Get to know us

The Intelligent Health Computing (IHC) Core is a collaborative hub of computational scientists dedicated to advancing medical data processing and multimodal, image-guided systems. We integrate radiology, pathology, surgical, and physiologic data with clinical outcomes to build robust, AI-ready datasets and deploy models that elevate decision support, precision diagnosis, and treatment planning.

What we do

• Multimodal data integration: Fuse whole-slide pathology, MRI/CT/PET, ultrasound, endoscopic video, wearable signals, and EHR variables into coherent, harmonized datasets.

• Image-guided systems: Develop algorithms and workflows that inform intra-procedural navigation, tumor margin assessment, and therapy response monitoring.

• AI/ML pipelines: Design and validate models for segmentation, detection, prognosis, and treatment effect prediction using classical ML and modern deep learning (including foundation and multimodal architectures).

• Computational pathology: Build high-throughput WSI pipelines (tiling, QC, stain normalization, self-supervision) and link spatial features to molecular and clinical endpoints.

• Clinical deployment: Translate prototypes into secure, containerized services with monitoring, auditing, and human-in-the-loop review.

• Methods research: Advance algorithms for domain shift, uncertainty estimation, interpretability, and federated/secure learning.

Capabilities

• Secure data environment: HIPAA-aligned governance, IRB-aware data onboarding, de-identification, and PHI-safe audit trails.

• Scalable compute: GPU-accelerated training and high-memory inference for large images and video; workflow orchestration for reproducible experiments.

• Standards & interoperability: DICOM, FHIR, OMOP, and FAIR data practices for shareable science and smooth clinical integration.

• Quality & validation: Prospective test design, reader studies, statistically rigorous benchmarking, and model monitoring to guard against drift.

Collaboration & Impact

IHC partners with clinicians, basic scientists, and industry to co-design solutions that matter at the bedside. Recent efforts include:

• Multimodal biomarkers for early treatment response in oncology.

• Image-guided decision support for interventional radiology and neurosurgery.

• Digital pathology pipelines linking morphometrics with genomics for risk stratification.