Stay ahead of the headlines

Dec 4, 2024·LinkedIn

Amazon has launched the Nova family of models to enhance generative AI capabilities across text, image, and video applications. The six models, available via Amazon Bedrock, are designed to optimize performance for various tasks, with Nova Micro focusing on low-latency text processing, Nova Lite supporting multimodal interactions, and Nova Pro balancing accuracy and efficiency for complex workflows. Upcoming models include the yet-to-be-released Nova Premier for complex reasoning and Nova Canvas, which specializes in advanced image generation and editing features.

Dec 3, 2024·Healthcare IT News

Researchers at the National Institutes of Health (NIH) have introduced TrialGPT, an AI tool aimed at improving the matching of volunteers to clinical trials by analyzing patient summaries for relevant medical and demographic information. With performance comparable to human clinicians, TrialGPT generates a ranked list of suitable clinical trials, allowing clinicians to efficiently discuss options with patients and explain enrollment criteria. In a study, clinicians using the tool were found to spend 40% less time on patient screenings without sacrificing accuracy. This development reflects a growing trend of leveraging AI in clinical trial recruitment, particularly to enhance health equity and trial diversity amid challenges exacerbated by the COVID-19 pandemic.

Dec 3, 2024·Radiology Business

In his address at the Radiological Society of North America's 110th Scientific Assembly, President Curtis P. Langlotz, MD, PhD, discussed the key components needed to advance artificial intelligence (AI) in medical imaging, including: eliminating the use of discs, promoting interoperability between systems, enhancing radiologist training in AI technologies, fostering collaboration among healthcare teams, and advocating for clear regulatory frameworks governing AI in radiology. By adopting these strategies, Langlotz believes the integration of AI in medical imaging can lead to improved patient outcomes and more efficient radiological practices.

Dec 3, 2024·Menlo VC

In 2024, spending on generative AI by enterprises has surged to $13.8 billion, marking a substantial shift from experimentation to execution, with 72% of decision-makers anticipating broader adoption of the technology. Despite this optimism, over a third of leaders lack a clear strategy for implementation. Funding trends show that 60% of investments are sourced from innovation budgets, with 40% from permanent budgets, indicating a long-term commitment to AI transformation. The demand for applications has risen dramatically, with $4.6 billion allocated for generative AI applications, as organizations explore an average of ten use cases, including code generation and support chatbots, for immediate deployment and tangible returns.

Dec 2, 2024·Healthcare IT News

The Radiological Society of North America's annual meeting in Chicago is drawing attention to the impact of artificial intelligence (AI) in imaging, with a KLAS study revealing over 300 FDA-approved AI tools currently in use. The report indicates that more than half of surveyed organizations now implement AI for imaging applications, a marked increase from 17% in 2018, suggesting a significant shift in the adoption of these technologies. At the conference, companies like Agfa and Aidoc are showcasing their latest AI innovations to enhance diagnostic efficiency in healthcare imaging.

Dec 2, 2024·LinkedIn

Although AI scribes have shown significant benefits in healthcare, expectations need to be tempered, according to Spender Dorn, Vice Chair & Professor of Medicine at UNC. The value needs to extend beyond documentation to effectively integrate with broader clinical processes, he said, citing a recent NJEM article. The analysis of DAX Copilot's impact raises important considerations about the need for robust evaluation of AI tools in clinical practice, as merely adopting technology does not guarantee improvements in efficiency or productivity among healthcare providers.

Dec 2, 2024·Becker's Hospital Review

Mercy, a healthcare system in Chesterfield, Missouri, represented the industry on Capitol Hill during the first CHAI on the Hill Day, organized by the Coalition for Health AI. As the sole healthcare organization at the event, Mercy engaged with policymakers to discuss the applications and regulatory needs of artificial intelligence in healthcare. President and CEO Steve Mackin underscored AI’s potential to improve patient care while advocating for responsible usage that prioritizes safety. Mercy is actively collaborating with companies like Microsoft and the Mayo Clinic to develop AI tools for patient communication and data analysis to enhance care delivery and operational efficiency.

