Powered by GPT-5.2 · Fisher's z · Cox / KM

Radiogenomic intelligence
for cohort research

Upload X-rays, CT, MRI and gene expression data. Get publication-grade gene–feature correlations, volcano plots, Kaplan–Meier survival curves, unsupervised phenotype clustering, AI-written insight reports and cross-study meta-analysis — all tied to the studies your lab is running.

3-day free trial · 4 scans · full analysis-module access (except sharing) · no card required. Paid plans from $ 100/mo.

Cohort insight preview

TLR4 → GLCM Entropy

+0.83
r
6.2e-7
p-value
128
N

Pathway: Innate immunity · TLR signaling

AI summary: TLR4 expression positively correlates with lesion texture heterogeneity, consistent with neutrophil infiltrate. Median-split KM shows worse 12-mo survival in TLR4-high subjects (p=0.014).
20+
Analysis modules
79
Genes in panel
100
Top correlations
GPT-5.2
AI insights

Everything in one platform

From scan to manuscript, fully integrated

Every analysis is bound to a Study/Project. Upload once — get correlations, pathway enrichment, survival curves, clusters and an AI insight report in the same workspace.

Chest imaging AI (X-ray / CT / MRI)

Upload chest X-rays, CTs or MRIs and run single-scan AI classification with GPT-5.2 Vision — clinically-framed prognosis output plus DICOM support.

Two workflows: Paired Radiogenomics & External Radiotranscriptomics

Choose your track at study creation. Paired (imaging + gene expression from the SAME patients) unlocks individual-level association testing; External (independent public transcriptomic datasets) runs cohort-level, hypothesis-generating analysis — each with the correct scientific guardrails and report labels.

Cohort gene–radiomic correlations (Spearman/Pearson + FDR)

Upload a subjects×genes matrix and compute true cross-cohort correlations against radiomic features — Spearman (default) or Pearson, with Benjamini-Hochberg FDR across all pairs, gene-symbol standardization, QC and paired subject-ID validation.

Pathway–radiomic correlation

Per-subject pathway activity scores (z-scored gene signatures across 8 curated pathways) correlated with radiomic features — more biologically meaningful than single-gene results, FDR-corrected.

Differential expression by imaging phenotype

Split the cohort by a radiomic feature (median/tertile) or diagnosis and find differentially-expressed genes — log₂FC, Mann–Whitney U, FDR, plus a volcano plot and top up/down tables.

Interactive volcano plot

Every gene-feature pair plotted as r vs −log₁₀(p). Filter by significance, |r| threshold or pathway, then click any point to drill into GeneCards and Reactome.

Pathway enrichment (ORA)

Hypergeometric over-representation with Benjamini-Hochberg FDR on built-in panels — see which pathways the significant correlations cluster into.

Prognosis: Kaplan–Meier + Cox regression

Upload subject survival CSVs, stratify by any gene or radiomic feature for Kaplan–Meier + log-rank, and fit real Cox proportional-hazards models (radiomic/gene/pathway predictors) with hazard ratios, 95% CI, FDR and concordance index.

Unsupervised phenotyping

KMeans + PCA on z-scored radiomic + gene matrix reveals subject subtypes with per-cluster top discriminating features and silhouette quality.

Scans Compare (Group-vs-Group)

Compare two scans — or two subject groups — side by side, with per-feature deltas and group-vs-group significance.

Gene co-expression network

Build correlation-driven gene co-expression networks with edge thresholding and hub-gene detection, plus differential co-expression across cohorts.

Protein–Protein Interaction (STRING-DB)

Map your significant genes onto STRING-DB protein interaction networks to expose functional modules and candidate driver proteins.

Imaging → external signature mapping

Track B: map each imaging phenotype group to the most biologically-consistent external transcriptomic signature (with confidence tiers) — cohort-level, hypothesis-generating, never patient-level.

Cross-dataset replication

Test whether a query signature replicates across multiple independent public datasets — per-dataset significance plus a strong/moderate/weak/not-replicated verdict.

Cross-study meta-analysis

Pool gene-feature correlations across studies via Fisher's z-transform with Cochran's Q + I² heterogeneity. Ready for supplementary methods.

GPT-5.2 cohort insights

A senior-analyst-style markdown report synthesizes correlations, survival and clusters into Key Findings, Pathway Patterns and follow-up hypotheses.

Research Chatbot (GPT-5.5 + Claude)

Ask questions about your studies in plain language — GPT-5.5 for fast Q&A and Claude Sonnet 4.6 for manuscript-grade writing, grounded in your cohort data.

Cough TB Pre-Screening (audio)

Imaging + Audio AI concordance

Record or upload a cough, add a WHO-style symptom screen, and get an explainable AI pre-screening estimate (elevated / low / inconclusive) from real acoustic features. Link it to a scan's subject and it feeds the AI manuscript. Research/pre-screening only — not a diagnosis.

Team & study sharing

Invite collaborators to any study to co-edit, run analyses together and review results in one shared workspace — leave or unshare anytime.

