AI for HEALTH Lab

Bundling medical AI and healthcare expertise to save lives

partners

Challenges per year

250+

AI for Health Engineers

25,000+

hrs of Engineering

[data-wf-bgvideo-fallback-img] { display: none; } @media (prefers-reduced-motion: reduce) { [data-wf-bgvideo-fallback-img] { position: absolute; z-index: -100; display: inline-block; height: 100%; width: 100%; object-fit: cover; } }

WHAT IS AI FOR HEALTH LAB?

Empowering Medical Professionals and Patients with AI

With the AI for Health Lab we pool resources and share knowledge for the advancement of public health. Working with sensitive patient data is a tricky exercise and so is using AI to do critical diagnosis decisions. Through this Lab we are able to organize more AI for Good Challenges; deliver more real life healthcare solutions to medical centers; and train our machine learning community in specific skills relevant to medical AI.

The Mission Goals Tech in Focus Challenges Results Participate

WHY AI LAB FOR HEALTH PROTECTION?

Improving the quality and access to healthcare

Data science is making its way into medicine beyond labs and clinical research. Healthcare providers increasingly onboard bioinformaticians, medical and clinical data scientists requiring an AI-heavy skillset. AI-powered solutions are assisting with diagnoses, patient care and expanding treatment support beyond medical facilities. We can’t be missing here.

We are lucky to have a flourishing collaboration ongoing thanks to our AI for Health Chapter based at the Eindhoven University of Technology. Working with health institutes and universities all over the country, they organize AI for Health Challenges and health topic events.

Thanks to their support, support of our partner ecosystem and our community of AI for Health engineers developing medical AI solutions, we are able to contribute towards UN Sustainable Development Goal 3: Good Health and Well-being.

Left: Sample training picture

Right: Corresponding masks for building a Generative Adversarial Network (GAN) that accurately segments chest X-Rays to detect COVID19 in hospitalized patients

UN Sustainable Development Goal 3: Good Health and Well-being

Ensure healthy lives and promote well-being for all at all ages.

Targets

Events

Publications

978

Actions

Find out more about Goal 3

Goal 3 infographic, source: unstats.un.org/sdgs/report/2022/

WHAT ARE THE GOALS OF THE LAB?

Make life-saving data available while preserving privacy

Improve the speed & accuracy of diagnostics in radiology

Foster early detection & prevention of diseases instead of treating symptoms

Make preventive diagnostics more affordable

Improve access to mental healthcare

Model the progress and impact of infectious diseases

WHAT AI TECH DO WE FOCUS ON?

Medical Imaging

The use of artificial intelligence in diagnostic medical imaging has shown impressive accuracy and sensitivity in the identification of imaging abnormalities. Assisting healthcare professionals with detection, segmentation and classification, it can be of immense value for screenings and precision medicine. Our AI for Health engineers used Generative Adversarial Networks to segment X-rays of lungs to detect Covid-19. They applied U-nets and mask-RCNN for instance segmentation of cervical vertebrae to predict patient developing hernia.

Sensitive Patient Data

Medical AI is teeming with sensitive patient data used to train machine learning models. We teach our AI engineers how to handle sensitive information without compromising the model accuracy. We deploy techniques like Federated Learning to train algorithms across multiple decentralized servers holding local data without sharing them. Watch a recording from a Federated Learning Masterclass we’ve organized for an AI for Health Challenge.

Explainable AI

For transparent algorithmic decision making we deploy the techniques of Explainable AI or sometimes called Trustworthy AI. In healthcare, it’s important to see how the machine learning model arrived at a certain decision. How does it assess which preterm baby is at risk of getting sepsis or what does the neural network pay attention to when segmenting a medical image (as seen on our heatmap of a lung scan)?

Biosensor Analysis

We collaborate with medical professionals to deploy state-of-the-art medical analysis in our Challenges. In order to develop a usable AI solution, the participants need to understand the relevant hardware and the medical procedures. To improve hernia detection models for example, it’s of great value to have a neuroscientist explain the principles of radiology imaging, hernia diagnosis and its treatment to Challenge participants. Or the analysis of electrocardiograms to prevent heart failure.

AI FOR Health CHALLENGES

Completed

July 17, 2023

AI against Toxic Clouds

Use Computer Vision to detect toxic clouds emitted by large factories and report the authorities on the sightings

View challenge View more

Completed

March 26, 2023

AI against Diabetes

Construct Computer vision models that estimate the risk for diabetic ulcer formation, as well as highlight areas that indicate the risk.

