HACID project

@[email protected]
18 Followers
4 Following
77 Posts

HACID aims at the study of hybrid human-artificial collective intelligence for open-ended domains like medical diagnostics and decision support for climate change adaptation policies.

The project aims at improving decision making in situations where knowledge fragmentation and information overload can strike.

HACID is a collaborative project funded under the Horizon Europe Programme, within the topic "AI, Data and Robotics at work".

Websitehttp://www.hacid-project.eu/
LinkedInhttps://www.linkedin.com/company/hacid-project/
Twitterhttps://twitter.com/hacid_project
Show threadHACID projectJun 25, 2024
2/8 🤝🤖🧠✨ We analysed 2,133 medical cases and 40,762 physician diagnoses from the Human Diagnosis Project to compare human-only, AI-only and hybrid collectives. The combination of AI and physician expertise produces better results than either alone.
Show threadHACID projectJun 4, 2024
Pooled decisions achieved highest accuracy when individual GPs used a Decision Support System (DSS) in their diagnostic process, showing that there can be synergy between DSS and collective intelligence approaches.
Show threadHACID projectJun 4, 2024
Using a vignette and actor-patient data set, we show that combining the decisions of multiple GPs increases diagnostic accuracy. The benefits were highest when weighing all diagnoses equally. Giving more weight to more confident or more senior GPs was not as effective.
HACID projectJun 4, 2024

New publication from HACID project showing that pooling the decisions of General Practitioners can increase diagnostic accuracy.

https://journals.sagepub.com/doi/10.1177/0272989X241241001?icid=int.sj-full-text.citing-articles.2

@mpib_berlin @stefanherzog

HACID projectMar 4, 2024
AI could make climate services more accessible and cheaper but do the risks outweigh the benefits?
We discussed this during a workshop aimed at surfacing current attitudes to using AI in the sector and to define a roadmap for safe implementation. Here are a few takeaways 🧵1/
Show threadHACID projectMar 4, 2024

The workshop was hosted at the @MetOffice headquarters and gathered 30 experts working in climate services, climate science and technology. The workshop was one of the first cross-sector meetings to explore the opportunities and risks of applying AI more widely.

Unlike AI for climate modelling, there’s been relatively little attention on how AI can support climate services, i.e. using climate data and information to support decision-making for climate change adaptation management 2/

Show threadHACID projectOct 30, 2023
Our solution substantially increases diagnostic accuracy: Single diagnosticians achieved 46% accuracy, pooling the decisions of ten diagnosticians increased this to 76%. Improvements occurred across medical specialties, chief complaints, and tenure levels
Show threadHACID projectOct 30, 2023
Using knowledge engineering methods, we introduce an automated solution to do this. We tested our solution on 1,333 medical cases, each of which was independently diagnosed by ten diagnosticians collected by the Human Diagnosis Project https://www.humandx.org/
The Human Diagnosis Project

Human Dx aims to empower anyone with the world’s collective medical insight.

Show threadHACID projectAug 1, 2023
Demonstration in #MedicalDiagnostics starts from interesting preliminary results, in which we aggregate responses from multiple users, thanks to the collaboration with Human Dx (https://www.humandx.org). Collective diagnoses achieve accuracy up to 80%, much better than the average individual results
The Human Diagnosis Project

Human Dx aims to empower anyone with the world’s collective medical insight.

HACID projectJun 7, 2023
The HACID project was presented at the 3rd conference of TAILOR, the 🇪🇺Network of Excellence for Trustworthy #AritificialIntelligence, held in the beautiful setting of the Certosa di Pontignano in Tuscany 🇮🇹
What is Trustworthy #AI, and how HACID contributes to it? 🧵👇🏽