}

Overview

This case study highlights how MediVerse's voice-driven VR interface enables sophisticated multimodal analysis of pediatric rhabdomyosarcoma (RMS) by integrating genomic and clinical data seamlessly, without SQL or manual navigation. Imported datasets include the Gene Expression Intensity Matrix and SVD Results.

Workflow

  1. Initial Scatter by Histology. Voice command: “Show a 3D scatter of V1, V2, V3 colored by Histology.”
    2. 3D scatter plot of spatial data by histology
    Fig. 5. Initial 3D scatter plot colored by Histology
  2. ERMS Samples Only. “Show a 3D scatter of V1, V2, V3 for ERMS samples only, colored by Histology.”
    3. 3D scatter of ERMS samples colored by histology
    Fig. 6. ERMS samples scatter colored by Histology
  3. Color by Age Status. “Show a 3D scatter of V1, V2, V3 for ERMS samples only, colored by AgeStatus.”
    4. 3D scatter of ERMS samples colored by age status
    Fig. 7. ERMS samples scatter colored by AgeStatus
  4. High-Risk ERMS Patients. “List all ERMS patients with AgeStatus equal to Unfavourable and survival_time less than 2 years.”
    5. Table of high-risk ERMS patients
    Fig. 8. Table of high-risk ERMS patients
  5. PAX3 Expression Line Chart. “Show a line chart of expression values across all samples for pax3.”
    6. Line chart of PAX3 expression across all samples
    Fig. 9. PAX3 Expression Across All Samples

Results & Impact

MediVerse enables intuitive, natural-language-driven exploration of integrated genomic and clinical datasets in VR. Researchers can uncover relationships between histological subtypes, prognostic age groups, and gene expression patterns, all within an immersive environment.