Computational Visualization Center
All projects

Healthcare AI / Parent directory

Parkinson's disease

One public entry point for CVC's Parkinson's research across imaging, biomarkers, patient heterogeneity, and interpretable longitudinal modeling.

Posterior motor states connected to clinical assessments, patient-level phenotypes, DaTSCAN, and MRI validation
Program overview. Clinical assessments become soft motor-state profiles that can be compared with imaging-associated measures.

One research program

From clinical signals to interpretable patient views

These efforts are related parts of one Healthcare AI program. The public summaries keep implementation details light while showing the visual logic behind each line of work.

Posterior motor states and imaging-associated validation for Parkinson disease

Longitudinal clinical states

Posterior-aware motor phenotyping

Soft motor-state assignments make heterogeneity visible across visits and connect patient-level patterns to DaTSCAN and MRI validation.

Open project
Posterior-aware motor phenotyping pipeline for the MICCAI 2026 accepted paper

Interactive paper page

MICCAI 2026 accepted paper

Explore the accepted conference paper through the visual method flow, posterior-state explorer, BGMM configuration browser, and empirical result gallery.

Open paper page
Pathway-anchored multimodal Parkinson disease imaging framework

Interpretable imaging

Pathway-anchored PD clustering

Multimodal imaging features are organized around disease-relevant circuits so clusters can be read as pathway-level signals.

Open project
Integrated genetic, molecular, wearable, and prodromal biomarker framework

Genetics, assays, and wearables

Integrated precision stratification

A multimodal framework brings genetic risk, molecular assays, wearable sensing, and prodromal measures into one uncertainty-aware view.

Open project
Actionable Intelligence Parkinson disease project preview

Project site

Actionable Intelligence

An external project surface for patient-specific SBR biomarker exploration and clinical visualization workflows.

Visit project site

Visual explainability

Show the structure before the implementation

High-level figures make the reasoning visible without exposing unpublished system details. Open any project above for the full technical narrative and paper citations.

Posterior-aware phenotyping workflow from longitudinal assessments through model selection, triage, and imaging validation
Figure 2. A posterior-calibrated workflow connects longitudinal clinical data, uncertainty-aware motor states, external generalization, and imaging validation.
Posterior motor-state explainability panels showing domain scores, component selection, and temporal predictability
Explainability panels show how motor domains and temporal relationships shape the model view.
DaTSCAN and MRI imaging validation panels for posterior motor states
DaTSCAN and structural MRI provide visible validation anchors for the clinical representation.

Papers & evidence

Technical depth lives at the bottom of the page

These links anchor the public overview to the research record. The project pages carry the supporting figures and detailed methods.

  1. 01

    Posterior-Aware Motor Phenotyping with Multimodal Imaging Validation in Parkinson's Disease

    H. M. Tirhekar, P. Yadav, C. Bajaj. MICCAI 2026, accepted paper 4053.

    Open paper page
  2. 02

    Posterior-calibrated multimodal motor states reveal longitudinal and imaging-associated heterogeneity in Parkinson's disease

    H. M. Tirhekar, P. Yadav, C. Bajaj. bioRxiv 2026.

    Open paper
  3. 03

    Pathway-Anchored Multimodal Clustering for Parkinson's Disease

    A. Vinod, A. S. Ellendula, S. Bhardwaj, et al. bioRxiv 2025.

    Open paper
  4. 04

    Integrated Genetic, Molecular, and Wearable Sensor Biomarkers Enable Bayesian Machine Learning-Driven Precision Stratification in Parkinson's Disease

    H. M. Tirhekar, P. Yadav, C. Bajaj. medRxiv 2025.

    Open paper