View challenge View more

Completed

January 8, 2023

AI for Preterm Babies

Infer the circadian rhythm of preterm babies from UMC patients monitoring data and develop ML models to regulate light and sound to improve their development in incubators.

View challenge View more

Completed

May 25, 2022

AI for Health - Heart Failure Detection

Developing a machine learning model to predict blood values from electrocardiograms as an indication of heart failure.

View challenge View more

Completed

March 23, 2022

AI for Health - Preventing Sepsis

Early prediction of the risk of a preterm born baby developing sepsis by extending an existing classification model with more features, exploring more advanced models and developing a new explainable time-series model.

View challenge View more

Completed

November 12, 2021

AI for Health - Predicting Deterioration

Adapting and evaluating the deterioration index (DI) model to better predict the risk of deterioration of patients in the hospital beds.

View challenge View more

Completed

October 13, 2021

AI for Health - Hernia Detection

Detecting hernia on X-ray images in collaboration with hospital radiologists.

View challenge View more

Completed

June 18, 2021

AI for Health - Covid Detection

Improve the workflow of radiologists by automating lung segmentation in thorax X-rays to detect Covid

View challenge View more

Completed

March 18, 2020

AI against Covid-19 Challenge

How can we predict the growth of new infections? How do we know if our measures against COVID-19 are working? What are the implications for how the virus is spreading?

View challenge View more

Submit an AI for Good Challenge

Do you have a problem that can be solved by applying machine learning technology? We’ll crowdsource an AI solution in 10 weeks.

Submit a challenge

THE LAB RESULTS

We get some machine learning in medicine done

Develop prediction models for blood values from electrocardiograms as an indication of heart failure
Early risk prediction of a preterm born baby developing sepsis
Develop explainable time-series models for prediction
Detect hernia on X-ray images of a anonymized patient dataset
Automatic segmentation and tracking of cervical vertebrae over time
Segment lungs using Generative Adversarial Models
Predict economic effect of Covid 19 and its snowball effect on wellbeing
Model the burden of Covid 19 on the healthcare system

SPOTS Robi Beninca launching the flying ranger UAV (unmanned aerial vehicle) & Poacher detection running on the thermal video camera footage on edge hardware of the UAV

Results: Watch and read

From our YouTube channel

From our Publications

February 1, 2024

Tracking Turtles: How AI helps conservationists to re-identify sea turtles

Discover how AI revolutionizes sea turtle conservation, enabling accurate re-identification with non-intrusive methods. This collaboration with Sea Turtle Conservation Bonaire utilizes cutting-edge AI to match sea turtle photos against a database, enhancing our understanding of their migratory patterns and aiding in their preservation. The solution employs innovative techniques for turtle face detection, feature extraction, and matching, proving pivotal for habitat preservation efforts.

November 20, 2023

From Pixels to Preservation: How AI Gives Rise to a Birdwatching Revolution

At the interface of ecology and the ever-increasing applications to use artificial intelligence for good, FruitPunch AI and the Swedish University of Agricultural Sciences (SLU) joined together to better understand how our ecosystems function. Over decades, passionate ornithologists have been taking thousands of images of birds across the world. In the AI for Eagles Challenge, we exploited this dataset to build a machine learning pipeline that would allow rapid assessment of a bird’s behavior, species, and age. Traditional manual labeling of such images is a time-consuming process, and the need for speed in assessing individual birds and population states underscores the urgency of the task. The successful implementation of AI has the potential to revolutionize the efficiency of bird image analysis. By automating the labeling process, the application can significantly accelerate the assessment of individual birds and contribute valuable insights into the broader state of bird populations.

August 25, 2023

Flying High with AI: Counting Pelican Breeding Pairs in the Danube Delta

Imagine flying in a small airplane over the vast wetlands of the Danube Delta on the shores of the Black Sea in Romania looking for patches of small white dots: great white pelicans (Pelecanus onocrotalus). While flying over the colonies researchers like Sebastian Bugariu from the Romanian Ornithological Society (ROS) take photos which will be used to count the number of breeding birds when back in the office. The number of breeding pairs has grown from ~5,000 pairs 15 years ago to recently ~18,000 pairs. Keeping good records of the breeding numbers is important but not an easy task. Back at the office, it can take weeks to go through the images and manually count the pelicans. Wouldn’t it be great if this process could be automated, freeing up time that could be dedicated to other important conservation? This is where the AI for Pelican challenge started.

