Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jun 21, 2026

Guidelines and Experience Using Imaging Biomarker Explorer (IBEX) for Radiomics
10:17

Guidelines and Experience Using Imaging Biomarker Explorer (IBEX) for Radiomics

Published on: January 8, 2018

Integrated Hyperparameter Optimization with Dimensionality Reduction and Clustering for Radiomics: A Bootstrapped

S J Pawan1, Matthew Muellner1, Xiaomeng Lei1,2

  • 1Radiomics Lab, University of Southern California, Los Angeles, CA 90033, USA.

Multimodal Technologies and Interaction
|March 30, 2026
PubMed
Summary

Related Concept Videos

Bootstrapping01:24

Bootstrapping

The term "bootstrap" originated in the 19th century as a metaphor for self-improvement or achieving something independently, without external assistance. This concept extends to statistical bootstrapping, a self-contained method for estimating population parameters through resampling, even though it can be computationally intensive. Developed by the American statistician Dr. Bradley Efron in 1979, bootstrapping provides a robust way to perform inference when the original sample size is small or...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Improving readability of layperson abstracts and summaries in oncology using task-specific large language model powered tool: results from the BRIDGE-AI 7 study.

JAMIA open·2026
Same author

Radiogenomic analysis of muscle-invasive bladder cancer using CT-based texture analysis.

Bladder cancer (Amsterdam, Netherlands)·2026
Same author

Metabolic and Molecular Correlates of Medulloblastoma: Identification of Molecular Groups Using In Vivo  1H-MR Spectroscopy.

Neuro-oncology·2026
Same author

Deep learning-based detection and segmentation of osseous metastatic prostate cancer lesions on computed tomography.

European journal of radiology artificial intelligence·2026
Same author

Machine learning based classification of aggressive and malignant renal tumors from multimodal data.

PLOS digital health·2026
Same author

Pain Perception During Transperineal and Transrectal Prostate Biopsy Under Local An-esthesia: a Prospective Analysis of a Multi-ethnic and Diverse Cohort.

International braz j urol : official journal of the Brazilian Society of Urology·2026
Same journal

Ability-Based Methods for Personalized Keyboard Generation.

Multimodal technologies and interaction·2022
See all related articles
This summary is machine-generated.

This study introduces a novel hyperparameter optimization method for radiomic data clustering. The best pipeline, Non-negative Matrix Factorization (NMF) with K-means, accurately identified simulated clusters and showed moderate clustering in real-world renal cell carcinoma data.

Area of Science:

  • Medical Imaging Analysis
  • Machine Learning in Oncology
  • Quantitative Imaging Biomarkers

Background:

  • Radiomics extracts quantitative features from medical images, creating high-dimensional data.
  • Unsupervised clustering aims to uncover biological insights from radiomic features.
  • Optimizing dimensionality reduction and clustering pipelines is crucial but underexplored.

Purpose of the Study:

  • To develop and evaluate a novel bootstrapping-based hyperparameter search for optimizing radiomics clustering.
  • To treat dimensionality reduction and clustering as an integrated process chain.
  • To compare different unsupervised learning pipelines on simulated and real-world data.

Main Methods:

  • A bootstrapping framework with 100 iterations guided hyperparameter search using Adjusted Rand Index (ARI) and Davies-Bouldin Index (DBI).
Keywords:
clusteringhyperparametersmachine learningradiomicsrenal cell carcinoma

Related Experiment Videos

Last Updated: Jun 21, 2026

Guidelines and Experience Using Imaging Biomarker Explorer (IBEX) for Radiomics
10:17

Guidelines and Experience Using Imaging Biomarker Explorer (IBEX) for Radiomics

Published on: January 8, 2018

  • Cluster assignments were generated via 10-fold cross-validation and grid search for hyperparameter combinations.
  • Ten unsupervised learning pipelines were evaluated using simulated data and multiphase CT radiomics from renal cell carcinoma.

    • Main Results:

    • Non-negative Matrix Factorization (NMF) and Spectral Clustering outperformed Principal Component Analysis (PCA) in simulations.
    • The optimal pipeline (NMF + K-means) identified all three simulated clusters with a Cramér's V of 0.9.
    • Real-world data showed moderate clustering effects, correlating with weak associations to clinical outcomes (AUROC 0.63).

    Conclusions:

    • The proposed bootstrapping approach effectively optimizes radiomics clustering pipelines.
    • NMF followed by K-means clustering is a promising pipeline for radiomic data analysis.
    • The study provides a framework for evaluating pipeline concordance and highlights moderate clustering in renal cell carcinoma radiomics.