Workshop Description

While computational methods in medical imaging have enabled us to detect and assess cancerous tumors and assist in their treatment, early detection of cancer precursors provides us with an opportunity for its early treatment and prevention. The survival rate of cancer is still low, and largely depends on the affected organ and how early it is diagnosed. The variable nature of the disease in different patients and the diverse imaging acquisition types involved for quantification of disease and treatment demands robust method designs. It is therefore critical to develop generalizable methods as part of a holistic early cancer detection ecosystem -- including different data analysis methods for various modalities and optimizing cross-modality fusion for improved detection and prognosis.

The workshop will invite researchers in the field of medical imaging around the central theme of data-driven cancer detection and treatment, and strives to address the challenges that are required to be overcomed to translate computational methods to clinical practice through well designed, generalizable (robust), interpretable and clinically transferable methods. Most current methods are developed on retrospective data that do not guarantee good representation of daily clinical procedures (e.g. domain gap). We aim to identify the new ecosystem that will enable comprehensive method validation and reliability of methods, setting up a new gold standard for sample size and elaborate evaluation strategies to identify failure modes of methods when applied to real-world clinical environments.

Workshop themes

Early detection and diagnosis of cancer: learning algorithms for lesion detection in medical images, staging, risk assessment, prediction of cancer outcome (in terms of life expectancy, survivability, progression, treatment sensitivity)
Image-guided intervention for cancer treatment: Image fusion, multi-modal registration, detection, segmentation, and tracking, computer-guided interventions, augmented reality for tumor delineation and tumor resection.
Real-world data exploration for cancer prediction: Big (imaging) data analysis, multimodal data analysis using active, semi and self-supervised learning, model-agnostic meta-learning, federated learning (sparsely labeled, data privacy), large foundational models (LLMs), and continual learning.
Cancer biomarkers: new predictive (visual) biomarker discovery in medical images, tumor data signatures, personalized cancer treatments, genomics and radiomics.
Clinically accepted evaluation methods: Identifying new evaluation metrics or gold standards (e.g., compared to widely used Dice or IoU), sample size standardization, biases, uncertainty estimation, best practices for validation, image simulation and synthesis techniques in cancer research.

Key technical themes

Medical image analysis methods targeted to the detection, diagnosis, treatment and/or monitoring of cancer that include a wide range of machine learning algorithms, workflows designed to improve patient outcome, data analysis and insights, and new imaging datasets.

Some key technical themes are listed below (not limited to):

Learning algorithms & workflows (e.g., semi and self-supervised, active learning, metric learning, model-agnostic meta-learning, unsupervised learning, outlier detection, LLMs, foundational models)
Data and label efficiency (e.g., limited data problem, data imbalance problem)
Model robustness and generalisability (e.g., edge-AI, cloud-AI, quantized models for clinical application)
Explainability, fairness and data privacy (e.g., model calibration, saliency mapping and federated learning for multi-center data)
Multimodal, and multi-instance learning

Important Dates

Paper submission begins: 2nd May 2026

Submission deadline: July 1, 2026

Paper decision notification: 31st July 2026

Camera ready submission: 1st August 2026

Workshop day: TBD

Submission

CMT submission website (TBD)

Accepted papers will be published in a joint proceeding with the MICCAI 2026 conference.

All papers should be formatted according to the Lecture Notes in Computer Science templates.

We recommend submission up to 8-pages and 2-pages of references (same as MICCAI main conference) for a double-blind peer review process

In addition, since the joint workshop has adhered to the double-blinded peer review process, we ask that you please follow the MICCAI2026 anonymity guidelines when preparing your intial submission.

Proceeding

Accepted papers will be published in LNCS as a separate CaPTion 2024 (MICCAI Workshop), proceeding

Call for Papers

Imaging themes (not limited to):

Optical imaging: Endoscopy, OCT, Hyperspectral imaging, opto-acoustics
CT/PET fusion, MRI
New imaging biomarkers
Multimodal imaging
Ultrasound

Clinical applications (not limited to):

Early cancer detection and diagnosis
Prognosis and prediction
Tumor characterisation, cancer staging
Longitudinal patient studies
Surgical data science
Digital histopathology
Phenotypic tumor correlation

Organising committee


Fons van der Sommen TU/e, Eindhoven, The Netherlands	Sharib Ali University of Leeds, UK	Noha Ghatwary AASTMT, Egypt

Bartek Papiez University of Oxford, UK	Yueming Jin National University of Singapore	Jiangbei Yue University of Leeds, UK