August 10, 2023

Solving automated wildlife taxonomy with AI

In the wild landscapes of Europe, camera traps have become essential tools for ecologists and wildlife researchers, sharing glimpses into the lives of the continent's diverse and fascinating animals. These devices have revolutionized our understanding of European wildlife, but their use comes with a challenge - the sheer volume of data they generate. To tackle this issue, a team of dedicated FruitPunchers recently developed an AI solution for the European Wildlife Challenge.

July 27, 2023

Listening to the Giants: Protecting Forest Elephants Through Audio Monitoring

In the dense rainforests of Central Africa, a captivating endeavor is underway, driven by researchers from Cornell University. Their mission? To track and protect elephants through the power of audio monitoring. These majestic creatures emit resounding rumbles at an incredibly low frequency, nearly imperceptible to the human ear. These deep calls, traversing vast distances, serve as a concealed communication network for elephants, veiled from our understanding for centuries.

March 23, 2023

Understanding seals with AI

On 6th December 2022, the FruitPunchAI community joined forces with researchers from Colgate University to build a facial recognition model that enables the non-invasive study and monitoring of harbor seals and other marine mammals. The team’s goal was to build upon the existing SealNet model developed by Colgate University to improve its accuracy, generalizability across sea species and mammals, and increase development velocity across data processing and model development workflows.

August 29, 2022

AI-powered Wildlife Conservation in Africa

An account of 10-week teamwork developing multiple machine learning and hardware pipelines to bring production-ready AI to edge hardware on flying rangers.

August 28, 2022

Model Optimization and Pruning of Poacher-detecting YOLOv5

Optimizing a YOLOv5 model for NVIDIA Jetson Nano to increase the inference speed and reduce memory footprint, focusing on inference speed not the absolute mAP.

August 27, 2022

User-friendly, Wilderness-proof MLOps

Developing a CI/CD pipeline that automatically retrains the existing poacher-detecting model, which can be deployed to the drone with improved performance.

August 26, 2022

Autonomous Flight and the Landing of a Fixed-Wing UAV

Developing a model for autonomous landing using a ‘traditional’ technique using GPS beacons a more experimental method of Reinforcement Learning.

Challenge results

Learning Resources

Artificial Intelligence

Time Series Forecasting

FOR INDIVIDUALS

How can you participate?

Become a certified AI for Health Engineer

Individual engineers who collaborate in the AI for Earth Challenges are trained in working with sensitive patient data; applying computer vision techniques for medical imaging and in other machine learning skills essential to medical AI.

Sign-up at the platform for free and fill your profile. Pick your interests and join corresponding communities. Create a skill tree you’ll be developing with every AI for Good activity.

Start my journey

Upgrade your skills with AI bootcamps & masterclasses

Learn how to apply your acquired AI knowledge in the real world. Join AI for Good challenges and collaborate in teams all over the globe.

Upgrade my skills

Every activity adds to your development. You’ll be accredited with badges for specific hard and soft skills and certifications after an accomplished challenge.

Build my skill tree

AI for Earth Skills

Computer Vision

Domain knowledge injection

GDPR

Explainable AI

Human aided AI

Differential Privacy

Federated Learning

Clustering

Anomaly detection

U-Net

Linear models

Wematic Segmentation

Grey box methods

Signal Processing

Biometric sensoring

Bioinformatics

Data anonimization

FOR INDIVIDUALS

How can you participate?

Become a certified AI for Wildlife Engineer

Individual engineers who collaborate in the AI for Wildlife Challenges are trained in edge computing, computer vision, model pruning and other skills essential for the applied AI projects.

Join the community

Sign-up at the platform for free and fill your profile. Pick your interests and join corresponding communities. Create a skill tree you’ll be developing with every AI for Good activity.

Start my journey

Start your personalized learning journey

Participate in a challenge

Join our machine learning focused learning events led by experts collaborating with the AI for Good community. Learn with the pros and from one another.

Join the community platform

Get certified

Every activity adds to your development. You’ll be accredited with badges for specific hard and soft skills and certifications after an accomplished challenge.

Get certified

For Organizations

Solve your challenges!

Challenge Partners

Medical & healthcare organizations and experts applying the results of the AI for Health Challenges. Do you have a medical or healthcare problem that can be solved with AI? Submit your challenge

Contribute!

Contributing Partners

Use their teams, expertise and resources to advance AI for Health Lab & Challenges.Can your experts or technology contribute to solving an AI for Health Challenge? Partner up! 👇

Get your community involved!

Community Partners

Use their reach to spread the AI for Good news, advocate for AI for Good Lab & Challenges.Do you have a network we could reach out to with medical AI and healthcare topics? Connect on Linkedin, Twitter or drop us a note 👇