RECORD
MMLN: Leveraging Domain Knowledge for Multimodal Diagnosis
Machine Learning · Submitted 9 Feb 2022
abstract compiled · 2026-09-23 · no language-model review
The authors' main claim The results show that the proposed method outperforms the state-of-the-art multimodal baselines in terms of accuracy and interpretability.
The compiled record
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
Open the compiled record
The abstract, sentence by sentence
Recent studies show that deep learning models achieve good performance on medical imaging tasks such as diagnosis prediction. Among the models, multimodality has been an emerging trend, integrating different forms of data such as chest X-ray (CXR) images and electronic medical records (EMRs). However, most existing methods incorporate them in a model-free manner, which lacks theoretical support and ignores the intrinsic relations between different data sources. To address this problem, we propose a knowledge-driven and data-driven framework for lung disease diagnosis. By incorporating domain knowledge, machine learning models can reduce the dependence on labeled data and improve interpretability. We formulate diagnosis rules according to authoritative clinical medicine guidelines and learn the weights of rules from text data. Finally, a multimodal fusion consisting of text and image data is designed to infer the marginal probability of lung disease. We conduct experiments on a real-world dataset collected from a hospital. The results show that the proposed method outperforms the state-of-the-art multimodal baselines in terms of accuracy and interpretability.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Recent studies show that deep learning models achieve good performance on medical imaging tasks such as diagnosis prediction. | background 1.00 | 0.03 | 0.02 | asserted 1.00 |
| s1 | Among the models, multimodality has been an emerging trend, integrating different forms of data such as chest X-ray (CXR) images and electronic medical records (EMRs). | background 1.00 | 0.04 | 0.02 | asserted 0.99 |
| s2 | However, most existing methods incorporate them in a model-free manner, which lacks theoretical support and ignores the intrinsic relations between different data sources. | gap 0.96 | 0.06 | 0.02 | asserted 1.00 |
| s3 | To address this problem, we propose a knowledge-driven and data-driven framework for lung disease diagnosis. | contribution 0.99 | 0.84 | 0.05 | asserted 0.95 |
| s4 | By incorporating domain knowledge, machine learning models can reduce the dependence on labeled data and improve interpretability. | background 0.90 | 0.07 | 0.05 | asserted 0.55 |
| s5 | We formulate diagnosis rules according to authoritative clinical medicine guidelines and learn the weights of rules from text data. | method 0.97 | 0.91 | 0.03 | asserted 0.97 |
| s6 | Finally, a multimodal fusion consisting of text and image data is designed to infer the marginal probability of lung disease. | method 0.99 | 0.89 | 0.05 | asserted 0.97 |
| s7 | We conduct experiments on a real-world dataset collected from a hospital. | method 0.93 | 0.96 | 0.02 | asserted 0.97 |
| s8 | The results show that the proposed method outperforms the state-of-the-art multimodal baselines in terms of accuracy and interpretability. | result 1.00 | 0.96 | 0.90 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
The paper
Original abstract
Recent studies show that deep learning models achieve good performance on medical imaging tasks such as diagnosis prediction. Among the models, multimodality has been an emerging trend, integrating different forms of data such as chest X-ray (CXR) images and electronic medical records (EMRs). However, most existing methods incorporate them in a model-free manner, which lacks theoretical support and ignores the intrinsic relations between different data sources. To address this problem, we propose a knowledge-driven and data-driven framework for lung disease diagnosis. By incorporating domain knowledge, machine learning models can reduce the dependence on labeled data and improve interpretability. We formulate diagnosis rules according to authoritative clinical medicine guidelines and learn the weights of rules from text data. Finally, a multimodal fusion consisting of text and image data is designed to infer the marginal probability of lung disease. We conduct experiments on a real-world dataset collected from a hospital. The results show that the proposed method outperforms the state-of-the-art multimodal baselines in terms of accuracy and interpretability.
Human review
How the literature uses it
What this paper does with the arXiv papers it cites
What it cites
Compares with 4 papers
-
MultiModal BiTransformers (MMBT) [ref] : the underlying pretrained models are BERT [ref] and ResNet [ref] .
Deep Residual Learning for Image Recognition · Comparison of MMLN and multimodal baselines · 0.90
-
MultiModal BiTransformers (MMBT) [ref] : the underlying pretrained models are BERT [ref] and ResNet [ref] .
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding · Comparison of MMLN and multimodal baselines · 0.90
-
MultiModal BiTransformers (MMBT) [ref] : the underlying pretrained models are BERT [ref] and ResNet [ref] .
Supervised Multimodal Bitransformers for Classifying Images and Text · Comparison of MMLN and multimodal baselines · 0.90
Discussion
Know something about this paper? Leave a comment, or highlight any passage above to comment on it.