Blanca Inigo Romillo
About me
My work has focused on applying Deep Learning models for various medical applications, including diagnosis, surgical planning, and Mixed Reality guidance. My goal is to improve patient outcomes ensuring AI systems remain interpretable for smooth clinical integration. At the ARCADE Lab, I am currently developing the intelligent components for a novel Multi-robot System for Semi-automated Image-guided Vertebral Augmentation.
Links
Publications
-
An Intrinsically Explainable Approach to Detecting Vertebral Compression Fractures in CT Scans via Neurosymbolic Modeling
SPIE Medical Imaging -
Impact of Cone-beam CT Noise Correlation on Self-supervised Denoising Strategies for Low Dose Breast CT Imaging
SPIE Medical Imaging
Link -
Fastsam-3dslicer: A 3D-Slicer Extension for 3D Volumetric Segment Anything Model with Uncertainty Quantification
MedAGI
Link -
Fastsam3d: An Efficient Segment Anything Model for 3D Volumetric Medical Images
MICCAI 2024
Link -
Intelligent Control of Robotic X-ray Devices Using a Language-promptable Digital Twin
IPCAI 2025
Link
Experience
CS PhD | ARCADE Lab (JHU)
Jan 24 - Present
- Leading the development of AI-based models and Mixed Reality tools for a Multi-robot System for Semi-automated Image-guided Vertebral Augmentation.
DL Engineer | Vectech
June 23 - Dec 24
- Optimized state-of-the-art fine-grained image classification models to introduce color distribution awareness into predictions.
- Created a SAM-based interactive interface that generates and fine-tunes YOLO-format segmentation masks in real time.
- Utilized GAN models to generate realistic intraspecies diversity features, employing them as an augmentation method to enhance variability within a fine-grained dataset.
- Effectively trained state-of-the-art object detection models on a challenging dataset of noisy and non-standardized phone images of ticks.
AI/DL Researcher | I-STAR Lab (JHMI)
Oct 22-July 23
- Under the mentorship of Dr. Alejandro Sisniega, evaluated the impact of noise correlation in cone-beam breast CT (bCT) images on self-supervised denoising methods to improve image quality without requiring matched low-dose and high-dose data.
Test Analyst | Roche Diagnostics
Aug 21-July 22
- Tested the software and hardware of the coagulation analyzer “Cobas t”.
- Led updates to the test plans in line with the latest developments of the device.
Radiation Oncology Researcher | Ramón y Cajal Hospital
Sept 20-Jan 21
- Conducted a literature review on various immobilization systems for breast cancer treatment with Stereotactic Body Radiotherapy (SBRT), including GammaPod®, CyberKnife®, and Bodyfix®.
- Led the design and development of a tumor segmentation app to train future radiologists in identifying cancerous tissue on biomedical images for radiotherapy.
- Tested the app with medical residents, demonstrating its effectiveness as a training tool.