Dec 1, 2024·Apple News

Microsoft has introduced a new feature for Teams that allows users to clone their voice using advanced AI technology. This capability aims to facilitate communication by creating a digital replica of a user's voice, which can be utilized during virtual meetings, particularly when an individual cannot speak due to illness or other commitments. The technology analyzes speech patterns to generate a personalized voice model that integrates seamlessly into calls. Microsoft highlights the importance of user privacy and has implemented measures to ensure control over voice data, allowing users to opt-in to the service.

Nov 26, 2024·Healthcare IT News

Mount Sinai Health System has established the Hamilton and Amabel James Center for Artificial Intelligence and Human Health at the Icahn School of Medicine to enhance healthcare delivery through AI integration. This 65,000 square-foot facility, located in Manhattan, brings together various research programs and houses around 40 principal investigators, along with 250 students and staff. The center aims to lead in AI research within U.S. healthcare, focusing on applying AI across genomics, imaging, and electronic health records, while also hosting institutes dedicated to digital health, genomic health, and personalized medicine. Mount Sinai has been actively exploring AI applications, including surgical video analysis and medical coding automation, to improve health services.

Nov 24, 2024·Harvard Business Review

Inefficiencies in healthcare practices hinder patient experiences, highlighting the need for more personalized care through smart processes and AI tools. Achieving such customization requires focusing on essential processes that prioritize quality data collection. Although electronic health records are widely adopted, the healthcare sector still struggles with effectively using patient data to enhance care, falling short of the personalization seen in other industries like retail. Robbie Hughes notes that redesigning and automating key care processes is critical for improving personalization, stressing the importance of leveraging patient data to tailor healthcare delivery.

Nov 21, 2024·LinkedIn

Allie K. Miller discussed a study comparing the diagnostic abilities of ChatGPT and human doctors, revealing that while physicians using ChatGPT achieved a diagnostic accuracy of 76%, ChatGPT alone had a 90% accuracy rate. The research identified two key challenges: physicians' overconfidence led them to dismiss ChatGPT's correct diagnoses, and many doctors lacked the training to effectively integrate AI tools into their practice. Miller highlighted the importance of fostering trust in AI to enhance diagnostic processes and urged medical professionals to pursue education that improves their competency in utilizing these technologies.

Nov 20, 2024·Healthcare IT News

The American College of Radiology (ACR) has launched the National Radiology Data Registry to evaluate the performance of artificial intelligence (AI) algorithms in radiology while collecting patient data and clinical metadata. This initiative will enable radiology practices to benchmark their results against national standards, enhancing the understanding of AI effectiveness in real-world settings. The ACR's Assess-AI platform will provide performance reports to clinical sites and developers, aiding in the continuous improvement of AI technologies. Dr. Christoph Wald noted the necessity of this registry as traditional radiology systems lack ongoing evaluation capabilities for AI, which is crucial for maintaining care quality. Key features of the registry include monitoring algorithm stability and ensuring consistency between AI results and radiology reports.

Nov 19, 2024·Sentara

Sentara Health is integrating artificial intelligence (AI) into its healthcare operations through an AI Oversight Program managed by a committee of senior leaders. The program is guided by eight principles that prioritize safety, ethics, and accountability while promoting responsible AI use. Key areas of focus include human oversight, technical robustness, and privacy, aiming to enhance patient care and safety. One application of this strategy is the Microsoft DAX Copilot, which aids healthcare providers in efficiently creating clinical notes, thus alleviating administrative tasks and enabling more patient-centric interactions.

