Luca Celebrano

Medical Student / MD Candidate at Università degli Studi di Napoli Federico II
Department of Clinical Medicine and Surgery · School of Medicine and Surgery · Naples, Italy
ORCID: 0009-0007-0931-3555 · LinkedIn · l.celebrano@studenti.unina.it

Computational Oncology · Biophysical Reaction-Diffusion Modeling · Spatial Multi-Omics · Digital Pathology AI

---

Project	What It Solves & Why It Matters	1-Click Access & Verified Production
👑 Spatial Continuum Oncology	Why do 68% of lung cancers relapse post-ADAURA? Biophysical reaction-diffusion solver resolving why drug-sensitive persister cells survive in a 34.1 $\mu$m unperfused sanctuary cuff ($K_0$ Bessel decay).	🎮 60 FPS Simulator · 🏥 Web App
🩺 MedTerminal Cardio-Toracico	High-Yield Medical Exam Prep & Active Recall Engine Piattaforma didattica interattiva per studenti di medicina che demolisce lo studio mnemonico noioso: modelli 3D WebGL (Three.js), monitor ECG dinamici per riconoscere aritmie e STEMI al volo, fonoteca Littmann reale e quiz d'esame commentati.	▶ Prova la Piattaforma Didattica ↗ React 19 · TypeScript · Three.js (Live Production)
👁️ InSituVision-AI	Ultra-Low-Cost (< €5) Diagnostic AI Deep learning Attention-MIL neural network predicting microvascular decoupling and relapse risk directly from routine H&E whole-slide images.	🌐 Live Vision Portal PyTorch Attention-MIL
📐 Spatial Oncology Master	Bessel $K_0$ Analytical Solver & MSKCC Validation ($N = 604$) Cylindrical reaction-diffusion transport solver, Monte Carlo parameter sensitivity (10k draws), and stage-adjusted Cox modeling on real clinical cohorts.	🐳 Docker GHCR · 🌐 Web Portal Springer Journal of Mathematical Biology

Translational oncology and computational biophysics program investigating non-genetic targeted therapy resistance, reaction-diffusion drug transport, and low-cost diagnostic AI in EGFR-mutant non-small cell lung cancer (NSCLC). Affiliated with the Department of Clinical Medicine and Surgery, School of Medicine and Surgery, Università degli Studi di Napoli Federico II, Naples, Italy.

Conditions for Pharmacological Sanctuaries in Perivascular Tumour Cords: A Reaction–Diffusion Analysis, a Regrowth-Kinetics Identity, and a Descriptive Survival Analysis of Histologic Subtype in Resected Lung Adenocarcinoma
Target Journal: Nature Communications (Under Consideration, Tracking NCOMMS-26-091227)
Preprint DOI: 10.21203/rs.3.rs-11530312/v1 | Zenodo Master Archive: 10.5281/zenodo.23754999

• Oral Keynote Master Slide Deck (12 Slides): 👉 Open Presentation
• Official Congress Poster (120×90 cm Vector): 👉 View Poster
• Author-Reviewed Lay Summary (Gist.Science): 👉 Read Plain-Language Digest
• Full Camera-Ready Manuscript PDF: 👉 Download PDF

Official clinical AI and computational biology laboratory of Università degli Studi di Napoli Federico II on Hugging Face:
🔗 huggingface.co/spatial-continuum-lab

┌────────────────────────────────────────────────────────────────────────┐
│               🤗 spatial-continuum-lab (Federico II)                   │
├────────────────────────────────┬───────────────────────────────────────┤
│ 🩺 Bedside CRI Decision Engine │ Official bedside clinical risk web app│
│    (Interactive Space)         │ https://huggingface.co/spaces/        │
│                                │ spatial-continuum-lab/                │
│                                │ spatial-continuum-clinical-app        │
├────────────────────────────────┼───────────────────────────────────────┤
│ 🧬 Biophysical Digital Twin    │ 60 FPS perivascular reaction-diffusion│
│    (Simulator Space)           │ https://huggingface.co/spaces/        │
│                                │ spatial-continuum-lab/simulator       │
├────────────────────────────────┼───────────────────────────────────────┤
│ 📊 MSKCC LUAD Spatial Cohort   │ Curated multicenter cohort (N=1,274)  │
│    (Public Dataset)            │ https://huggingface.co/datasets/      │
│                                │ spatial-continuum-lab/                │
│                                │ mskcc-lung-adenocarcinoma-spatial-... │
└────────────────────────────────┴───────────────────────────────────────┘

