AI-ECG system saves lives in hospitals ❤️‍🩹

Hey!

Welcome to this week edition of Supermedic!

ChatGPT is no longer the top medical model: Google's Med-Gemini now takes the crown. The AI-ECG system is saving lives in hospitals. Generative AI are amazing, but understanding their limitations is crucial. Meanwhile, China is rapidly advancing in brain implants, aiming to compete with Neuralink.

Let’s get into it!

Victor

TODAY’S MENU

• AI-ECG System Saves Lives in Hospitals

• Google Introduces Med-Gemini, Can it Beat GPT-4?

• AI Tool Predicts Treatment Outcomes for Stroke Prevention

• ChatGPT Fails at Heart Risk Assessment

• China Races to Create Brain-Computer Interfaces​​

Read time: under 7 minutes

AI SAVES LIVES

AI-ECG System Saves Lives by Predicting Patient Risk in Hospitals

Researchers have developed an AI-enabled electrocardiogram (AI-ECG) that significantly improves the detection of critical health risks in hospitalized patients, reducing mortality rates. This study published in Nature Medicine has been conducted on nearly 16,000 patients at 2 Taiwan hospitals.

How It Works:

AI-ECG analyzes standard ECG data and transforms it into a percentile risk score. Patients are categorized as low or high risk based on this score. High-risk patients trigger an alert to physicians, prompting immediate and appropriate interventions.

Proven Impact:

The study show that AI-ECG significantly outperformed traditional methods by predicting not only general mortality but also specific risks such as arrhythmia-related deaths. In the intervention group, where AI-ECG alerts were active, there was a notable reduction in the mortality rate from 23% to 16% compared to the control group.

Why It Matters:

Looking forward, researchers anticipate that AI-ECG could be routinely used to monitor all hospitalized patients, providing real-time risk assessment and prompting timely interventions to improve patient outcomes and potentially save lives.

Integration of AI-ECG into hospital systems could revolutionize critical care, leading to a more personalized and effective approach to patient management.

THEY SUPPORT US ❤️

Work lesser & drive 10x more impact using AI

HIGHLY RECOMMENDED: A Power-packed workshop (worth $199) for FREE and learn 20+ AI tools to become 10x more efficient at your work.

👉 Become an AI Genius in 3 hours. Register here (FREE for First 100) 🎁
In this workshop you will learn how to: 

✅ Simplify your work and life using AI

✅ Do research & analyze data in seconds using AI tools

✅ Automate repetitive tasks & save 10+ hours every week

✅ Build stunning presentations & create content at lightning speed

👉 Register Now! (FREE for first 100 people only) 🎁

CHATBOT

Google Introduces Med-Gemini, an AI Healthcare Model that Beats ChatGPT

Researchers from Google and DeepMind have developed Med-Gemini, a new family of highly capable multimodal AI models specialised for medicine.

Med-Gemini stands out from other AI models in several ways:

• Superior Performances: Google claims that its new LLM outperforms competitors like GPT-4 in accuracy and nuanced reasoning for complex medical tasks.

• Enhanced Data Extraction: Med-Gemini can effectively navigate and analyze electronic health records (EHRs), extracting crucial information that might be hidden within the vast amount of data.

• Clinical Diagnosis and Treatment Support: The model can analyze patient data and images, potentially suggesting diagnoses and treatment options. A successful case demonstrated Med-Gemini's ability to accurately diagnose a skin lesion and recommend appropriate treatment.

Despite its promising capabilities, Med-Gemini is still under development and requires further refinement before being deployed in real-world clinical settings.
The model hasn’t been release for public use yet but you can access the open research paper here.

RESEARCH

AI Tool Predicts Treatment Outcomes for Stroke Prevention

Researchers from The Ohio State University have developed an AI model that shows promise in predicting the effectiveness of stroke treatments. Named CURE (CaUsal tReatment Effect estimation), this model uses massive datasets, including de-identified health care claims, to simulate the conditions of a randomized clinical trial—traditionally the gold standard in treatment evaluation.

How It Works:

CURE is pre-trained on broad health data and then fine-tuned with specific details about stroke risks and therapies to assess the potential impact of each treatment on individual patients. This approach has proven to outperform several existing models and aligns closely with the outcomes of actual clinical trials.

Why It Matters:

This method could dramatically accelerate the process of clinical trials by identifying a narrower set of effective treatments early on, saving time and resources while advancing personalized patient care. Additionally, the integration of biomedical knowledge graphs enhances the model’s ability to fill gaps in patient data, further refining treatment predictions.

CARDIOLOGY

ChatGPT Fails at Heart Risk Assessment: Study Raises Concerns

While ChatGPT shows promise in many medical areas, a new study cautions against relying solely on it for real-world medical assessments, such as determining whether a patient with chest pain requires hospitalization.

Here's why:

• Inconsistent results: ChatGPT provided different risk assessments (low, medium, high) for the same patient data.

• Doesn't match traditional methods: It fails to align with standard scoring systems traditionally used by doctors, such as the TIMI and HEART scales.

• Randomness in responses: ChatGPT is designed to vary its responses for natural language, but this inconsistency is dangerous in healthcare settings requiring a definitive answer.

While not ideal for risk assessment, ChatGPT's ability to analyze vast amounts of medical data can be a valuable tool for early diagnosis. It can assist doctors by suggesting multiple possibilities for complex cases, aiding them in reaching accurate conclusions.

BRAIN IMPLANTS

China Races Ahead in Brain-Computer Interface Tech, Raising Ethical Concerns

At a recent tech forum in Beijing, a Chinese company unveiled a brain-computer interface that allowed a monkey to control a robotic arm just by thinking about it.

Key insights:

• A company video showcases a monkey with its hands restrained using the BCI to grasp a strawberry with a robotic arm.

• The system, similar to Elon Musk's Neuralink, employs soft electrode filaments implanted in the brain.

• China was previously lagging behind the US in invasive BCIs but is now rapidly catching up.

• China plans to expand this technology beyond medical applications to include "cognitive enhancement, sleep regulation, memory regulation, exoskeletons...etc"

• Due to fewer ethical constraints and more government support, China is more likely to widely adopt BCI technology for commercial and military sectors.

China's advancements in BCI technology are significant. While exciting for medical applications, ethical considerations and potential military use raise concerns. The US needs to maintain a strong presence in this field to ensure responsible development and avoid falling behind.

Thanks for reading!

Let’s improve the newsletter together!🌟