TinyVit-LightGBM: A lightweight and smart feature fusion framework for IoMT-based cancer diagnosis
Bibliographic record
Abstract
Cancer remains a leading global health issue, where accurate and timely diagnosis is critical for effective treatment. The Internet of Medical Things (IoMT), an interconnected network of medical devices, offers real-time multimodal and multi-source data acquisition and analysis, facilitating remote monitoring and improving diagnostic precision. However, IoMT-based diagnostic frameworks face major challenges, including limited computational resources of IoMT devices, difficulties in integrating multimodal data from diverse sources, and the necessity for interpretable models to enhance clinical trust. To address these issues, we propose TinyViT-LightGBM, a lightweight and smart multimodal data fusion framework optimized for breast cancer diagnostics in resource-constrained IoMT environments. TinyViT, an efficient Vision Transformer, extracts features from multi-source histopathology images, combined with mammograms and clinical-genetic data through a comprehensive fusion strategy. By using LightGBM for classification, the framework not only achieves high diagnostic accuracy but also enhances interpretability by identifying the most critical diagnostic features. The proposed framework achieves state-of-the-art diagnostic performance, with 97.8% accuracy, a 6.5% improvement over existing methods, alongside gains in precision (97.2%), recall (99.1%), and F1-score (98.1%). Additionally, its low false positive rate (0.0058) and computational efficiency on IoMT devices underscore its scalability and suitability for real-world healthcare applications. • TinyViT-LightGBM is a novel multimodal data fusion framework for IoMT-based breast cancer diagnostics. • The model addresses the issues of computational constraints, multi source data, and data interpretability. • The LightGBM classify the vital feature for accurate decision-making. • The model achieves better diagnostic results with 97.8% accuracy, 97.2% precision, 99.1% recall, and 98.1% F1-score.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".