Knowledge Graphs to facilitate Domain Adaptation?
A biomedicine study case
February 23, 2024
Edouard Albert-Roulhac &
Abdelhakim Sehad
What is a Knowledge Graph (KG)?
How do Knowledge Graphs work?
Applications of KG
Why are KG’s valuable in biomedicine and healthcare?
Why Knowledge Graphs in healthcare?
Research question
Can Multimodal Knowledge Graphs enable domain adaptation?
Relevant work
(https://arxiv.org/pdf/2305.19979.pdf)
What is BioKG?
Their approach
Some of the models
Lack of Multimodal Knowledge Graphs
(https://arxiv.org/pdf/2203.08893.pdf)
Introduces REMAP to build a high quality Knowledge Graph of relations between diseases
(https://arxiv.org/pdf/2002.08277.pdf)
Builds a Knowledge Graph specialized in radiology
These methods build KGs from image & text datasets but the resulting graph is not multimodal
Dataset
Chest X-Ray reports
Generate reports from X-Ray images
Variants include other tasks for chest X-Ray:
Experiments & Results
RaDialog results on the Chest X-Ray Report benchmark (SOTA)
CE: clinical efficacy
BS: BertScore
B-i: Bleu
MTR: METEOR
R-L: Rouge
Source: https://arxiv.org/pdf/2311.18681.pdf
Radiology Report Generation
Feed results from a multi-label feature extractor to the LLM
This results in little text information
Source: https://arxiv.org/pdf/2311.18681.pdf
Approach
Our approach is to incorporate knowledge from a graph to the LVM
KG embedding model will be used on top of the feature extractor
The intuition
Timeline
Thank You !
Let us know if you have any questions or comments !