Research Repository	Biophysical & Algorithmic Scope	Clinical Cohort / Evidence	Interactive Demos & Registries
👑 Spatial Continuum Oncology	Grand Unified Theory: Full reaction-diffusion continuum, modified Bessel $K_0$, regrowth kinetics, Bedside CRI, and Phase III SG-ADAURA trial allocation.	Multi-scale continuum • Zero third-party dependencies • Pure Python stdlib	Grand Unified Portal
Reaction-Diffusion Solver	Biophysical Transport: Perivascular cords, sanctuary sleeves ($\Delta r = r_{\text{ox}} - r_{\text{crit}}$), and multi-scale pharmacokinetics.	MSKCC clinical survival validation ($N = 604$) • SciPy ODE solver	Module 1: Bedside CRI
Study 4: InSituVision-AI	Ultra-low-cost (< 5 €) diagnostic AI: Attention-MIL neural network predicting microvascular decoupling directly from routine H&E WSIs.	$N = 570$ TCGA-LUAD cases • PyTorch Attention-MIL deep neural network	Live Vision Portal
Study 3: Massive Validation	Multi-cohort survival validation: Microvascular decoupling truncates median PFS from 78.6 to 24.0 mo ($p &lt; 10^{-15}$, $\text{HR} = 19.45$).	$N = 854$ patients (TCGA-LUAD + GSE31210 + GSE301973)	Module 5: Evidence
Study 2: Spatial Architecture	Subcellular transport limits: Microvascular decoupling in dense homotypic tumor sheets exceeding the $\lambda \approx 624,\mu\text{m}$ diffusion limit.	10x Visium HD ($211{,}104$ bins at $2,\mu\text{m}$ subcellular resolution)	Module 2: Visium HD
Study 1: InSituGame	In situ evolutionary game assay: Infers tumor subclonal payoff matrices ($A$) and evolutionary adaptive therapy regimens from spatial transcriptomics.	Pan-cancer ($20{,}093$ clinical spots) • Replicator dynamics solver	Module 4: InSituGame

All computational pipelines and CLI solvers are permanently packaged, versioned, and indexed in official public package managers:

# 1. Install the Grand Unified Spatial Continuum Framework (Flagship CLI)
conda install -c bioconda spatial-continuum

# 2. Or run via BioContainers Docker (Cloud / Cluster)
docker run --rm quay.io/biocontainers/spatial-continuum:1.0.2--pyhdfd78af_0 cri-calc --help

# 3. Or run via BioContainers Singularity (HPC / Hospital Clusters without root)
singularity exec https://depot.galaxyproject.org/singularity/spatial-continuum:1.0.2--pyhdfd78af_0 cri-calc --help

# 4. Install the In Situ Evolutionary Game Assay Engine
conda install -c bioconda insitugame

# 5. Install the Reaction-Diffusion Biophysical Sanctuary Solver
conda install -c bioconda spatial-sanctuary

An interactive multimodal clinical simulation and educational platform developed for medical students at Università degli Studi di Napoli Federico II:

• 3D Interactive WebGL Anatomy: Anatomical cardiac and thoracic models powered by Three.js.
• Simulated Real-Time ECG Telemetry: Dynamic waveform rendering across normal sinus rhythm, conduction blocks, and acute STEMI/NSTEMI patterns.
• Remastered Clinical Auscultation: High-fidelity audio syntheses of cardiac murmurs (AS, MR, AR) and respiratory lung sounds.
• Active Recall Diagnostic Modules: Case-based clinical reasoning puzzles and bedside medical calculators.
• Modern Web Stack: Built with React 19, TypeScript, Three.js, and Tailwind CSS v4.
  👉 Explore the live educational platform ↗

A verified record of peer-reviewed and merged contributions across bioinformatics, computational oncology, and core software ecosystems:

• Primary Institutional Affiliation:
  Department of Clinical Medicine and Surgery, School of Medicine and Surgery
  Università degli Studi di Napoli Federico II, Naples, Italy
• Academic Email: l.celebrano@studenti.unina.it
• Personal Contact: luca.celebrano1@gmail.com
• ORCID: 0009-0007-0931-3555
• LinkedIn Profile: linkedin.com/in/luca-celebrano
• GitHub Profile: @Lukecele
• Hugging Face Hub: @spatial-continuum-lab