Inside a virtual hospital where we train AI doctors👩🏻‍⚕️

Hey!

Welcome to this week’s edition of Supermedic, where we explore the latest developments in artificial intelligence and its transformative impact on healthcare.

Let’s get into it!

Victor

TODAY’S MENU

• Inside a Virtual Hospital Where We Train AI Doctors

• ChatGPT Gets a Major Upgrade with GPT-4o

• AI in Radiology: A Boon or Bane for Radiographers? (SURVEY)

• The Top 10 AI Healthcare Companies in 2024

• AI Coaches Fall Flat for Weight Loss (STUDY)

Read time: under 5 minutes

FUTURE IS NOW

Inside a Virtual Hospital Where AI Doctors Are Trained

Researchers from the Institute for AI Research (AIR) at Tsinghua University (China) have created a virtual hospital simulation to train AI doctors.

Welcome to Agent Hospital, where AI agents roleplay as doctors, nurses, and patients, simulating the entire hospital experience from disease onset to post-hospital follow-up.

The goal? To create AI doctors that can handle tens of thousands of cases in a matter of days, a feat that would take human doctors years to achieve.

How it works:

• AI agents are assigned roles such as doctors, radiologists, nurses, receptionists…etc.

• Patient agents move through the simulated hospital process, just like real patients.

• AI doctors treat a potentially infinite number of patients, gaining experience rapidly.

The researchers believe this novel training strategy could make medical AI agents smarter and more efficient than traditional methods like pre-training and fine-tuning.

Looking in the Future

In a visionary future, every physical hospital could be accompanied by a parallel digital twin – a hospital simulation that mirrors its real-world counterpart. This virtual hospital would be equipped with access to the same wealth of information available to human doctors, including patient histories, test results, and real-time monitoring data.

Imagine a scenario where a human doctor, at any given moment, requires a second opinion. The human doctor could engage in a discussion with its virtual agent about the patient in question, seeking guidance on the next steps in treatment, weighing the advantages and disadvantages of various approaches, and more.

This kind of context-aware support, available 24/7, could pave the way for a world where AI and doctors work closely together to provide better patient care.

THEY SUPPORT US 🩵

Learn How AI Impacts Strategy with MIT

As AI technology continues to advance, businesses are facing new challenges and opportunities across the board. Stay ahead of the curve by understanding how AI can impact your business strategy.

In the MIT Artificial Intelligence: Implications for Business Strategy online short course you’ll gain:

• Practical knowledge and a foundational understanding of AI's current state

• The ability to identify and leverage AI opportunities for organizational growth

• A focus on the managerial rather than technical aspects of AI to prepare you for strategic decision making

LEARN MORE

CHATBOT

ChatGPT Gets a Major Upgrade with GPT-4o

OpenAI has released a significant update to ChatGPT, introducing the new GPT-4o model. This multimodal AI can understand text, images, video, and audio, offering enhanced reasoning, processing, and natural language capabilities.

What’s New?

• GPT-4o is now available for free ChatGPT users.

• Multimodal capabilities spanning visual, audio, and text interaction, including real-time use with phone cameras.

• Human-like conversation abilities, allowing users to interrupt the AI assistant during interactions.

• Operates across 50 languages, now enabling real-time translation.

• Desktop app for Mac users (Windows app is coming later this year)

How to Access GPT-4o:

1. Sign in to your ChatGPT account or create a new one.

2. Check the model drop-down menu for GPT-4o. If it's not listed, you don't have access yet (the rollout is gradual, with access being granted in batches).

3. Start chatting with GPT-4o, keeping in mind the rate limits for free users.

4. You can change the AI model during a conversation by selecting the sparkle icon ✨

5. Upload files (images, videos, PDFs) for analysis and ask questions about the content.

RADIOLOGY

AI in Radiology: A Boon or Bane for Radiographers?

A recent survey of nearly 400 radiographers, published in BMC Medical Ethics, reveals mixed feelings about the integration of AI in radiology. While AI offers potential benefits such as improved diagnostic accuracy and workflow efficiency, radiographers are apprehensive about the ethical challenges it presents.

Key Concerns:

• Transparency and justification of AI decision-making (40% of respondents)

• Ethical dilemmas, including bias, data privacy, and ownership (44%)

• Potential job displacement and the need for further training

The study's author, Faten Mane Aldhafeeri, suggests implementing ongoing professional development programs specific to AI in radiology. This would empower radiographers to navigate ethical challenges and contribute to the responsible use of AI in healthcare.

BUSINESS

What Are the Top 10 Healthcare AI Companies?

Explore the ten leading companies at the forefront of AI advancements in hospitals, clinics, and the pharmaceutical industry:

1. Google DeepMind: Launched AlphaFold 3 to model protein and DNA structures, aiding scientific discoveries.

2. Microsoft Cloud for Healthcare: Streamlines workflows and unlocks insights through AI-powered analytics.

3. Merative LP: Formerly IBM Watson Health, offers AI-driven clinical and operational decision-making tools.

4. Amazon Web Services (AWS): Provides cloud infrastructure for optimizing drug discovery and treatment.

5. Intel: Enhances healthcare access and affordability with AI use cases like medical imaging and predictive analytics.

6. Nvidia: Offers advanced imaging and predictive models through its suite of healthcare microservices.

7. Siemens Healthineers: Automates clinical routines with scalable AI solutions.

8. Philips Healthcare: Provides diagnostic tools and an AI Manager for radiology departments.

9. Komodo Health: Offers data-driven insights to improve patient care and reduce disease burden.

10. Babylon Health: Connects patients with healthcare professionals via an AI-powered platform for virtual consultations.

Source: https://healthcare-digital.com/top10/top-10-healthcare-ai-companies

CHATBOT

AI Coaches Fall Flat for Weight Loss, Study Finds

A new Northwestern Medicine study reveals that technology alone is not enough to replace human coaches in obesity treatment. The study compared two approaches: one using only technology and the other combining technology with human coaching.

Key Findings:

• Participants who initially received only technology had worse weight loss outcomes compared to those who had a human coach from the start.

• Even after intensifying treatment for those with suboptimal weight loss, the disadvantage persisted for six months.

• 25% of participants using only technology achieved clinically meaningful weight loss (5% of body weight) after six months without treatment augmentation.

Implications:

While the need for low-cost, technology-driven obesity treatments is urgent, current AI is not advanced enough to replace human coaches entirely. Researchers are working to understand what makes human coaches successful and how AI can better imitate their emotional tone and context awareness.

The study highlights the importance of matching treatment type and intensity to individual needs and preferences to effectively manage the obesity epidemic.

Other AI Healthcare news :

• Market Prediction: AI in healthcare market projected to reach $286 Billions by 2031, growing at a CAGR of 38.1% in the forecast period (2024-2031).

• Mental Health: New study shows AI can analyze social media posts to measure depression and anxiety more accurately than traditional surveys.

• Cardiology: UVA Health researchers created a new model called CARNA that predict heart failure outcomes, available to all clinicians.

• IHI (Innovative Health Initiative) released a comprehensive report on the safety benefits, risks, and recommendations for generative AI applications in healthcare.

• Counting Electric Sheep: New paper examines potential power and pitfalls of harnessing AI for sleep medicine.

Thanks for reading!

Let’s improve the newsletter together!🌟