MDF-Net for abnormality detection by fusing X-rays with clinical data - crossword
Across
- 5. Adjective used to describe deep learning that enhances diagnostic accuracy by using multiple data sources (like images and text) rather than vision alone.
- 6. The category of data that includes patient history, vital signs, and lab results, which acts as "critical context" for the diagnosis.
- 9. The "trap" encountered when using larger models like ResNet-50, where the model memorized training data instead of learning generalizable patterns.
- 10. A specific type of sign (like Heart Rate or Temperature) included in the structured clinical data used for the study.
Down
- 1. A term describing the nature of image data (high-dimensional, dense pixel arrays) as opposed to the linear, 1D nature of clinical data.
- 2. The medical professional whose workflow serves as the model for the study; they do not diagnose in a vacuum but combine visual evidence with context.
- 3. The specific "backbone" architecture that was selected as the "winner" because it offered the optimal balance of computational efficiency and accuracy.
- 4. The core strategy of the MDF-Net architecture, occurring at both "Early" and "Late" stages to merge different data dimensions.
- 7. A critical lung condition mentioned in the case study that appears as "hazy, diffuse opacities" and can look very similar to Pneumonia on an X-ray.
- 8. The approach that combines visual evidence with clinical context, which the presentation states moves AI closer to the reasoning of a human doctor.