Nov 19, 2024·LinkedIn

Medication errors are a critical issue in the U.S. healthcare system, responsible for significant patient harm and economic costs. Despite technological advancements aimed at enhancing medication safety, such errors remain prevalent, accounting for a substantial portion of medical errors in hospitals. In response, clinical pharmacotherapists at Houston Methodist Hospital created the Medication Administration Protection System (MAPS), a software tool that integrates with electronic health records to leverage real-time data and machine learning. This system aims to detect medication errors and provide healthcare providers with immediate, actionable alerts, ultimately improving patient safety and maintaining efficient workflows.

Nov 19, 2024·Washington University in St. Louis

Washington University School of Medicine and BJC Health System have launched the Center for Health AI to transform healthcare delivery through artificial intelligence. This initiative aims to personalize patient care and improve efficiency by integrating AI to streamline administrative tasks and enhance clinician workflow. With tools already in use to assist with documentation and optimize scheduling, the center seeks to reduce clinician burnout and improve patient experiences. Philip R.O. Payne, PhD, will serve as the inaugural chief health AI officer, leading efforts that also prioritize educating future healthcare professionals in AI applications.

Nov 14, 2024·Becker's Hospital Review

Sanford Health has created a predictive AI model to improve early colon cancer detection by identifying individuals at higher risk using a comprehensive analysis of 85 different variables, including factors beyond standard age, family history, and lifestyle. This model leverages machine learning and electronic medical records to help physicians prioritize patients who may require urgent screenings, thus optimizing resource allocation for colonoscopies. The initiative aligns with recent recommendations to lower the age for routine screenings, reinforcing the importance of timely access to colorectal cancer care.

Nov 12, 2024·HealthLeaders Media

AI in healthcare, Jim Blum, CMIO at University of Iowa Health Care, stresses the need for a balanced approach in its integration. In a recent HealthLeaders podcast, he advocates for implementing AI as a tool to solve specific clinical issues rather than as an isolated solution, emphasizing that it should enhance workflows without adding complexity. UI Health Care has introduced two AI tools—an ambient transcription platform and a chart mining platform—that have gained traction among physicians, illustrating the effectiveness of involving clinicians and IT staff in evaluating AI technologies. Blum proposes simplified training for these tools to promote intuitive use, allowing clinicians to adapt at their own pace while ensuring access to support when needed.

Nov 11, 2024·Healthcare IT News

Beth Israel Deaconess Medical Center's OpenNotes Lab is launching a project in collaboration with AI developer Abridge to assess the effectiveness of AI in creating patient visit summaries. This initiative aims to improve the communication between patients and clinicians, ensuring greater transparency and health equity. Abridge's technology will generate structured clinical notes at an 8th-grade reading level, covering crucial aspects such as diagnoses and medications. Patient feedback on the AI-generated notes will be collected through focus groups while maintaining patient privacy. Since its inception in 2010, OpenNotes has enhanced patient engagement and is now a significant movement supported by the 21st Century Cures Act, requiring healthcare providers to share clinical notes with patients.

Nov 7, 2024·MedCity News

Generative AI is generating optimism in the healthcare sector for its potential to alleviate burnout, enhance operational efficiency, and improve patient care. Jason Hill, Chief Innovation Officer at Ochsner Health, highlights the need for standard governance and frameworks to ensure safe and effective AI use. He foresees a future of inevitable AI regulation, focusing on technological accuracy and operational readiness. Hill emphasizes the importance of a thorough evaluation process for AI tools in clinical settings, advocating for systematic reviews to assess their effectiveness before adoption.

Nov 7, 2024·Mayo Clinic Platform

Generative AI is making notable strides in the healthcare sector since the launch of ChatGPT in late 2022, attracting around 200 million weekly users, although concerns about misinformation persist. Advances in AI functionality, particularly through the development of transformers and attention mechanisms by data scientists, are improving diagnostic accuracy, as shown by research from the University of Cologne that indicates aggregating outputs from multiple large language models (LLMs) enhances trustworthiness in clinical settings. While these developments are promising, ongoing efforts to evaluate the effectiveness and safety of AI in real-world healthcare applications are essential, with tech companies like OpenAI actively working to enhance the reliability of their AI models.