Built for publication

Track-aware reports (Paired Radiogenomics vs External Radiotranscriptomics) with AI executive summary, suggested manuscript titles, mandated safety statements, and one report exported as CSV/JSON/PDF/DOCX — plus read-only share links and GeneCards/Reactome deep links throughout.

Simple, transparent pricing

Pick the tier that fits your lab

Pay monthly in USD via PayPal. Cancel any time.

Free Trial (3 days)

Free

4 scans / trial

  • X-ray / CT / MRI uploads
  • Single-scan AI analysis (GPT-5.2 Vision)
  • Per-cohort gene–feature correlations
  • Volcano plots
  • Pathway enrichment (ORA)
  • Team & study sharing
  • All CSV / JSON / PDF / DOCX exports
  • Scans Compare (incl. Group-vs-Group)
  • Kaplan–Meier survival curves
  • Unsupervised phenotype clustering
  • Gene co-expression network
  • Protein–Protein Interaction (STRING-DB)
  • Cross-study meta-analysis
  • AI-generated article reports (GPT-5.2)
Start free trial

Student

$ 100/ month

50 scans / month

  • X-ray / CT / MRI uploads
  • Single-scan AI analysis (GPT-5.2 Vision)
  • Per-cohort gene–feature correlations
  • Volcano plots
  • Pathway enrichment (ORA)
  • Team & study sharing
  • All CSV / JSON / PDF / DOCX exports
  • Scans Compare (incl. Group-vs-Group)
  • Kaplan–Meier survival curves
  • Unsupervised phenotype clustering
  • Gene co-expression network
  • Protein–Protein Interaction (STRING-DB)
  • Cross-study meta-analysis
  • AI-generated article reports (GPT-5.2)
Choose Student
Most popular

Researcher

$ 450/ month

Unlimited scans

  • X-ray / CT / MRI uploads
  • Single-scan AI analysis (GPT-5.2 Vision)
  • Per-cohort gene–feature correlations
  • Volcano plots
  • Pathway enrichment (ORA)
  • Team & study sharing
  • All CSV / JSON / PDF / DOCX exports
  • Scans Compare (incl. Group-vs-Group)
  • Kaplan–Meier survival curves
  • Unsupervised phenotype clustering
  • Gene co-expression network
  • Protein–Protein Interaction (STRING-DB)
  • Cross-study meta-analysis
  • AI-generated article reports (GPT-5.2)
Choose Researcher
Flexible

Pay Per Module

$ 50/ module / 30 days

Unlimited usage within paid modules · pay only for what you need

  • Pick 1 or more of 15 modules
  • X-ray / CT / MRI uploads
  • Single-scan AI (GPT-5.2 Vision)
  • Volcano plots & Pathway ORA
  • Kaplan-Meier survival
  • Phenotype clustering
  • Gene & PPI networks
  • Cross-study meta-analysis
  • AI-generated article (GPT-5.2)
  • Research Chatbot (GPT-5.5 + Claude Sonnet 4.6)
  • Cough TB pre-screening (audio) — RM/USD 100
  • All exports (CSV/JSON/PDF/DOCX)
  • Add more modules anytime
  • No bundle. No commitment.
Pick your modules

All plans are billed in USD via PayPal. Need a Lab/Institution plan with 10+ seats? Contact us.

Loved by researchers

What researchers are saying

5.0 / 5 · 4 reviews

Researchers highlight RespiraDiagnose for its impressive accuracy in prognosis classification and the valuable insights it provides into the biological associations of radiomic features. With its user-friendly interface and affordable pay-per-module subscription, it has proven to be a vital tool for data analysis and publishing impactful research in the field.

AI-summarized from verified researcher reviews

“I published my paper "Artificial Intelligence-Driven Prognostic Classification of COVID-19 Using Chest X-rays: A Deep Learning Approach" and compared my results using Microsoft CustomVision with the results from RespiraDiagnose - accurate and much better in terms of the details of prognosis. I like the Gene Network and Protein network interactions.”

Alfred Simbun

Researcher · MSU Shah Alam

“I published my paper "Classification of COVID-19 in chest X-ray images using DeTraC deep convolutional neural network" few years ago and stumbled upon RespiraDiagnose. It gave me the same accuracy of prognosis classification. Kudos!”

Asmaa Abbas

Researcher

“I really think that RespiraDiagnose has helped me to understand the biological association with every individual radiomic features of my xray images.”

Zahid

Researcher · UiTM Shah Alam

“I used this to scan my xray images two years ago and I saw so many new things being offered by this portal with affordable pay-per-module subscription package. It really helped me to analyze my data and publish my first paper.”

Isaac

Researcher · Management & Science University (MSU) Shah Alam

From DICOM to discussion section.

Start with the 3-day free trial. No credit card. Four scans and full analysis-module access — cohort correlations, survival, gene networks and more. Upgrade in one click when your trial ends.

Create your account