Nichula S. Wasalathilaka
Research for a more trustworthy world

Nichula S. Wasalathilaka

Building trustworthy AI systems for medical vision, remote sensing, and scientific discovery.

Final-year Electrical and Electronic Engineering undergraduate at the University of Peradeniya, Sri Lanka, and a visiting research student at MBZUAI, Abu Dhabi. I build uncertainty-aware, distribution-robust models for high-stakes scientific imaging, across trustworthy medical AI, Earth observation, and scientific machine learning.

Download CV (coming soon) Contact Me
Portrait of Nichula Sathmith Wasalathilaka
A prototype embedding space holding four coloured clusters of known medical-imaging classes, with a stream of test images passing above it and a panel of uncertainty-aware predictions alongside, in which an unfamiliar retinal image is flagged in red as a novel class.

Medical AI

Reliable AI for healthcare imaging

 A satellite imaging a populated valley from orbit, beside a stack of co-registered raster layers — true colour, infrared, radar and thermal — and panels showing a colour-coded land-cover segmentation, an infrared response map and a terrain elevation model.

Remote Sensing

AI for a more sustainable planet

 A translucent human chest showing the bronchial airway tree coloured by simulated airflow speed, with a pipeline strip running from CT slice to segmented airway to surface mesh to flow solution, a magnified view of the mesh at a bifurcation, and a series of progressively branching generated airway trees.

Scientific Computing

Machine learning for physical and biomedical systems

Selected Papers

  1. Chest X-ray with Graph-of-Differences anatomy-graph nodes over the lung and heart regions

    Graph-of-Differences: Anatomy-Structured Difference Alignment for Medical Image Re-Identification

    N. Wasalathilaka, A. Das, I. Razzak, D. Mahapatra

    MICCAI 2026

    Paper arXiv

  2. PROTON architecture diagram showing the frozen vision-language model and online prototype bank

    PROTON: Prototype-based Test-time Online OOD Detection for Medical Vision-Language Models

    A. Das, N. Wasalathilaka, Y. Lu, A. Dukre, D. Mahapatra, S. Khan, I. Razzak

    MICCAI 2026

    Paper arXiv

  3. Mamba-FCS encoder overview with the pre-change and post-change satellite inputs

    Mamba-FCS: Joint Spatio-Frequency Feature Fusion, Change-Guided Attention, and SeK Inspired Loss for Enhanced Semantic Change Detection in Remote Sensing

    B. Wijenayake, A. Ratnayake, P. Sumanasekara, R. Godaliyadda, P. Ekanayake, V. Herath, N. Wasalathilaka

    IEEE J-STARS · 2026

    Paper Website

All publications →

Recent Updates

  • Jun 2026 Two papers published at MICCAI 2026, Strasbourg, France.
  • Jun 2026 RespGeomLib published at IEEE Mercon 2026.
  • Mar 2026 Two papers accepted to IGARSS 2026, Washington, D.C., USA.
  • Feb 2026 Mamba-FCS published in IEEE J-STARS.
Research network

Global Research Network

Coauthors, collaborations, and conference presentations across institutions and venues.

  • Research base
  • Collaborator / coauthor
  • Conference / presentation
    Leave a note

    A small wall from amazing people.

    Kind words, shared moments, and lasting connections.

      Leave a note

      Optional — first name or initials are enough.

      Maximum 280 characters.

      0 / 280

      Submitted notes are reviewed before appearing publicly.

      Notes from visitors