Featured Story

Regulations

AI Regulation in Europe: A Balancing Act Between Innovation and Ethics

View Sources,2,3,4 

The European Union's proposed legislation on Artificial Intelligence (AI) has sparked a heated debate among tech executives and lawmakers. While the EU aims to ensure a human-centric and ethical development of AI, critics argue that the stringent rules could stifle innovation and impose high compliance costs on firms. AI has been at the forefront of technological innovation, promising to revolutionize various sectors from healthcare to finance.

The EU's Proposed AI Legislation

The EU's proposed AI legislation, which has been endorsed by the Internal Market Committee and the Civil Liberties Committee, aims to ensure that AI systems are overseen by people, safe, transparent, traceable, non-discriminatory, and environmentally friendly. The rules follow a risk-based approach, establishing obligations for providers and users depending on the level of risk the AI can generate. AI systems with an unacceptable level of risk to people’s safety would be strictly prohibited, including systems that deploy subliminal or purposefully manipulative techniques, exploit people’s vulnerabilities, or are used for social scoring.

Criticism from Tech Executives

Despite the EU's intentions, the proposed legislation has been met with criticism from tech executives at over 160 firms, including Meta, OpenAI, and Cellnex. These critics argue that the EU rules could cause high compliance costs and unfair liability risks to firms experimenting with AI technology. They also express concerns that the stringent rules could trigger leading AI researchers to migrate to friendlier jurisdictions to continue innovating with the technology.

The Impact on Innovation and the Tech Industry

The controversy surrounding the EU's proposed AI legislation highlights the delicate balance between fostering innovation and ensuring ethical and safe use of technology. While the EU believes its proposed rules are balanced to encourage a competitive environment for AI firms to thrive, critics argue that the legislation could inadvertently stifle innovation and impose undue burdens on tech companies. This tension underscores the challenges of regulating a rapidly evolving field like AI, where the potential benefits of innovation must be weighed against the risks of misuse and ethical concerns.

As the debate continues, it's clear that the EU's proposed AI legislation will have far-reaching implications for the future of AI in Europe and beyond. Whether these rules will ultimately serve as a blueprint for other regions or a cautionary tale remains to be seen. What is certain, however, is that the conversation around AI regulation is far from over, and the decisions made today will shape the trajectory of AI development for years to come.

Stay informed about the latest developments in AI regulation by subscribing to our newsletter. Share your thoughts and join the conversation by leaving a comment below. Your insights could help shape the future of AI!

#Regulation, #EU

Drug Development

First Generative AI Drug Enters Phase II Trials

View Source

Insilico Medicine, a clinical-stage biotechnology company powered by generative artificial intelligence (AI), has announced the initiation of Phase II clinical trials for INS018_055, the world's first anti-fibrotic small molecule inhibitor discovered and designed using generative AI. The study is a randomized, double-blind, placebo-controlled trial aiming to assess the safety, tolerability, pharmacokinetics, and preliminary efficacy of a 12-week oral dosage of INS018_055 in subjects with Idiopathic Pulmonary Fibrosis (IPF). The company plans to recruit 60 subjects at about 40 sites in the U.S and China. This marks a significant milestone in the application of AI in drug discovery and development, potentially offering a new treatment option for patients worldwide. For more information,

#AI #Healthcare #DrugDiscovery #MachineLearning

Diagnostics

VA researchers working on artificial intelligence that can predict prostate cancer

View Source

The Department of Veterans Affairs researchers are developing an artificial intelligence (#AI) algorithm to predict aggressive prostate cancer. The research study, which began on July 1, is being conducted across 14 sites and will analyze data from over 5,000 veterans diagnosed with high-risk prostate cancer. The AI model will use diagnostic images, high-resolution scans of prostate biopsies, and socioeconomic variables to identify patterns indicative of aggressive prostate cancer. The study is unique in its incorporation of socioeconomic conditions into a predictive model for high-risk prostate cancer. The researchers are also leveraging cloud storage, computing infrastructure, and unique databases, including the Million Veteran Program and the Prostate Cancer Foundation-VA partnership. The infrastructure developed by this research will serve as a hub for future discovery and could be used to develop similar AI algorithms for other cancers and diseases. Prostate cancer is the second most common cancer among men in the United States and makes up 30% of new cancer diagnoses in the Department of Veterans Affairs. #AI #Healthcare #ProstateCancer #VeteransAffairs

AMA Approves CPT III Code for icometrix's AI-Related Brain MRI Software

View Source

The American Medical Association (AMA) has issued a Current Procedural Terminology (CPT) code for icometrix's FDA-cleared, AI-related brain MRI quantification software, creating a path to reimbursement. This milestone makes icometrix, a leading brain imaging AI company, part of the standard of care. The software provides clinically relevant metrics for cerebral MRI scans, aiding clinicians in diagnosing, monitoring, and assessing treatment responses for brain disorders such as multiple sclerosis, Alzheimer's Disease, Epilepsy, Traumatic Brain Injury, Stroke, and more.

The new CPT codes will allow hospitals and imaging centers in the US to submit claims for icometrix's AI-based analysis of brain MRI scans. This recognition is a significant step towards integrating precision medicine into the management of neurological conditions. The demand for brain MRI procedures is increasing due to shifting demographics and aging populations. The use of AI in neuroimaging can help manage the growing burden of neurologic diseases.

