials are considering what remedies to ask a federal judge to order against the search giant. They are discussing various proposals, including breaking off parts of Google, such as its Chrome browser or Android smartphone operating system forcing Google to make its data available to rivals, or mandating that it abandon deals that made its search engine the default option on devices like the iPhone, sources told The New York Times. Meanwhile, Microsoft’s $10 billion investment and ongoing relationship with OpenAI, the $80 billion startup behind ChatGPT, is also under investigation amidst fears the tech giant may be able to “exert undue influence or gain privileged access in ways that could undermine fair competition” in the AI arms race.
DISINFORMATION
Brazil Takes On Disinformation and Elon Musk
X began to go dark across Brazil on August 31 after the nation’s Supreme Court blocked the social network because its owner, Elon Musk, refused to comply with Judge Alexandre de Moraes orders to suspend certain accounts that the government said were spreading disinformation and hate speech. “The battle of the titans, between de Moraes and Musk, reminds us of how powerful, political and provocative tech leaders have become,” Marietje Schaake, the international policy director at Stanford University’s Cyber Policy Center, told The Guardian. “Brazil won’t be that last country to seek accountability or to put up guardrails.”
To read more of The Innovator’s News In Context articles click here.
