TRACE: Transformer-Based Risk Assessment for Clinical Evaluation
We present TRACE (Transformer-based Risk Assessment for Clinical Evaluation), a novel method for clinical risk assessment based on clinical data, leveraging the self-attention mechanism for enhanced feature interaction and result interpretation. Our approach is able to handle different data modaliti...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11028068/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850219614985781248 |
|---|---|
| author | Dionysis Christopoulos Sotiris Spanos Valsamis Ntouskos Konstantinos Karantzalos |
| author_facet | Dionysis Christopoulos Sotiris Spanos Valsamis Ntouskos Konstantinos Karantzalos |
| author_sort | Dionysis Christopoulos |
| collection | DOAJ |
| description | We present TRACE (Transformer-based Risk Assessment for Clinical Evaluation), a novel method for clinical risk assessment based on clinical data, leveraging the self-attention mechanism for enhanced feature interaction and result interpretation. Our approach is able to handle different data modalities, including continuous, categorical and multiple-choice (checkbox) attributes. The proposed architecture features a shared representation of the clinical data obtained by integrating specialized embeddings of each data modality, enabling the detection of high-risk individuals using Transformer encoder layers. To assess the effectiveness of the proposed method, a strong baseline based on non-negative multi-layer perceptrons (MLPs) is introduced. The proposed method outperforms various baselines widely used in the domain of clinical risk assessment, while effectively handling missing values. In terms of explainability, our Transformer-based method offers easily interpretable results via attention weights, further enhancing the clinicians’ decision-making process. |
| format | Article |
| id | doaj-art-87e0878f5db744d0838b4bb23a0b3149 |
| institution | OA Journals |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-87e0878f5db744d0838b4bb23a0b31492025-08-20T02:07:19ZengIEEEIEEE Access2169-35362025-01-011310172110173410.1109/ACCESS.2025.357797311028068TRACE: Transformer-Based Risk Assessment for Clinical EvaluationDionysis Christopoulos0https://orcid.org/0009-0006-5386-7902Sotiris Spanos1https://orcid.org/0009-0000-1436-1876Valsamis Ntouskos2https://orcid.org/0000-0003-1810-7802Konstantinos Karantzalos3https://orcid.org/0000-0001-8730-6245Remote Sensing Laboratory, National Technical University of Athens, Athens, GreeceRemote Sensing Laboratory, National Technical University of Athens, Athens, GreeceRemote Sensing Laboratory, National Technical University of Athens, Athens, GreeceRemote Sensing Laboratory, National Technical University of Athens, Athens, GreeceWe present TRACE (Transformer-based Risk Assessment for Clinical Evaluation), a novel method for clinical risk assessment based on clinical data, leveraging the self-attention mechanism for enhanced feature interaction and result interpretation. Our approach is able to handle different data modalities, including continuous, categorical and multiple-choice (checkbox) attributes. The proposed architecture features a shared representation of the clinical data obtained by integrating specialized embeddings of each data modality, enabling the detection of high-risk individuals using Transformer encoder layers. To assess the effectiveness of the proposed method, a strong baseline based on non-negative multi-layer perceptrons (MLPs) is introduced. The proposed method outperforms various baselines widely used in the domain of clinical risk assessment, while effectively handling missing values. In terms of explainability, our Transformer-based method offers easily interpretable results via attention weights, further enhancing the clinicians’ decision-making process.https://ieeexplore.ieee.org/document/11028068/Self-attentiontransformer encodertabular dataclinical datamelanomaheart attack |
| spellingShingle | Dionysis Christopoulos Sotiris Spanos Valsamis Ntouskos Konstantinos Karantzalos TRACE: Transformer-Based Risk Assessment for Clinical Evaluation IEEE Access Self-attention transformer encoder tabular data clinical data melanoma heart attack |
| title | TRACE: Transformer-Based Risk Assessment for Clinical Evaluation |
| title_full | TRACE: Transformer-Based Risk Assessment for Clinical Evaluation |
| title_fullStr | TRACE: Transformer-Based Risk Assessment for Clinical Evaluation |
| title_full_unstemmed | TRACE: Transformer-Based Risk Assessment for Clinical Evaluation |
| title_short | TRACE: Transformer-Based Risk Assessment for Clinical Evaluation |
| title_sort | trace transformer based risk assessment for clinical evaluation |
| topic | Self-attention transformer encoder tabular data clinical data melanoma heart attack |
| url | https://ieeexplore.ieee.org/document/11028068/ |
| work_keys_str_mv | AT dionysischristopoulos tracetransformerbasedriskassessmentforclinicalevaluation AT sotirisspanos tracetransformerbasedriskassessmentforclinicalevaluation AT valsamisntouskos tracetransformerbasedriskassessmentforclinicalevaluation AT konstantinoskarantzalos tracetransformerbasedriskassessmentforclinicalevaluation |