#AI #Healthcare #Neurology #PrecisionMedicine #icometrix #BrainMRI #ArtificialIntelligence #MachineLearning 

Startup

Medical Diagnostics with AI-Powered Osteoporosis Solution by Promedius

View Source

Promedius, a Korean startup, is making strides in healthcare with its AI-powered solution for diagnosing osteoporosis. The company's innovative approach allows for the triaging of suspected osteoporosis patients using chest X-ray images, a significant shift from the traditional DXA test. This new method facilitates early diagnosis, as it enables individuals identified as suspicious for osteoporosis to undergo DXA bone density testing sooner.

The company's focus on X-ray imaging aims to bridge the gap in medical resources and improve patient outcomes. To further its mission, Promedius has actively pursued global partnerships with major DXA equipment manufacturers and global pharmaceutical companies specializing in osteoporosis treatments.

The company has secured prior funding of $2 million from prominent investors like The Big Bang Angels and Korea Credit Guarantee Fund and is currently seeking an additional $10 million to fuel its growth and expand its global footprint. Dr. Hyun-Jin Bae, with a Ph.D. background in image processing, data analysis, and big data, is at the core of Promedius.

Gradient Health Secures $2.75M Funding to Expand its AI-Driven Medical Data-Sharing Platform

View Source

Gradient Health, a prominent medical AI data-sharing company, has successfully closed a funding round of $2.75 million, led by ReMY Investors & Consultants. This investment will facilitate the company's mission to assemble the world's largest and most comprehensive annotated medical imaging library on a secure, unified platform. The platform is expected to be a game-changer in healthcare research, providing a vast collection of annotated medical images that will enable researchers to unlock new insights and minimize bias, thereby fueling advancements in diagnostics and therapies.

The platform offers a range of features designed to facilitate compliant research and collaborations, employing security measures to ensure the privacy and compliance of sensitive patient information. It also provides tools for developers, including machine-learning-ready formatting and the ability to search by diseases and imaging modality.

Gradient Health's data-sharing platform has been recognized for its potential to revolutionize healthcare research. The company aims to collaborate with academic institutions, healthcare organizations, and technology partners to collectively advance medical knowledge and improve patient care

Agriculture

Laser-Armed Robot terminates weeds without herbicide

View Source, 2 

The global herbicide market, valued at $33.65 billion in 2020, is set to be disrupted by Carbon Robotics' autonomous weed elimination robot, "Bud". This innovative machine uses artificial intelligence (AI) to distinguish weeds from crops and employs a high-power laser to eradicate the weeds, reducing the need for herbicides and labor, thus enhancing crop yields and saving costs.

"Bud" is equipped with an Nvidia AI processor that collects data from a dozen high-resolution cameras to feed its crop and weed computer vision models. It also carries lighting for night-time operation. The robot features eight independent weed-killing units, each with a 150-watt laser that can fire every 50 milliseconds with 3 mm accuracy, capable of eliminating over 100,000 weeds per hour.

The robot's development faced challenges, including generating sufficient power in a mobile platform and distributing it to all the lasers and cooling systems. Despite these hurdles, the team at Carbon Robotics found the cross-industry collaboration rewarding.

The robot's accuracy in avoiding crops while eliminating weeds is crucial. It uses deep learning to create a furrow detection model, which guides the robot's path without the need for GPS or location waypoints. The robots can operate in fields ranging from 200 acres to tens of thousands of acres, clearing 15-20 acres per day, and replacing several deployments of hand-weeding crews. They have been tested on specialty crop farms, working on fields with a variety of crops.

 Research 

MIT's BioAutoMATED: A Leap Forward in Democratizing AI for Biological Research

View Source

Researchers at MIT's Abdul Latif Jameel Clinic for Medical Engineering and Science have developed BioAutoMATED, an automated machine learning (ML) system designed to generate AI models for biology research. The system, detailed in an open-access paper published in Cell Systems, aims to democratize AI for research labs by eliminating the need for extensive machine-learning expertise.

BioAutoMATED can select and build an appropriate model for a given dataset and even handle the labor-intensive task of data preprocessing. This significantly reduces the time and financial cost associated with recruiting machine-learning researchers and formatting datasets for model development. The system's repertoire of supervised ML models includes binary classification models, multi-class classification models, and regression models.

The team at MIT believes that BioAutoMATED can help lower barriers for biology-centric labs that need to invest in significant digital infrastructure and AI-ML trained human resources before they can even see if their ideas are poised to pan out. The open-source code is publicly available and easy to run, encouraging collaboration and improvement from the larger community.

The development of BioAutoMATED represents a significant step forward in the integration of AI and ML in biological research, potentially revolutionizing the field and making AI more accessible to researchers worldwide.

Prediction of multiple conformational states by combining sequence clustering with AlphaFold2

View Source

In a recent study published on researchers have demonstrated a significant advancement in the capabilities of AlphaFold2 (AF2), a tool that has already revolutionized structural biology with its accurate predictions of single protein structures and protein-protein complexes. The study reveals that AF2, when combined with a method of clustering multiple sequence alignments (MSAs) by sequence similarity, can predict alternate states of known metamorphic proteins with high confidence.

This development in bioinformatic methods, combined with experimental testing, could have a profound impact on predicting protein energy landscapes, essential for understanding biological function. However, it's important to note that these are predictions and further experimental testing is needed to confirm these findings.

The study is a significant step forward in the use of #AI and #MachineLearning in #StructuralBiology, and it opens up new possibilities for understanding protein function and disease-causing point mutations.