Dear friends,
2021 was a very challenging but also very exciting year for MI4People!
After more than a year of preparation, planning, and engagement with German laws and bureaucracy, we have finally founded MI4People. It took us many discussions, sleepless nights of work, and doubts, but it paid off. This hard preparation work enabled us to take off quickly, so, that even though MI4People exists only for a couple of months, we have already reached a number of milestones:
We are continuously increasing our reach into social media, especially on LinkedIn. On various platforms, ca. 250 people are already following us including many talented MI-Experts and experienced senior IT-industry veterans (like CxOs, VPs, and Directors)
We already have 10+ volunteers who are either actively contributing to our current work or have been selected for our upcoming projects
We have already started our first donation campaign that could reach over 25 donors and collect over 2.300 €
But most importantly, we have already kicked off our first project! Our Soil Quality Evaluation System will be a platform that will support small farmers in developing countries by providing them with valuable information about their soil quality based on satellite images. This knowledge will enable smallholders to manage their soil in a more sustainable manner, increase their yield, and contribute to creation of stable food supply chains in regions that are plagued by a permanent danger of hunger outbreaks. We have already developed the first AI model that can predict organic carbon content in the soil in Africa and intend to use it as a showcase to get additional financing for the project. You can access our showcase via this link.
At this point we would like to thank all of you very much! Whether you supported us financially or on social media, or with advocacy within your network, you helped us a lot and encouraged us to continue.
Special thanks go to our volunteers who work for MI4People in their free time and help us with their great expertise!
We look forward to the new year and new challenges.
The entire MI4People team wishes you a happy and healthy new year.
Let us together make the world a better place for all of us.
Your MI4People Team
News
This edition is a bit different compared to our usual newsletter. This time, we do not consider news about MI for Public Good in various areas of Public Good but focus on only one area – healthcare. Let us know whether you like this small experiment.
Improving Mental Health Therapy with Help of AI
An Israeli startup, Eleos Health, is bringing an artificial intelligence assistant into the therapy room. Their AI system runs in the background of therapy sessions and generates the baseline for therapist’s clinical notes. It is trained to identify relevant information from the audio and was developed in a close collaboration with clinical psychologists. Thereby, the system not only transcripts the audio and identifies important passages, much more, it is able to spot most discussed subjects, what were the emotions expressed, and what techniques did the therapist use.
Eleos Health solution enables therapists to save time on administrative work and documentation which easily take 20% of the therapist’s work time. So, therapists who use this AI assistant are able to spend more time on the actual treatment of their patients and are less threatened by the risk of a burnout from documentation.
You can find more on this encouraging startup, its progress, and the personal motivations of its founder in this article.
AI Opens Incredible Perspective for Discovery of New Drugs by Cracking the Code of Protein Complexes
In the last two years, DeepMind’s AlphaFold2 and University’s of Washington RoseTTAFold has made a lot of headlines because these AI systems learned how to predict 3D structure of proteins. Since proteins are important building blocks of our bodies and are at the heart of most medication, the ability to gain knowledge about their structure quickly is like a holy grail for medicine because the time and cost for the development of new drugs can be drastically reduced.
However, one important detail was missing yet: proteins often don’t operate alone, rather, they associate into small groups (also called complexes) that interact to perform critical tasks in human cells. Now, the team from University of Washington has recently found a way to apply AI to tackle also this challenge. Using both AlphaFold and RoseTTAFold, they modified the programs to predict which proteins are likely to build complexes and sketched up the results into 3D models.
Using this method researchers predicted hundreds of protein complexes that are either entirely new or were structurally uncharacterized till now. This fundamental research is a breakthrough and will enable the search for new insights into how our cells grow, function, reproduce, age, and die. Correspondingly, it will enable the medical research and pharmaceutical industry to enter the new age of drug development and save many lives.
AI as a Promising Technology to Predict who Will Develop Dementia
A new study from University of Exeter and The Alan Turing Institute in the UK has shown that machine learning algorithms can accurately predict which people who attend memory clinics will develop dementia within a two years period.
Researcher have analyzed clinical records of over fifteen thousand patients of memory clinics in the United States. Some of these patients got the diagnosis of dementia within two years. The research team has used various machine learning algorithms to prospectively predict which of these patients would get dementia based on clinically relevant variables collected during the initial visits of the patients. The study has shown that AI considerably outperforms existing conventional methods. In addition, AI could identify 80% of initially inconsistent diagnosis (positive diagnosis that were reversed later). In other words, applying AI instead of conventional methods could also significantly reduce the unnecessary stress for affected patients and their families that a wrong diagnosis could cause.
While this study represents ongoing research, we might expect real AI-driven solutions for combating this widespread disease in the near future.
Concluding Remarks
The above methods and technologies are only few recent examples of Machine Intelligence applications in the realms of medicine, pharmacy, and therapy. In fact, there are many more existing applications, ongoing research and even more potential use cases that have not been revealed yet. While we can be confident that our life expectation, health, and overall wellbeing will be enhanced by smart technologies in the foreseeable future, we should keep in mind that many people and even whole countries do not have even basic functioning healthcare system. These people are completely excluded from the technological advance in medicine and would not share our confidence about the future. We should do our best to let these people also benefit from the current MI-driven technological revolution in healthcare. This requires both political willingness and novel approaches to deliver the value generated by high-tech to people in developing regions.
Comments