Safe and Explainable AI for Cardiology, Breast Cancer, and Sepsis
This study is looking at how to make Artificial Intelligence (AI) safer and easier for doctors to understand when used in patient care. It's an observational study, meaning researchers will look at existing information rather than giving new treatments. The study focuses on AI-PERSONALIZED CLINICAL DECISION SUPPORT, which uses AI to help doctors make personalized decisions for patients. Researchers will be developing new AI learning algorithms and ways to explain AI recommendations, as well as methods to ensure AI safety. They will measure success by how well these new algorithms and explanation methods are developed and evaluated over 18 to 36 months. You could be eligible if you are 18 years or older and have been treated at Penn Medicine hospitals for cardiology, sepsis, or oncology conditions since 2017.
- Study design
- This is an observational study that plans to include 300,000 participants. It is not specified if it is randomized or blinded.
- What's involved
- Not specified in the trial record.
- Compensation
- Not stated in the trial record.
- Follow-up
- The study involves developing and evaluating AI methods over periods ranging from 18 to 36 months.
AI-generated from the public study record. Only the study team can confirm whether you're eligible — confirm details with them before making decisions.
Safe and Explainable AI
At a glance
Conditions
Where it's being run
1 sites across 1 statesWho to contact
Opens a ready-to-send draft in your own email app — review before sending.
What this trial measures
- Neurosymbolic Learning AlgorithmsPrototype and develop new learning algorithms; 18 months. Benchmark and evaluate the learning algorithms; 24 months. Publish research results; 24 months
Develop and evaluate novel algorithms for training neurosymbolic models. We will develop data- and compute-efficient algorithms for end-to-end training of neurosymbolic models. This task will reduce the burden on clinician experts to provide fine-grained labels on voluminous EHR data.
- Explanation MethodsPrototype and develop new explanation algorithms; 18 months. Derive certified guarantees for explanations; 18 months. Benchmark and evaluate the explanation algorithms; 24 months. Extend certificates to new properties and tasks; 30 months. Publ
We will develop new explainable AI techniques that come with verifiable guarantees. These guarantees will enable trust and transparency in AI at a fundamental level.
- Methods for Safety GuaranteesPrototype and develop new rule learning algorithms; 30 months. Scale rule learning algorithms to larger data settings; 36 months. Incorporate new primitives to express complex rules; 36 months. Implement rule learning algorithms on baseline tasks
We will develop new algorithms that can scalably extract complex logical rules governing safety within the data that have statistical guarantees. These techniques will be rooted in statistical analysis and assist users in identifying out of distribution data and detecting anomalies.