Professor
Konstantinos Kamnitsas
MSc PhD
Associate Professor of Engineering Science
Non-Tutorial Fellow at Wolfson College
Email:
College: Wolfson College

Konstantinos Kamnitsas is Associate Professor of Engineering Science (Medical Imaging) at the Department of Engineering of the University of Oxford, and Non-Tutorial Fellow at Wolfson College. He is co-director of the EPSRC CDT in Healthcare Data Science (2024-). His research focuses on Machine-Learning (ML) and primarily deep neural networks for medical image analysis. His work has two main goals:

  • Develop reliable, transparent and accountable AI models for safe use in healthcare.
  • Empower radiologists, clinicians and researchers with intelligent ML-based tools to better address their research questions and needs of clinical workflows.

Konstantinos completed his PhD at Imperial College London in 2019, where he pioneered development of 3-dimensional neural networks for analysing volumetric medical data, such as MRI and CT, and methods for improving generalization to heterogeneous data. His work won various awards, among which international competitions for segmentation of cancer and stroke lesions. He previously obtained an MSc in Computing Science from Imperial College, and Diploma in Electrical and Computer Engineering from Aristotle University of Thessaloniki, Greece. He has also conducted research in industry, such as at Microsoft Research and Kheiron Medical Technologies. He became Lecturer of Computer Science at the University of Birmingham in 2021, before moving to Oxford in 2022. He sits on the Editorial Board of the Medical Image Analysis (MedIA) journal.

We develop machine learning (ML) methods for medical image interpretation and analysis, driven by two main aims:

  • To develop more reliable and transparent machine learning models to catalyse safer integration of the technology in real-world applications.
  • To facilitate clinical research in a variety of applications (segmentation, detection, reconstruction, etc) and imaging modalities (MRI, CT, Mammography, X-rays, etc).

Therefore we investigate a variety of methodologies such as:

  • State-of-the-art neural networks for image understanding and analysis
  • Estimating model uncertainty or detecting potential failure of prediction for safe ML (due to corrupted input, unknown pathology, etc)
  • Identifying and alleviating bias in a model for fair ML in healthcare (domain adaptation, causality, etc)
  • Learning from decentralised data to enable international collaborations (federated learning, etc)
  • How to learn useful information from unlabelled data, multi-modal data, and more.

Biomedical Image Analysis

Mr Harry Anthony - DPhil Student
Ms Anjun Hu - DPhil Student
Mr Yasin Ibrahim - DPhil Student
Miss Ziyun Liang - DPhil Student
Mr Felix Wagner - DPhil Student
Miss Hermione Warr - DPhil Student
Mr Wentian Xu - DPhil Student

Deep Learning Methods for Medical Image Analysis

We develop neural-network based tools for detection/segmentation of pathologies (tumors, injuries, etc) and tissues of interest for a variety of imaging modalities (MR/CT/Mammography/Xrays/etc). Our aim is to empower clinical researchers with tools to tackle variety of tasks. Among other tools, we have developed and maintain DeepMedic, an open-source, easy-to-use Deep Learning segmentation tool for clinical research.

Detecting Failures of Neural Networks after Deployment

Performance of machine learning models may degrade when they process data that differ from those used during training (distribution shift). This poses a challenge for safe deployment of ML models in medical imaging due to data heterogeneity. We develop methods for detecting distribution shift, uncertainty estimation, and out-of-distribution detection to ensure reliable deployment of ML/AI models. Supported by EPSRC.

Federated Learning

We develop algorithms and optimization techniques that enable training neural networks on decentralised medical databases, held physically at multiple different institutions. Our aim is to enable international collaborations towards learning models that generalize better while preserving privacy of medical data. Supported by EPSRC.

CENTER-TBI

An international collaboration that aims to improve the care for patients with Traumatic Brain Injury (TBI). Our methods contribute in analysing this pathology in MRI and CT data and extract novel insights on the disease.

View a list of Professor Kamnitsas’ publications on Google Scholar.

 

Are you interested in studying for a PhD (DPhil) in Engineering Science with me? Below are the main routes:

  • Direct application for a DPhil at the Department of Engineering Science (via this link). Deadline 1 December 2023 (noon UK time). You will need to list me as supervisor. This is the preferred option for students that want to work specifically with me and have defined a research project/direction that fits my interests, which will need to be discussed within their submitted personal statement. Top applicants are notified around Febr-March. PhD duration is 3-4 years, starting next Oct. Note that although all successful applications are considered for scholarships by the Uni/Dept (see prev link), these are limited and very competitive. Applicants are strongly recommended to explore what other external scholarship opportunities they can pursue for supporting their studies.
  • Apply for a position at one of the Centers of Doctoral Training (CDT) that I am affiliated with: CDT on Health Data Science (HDS, link) or CDT on Autonomous Intelligent Machines and Systems (AIMS, link). This is a 4 year program, where during the 1st year, along with taught courses, students get to choose between projects proposed by affiliated faculty to pursue during the next 3 years for their PhD. CDT websites discuss, deadlines and funding of successful applicants.
  • Other opportunities will be advertised here when available.

Please have a look at my lab’s research areas and my Google Scholar to find my publications and identify my research interests. You can email me for specific follow up questions (e.g. whether a specific research question is of interest if you are considering a direct DPhil application). Please insert “[DPhil EngSci KK]” in your email subject to show you have read the above instructions. I apologise in advance for delays in replying, as the volume of emails can be quite large.

Preface DART 2023
Koch L,  Cardoso MJ,  Ferrante E,  Islam M,  Jiang M,  Kamnitsas K,  Rieke N,  Tsaftaris SA,  Yang D,  et al. (2024)
Semi-Supervised Learning for Deep Causal Generative Models
Ibrahim Y,  Warr H,  Kamnitsas K,  et al. (2024)
Author Correction: Federated learning enables big data for rare cancer boundary detection.
Pati S,  Baid U,  Edwards B,  Sheller M,  Wang S-H,  Reina GA,  Foley P,  Gruzdev A,  Karkada D,  Davatzikos C,  Sako C,  Ghodasara S,  Bilello M,  Mohan S,  Vollmuth P,  Brugnara G,  Preetha CJ,  Sahm F,  Maier-Hein K,  Zenk M,  Bendszus M,  Wick W,  Calabrese E,  Rudie J,  Villanueva-Meyer J,  Cha S,  Ingalhalikar M,  Jadhav M,  Pandey U,  Saini J,  Garrett J,  Larson M,  Jeraj R,  Currie S,  Frood R,  Fatania K,  Huang RY,  Chang K,  Balaña C,  Capellades J,  Puig J,  Trenkler J,  Pichler J,  Necker G,  Haunschmidt A,  Meckel S,  Shukla G,  Liem S,  Alexander GS,  Lombardo J,  Palmer JD,  Flanders AE,  Dicker AP,  Sair HI,  Jones CK,  Venkataraman A,  Jiang M,  So TY,  Chen C,  Heng PA,  Dou Q,  Kozubek M,  Lux F,  Michálek J,  Matula P,  Keřkovský M,  Kopřivová T,  Dostál M,  Vybíhal V,  Vogelbaum MA,  Mitchell JR,  Farinhas J,  Maldjian JA,  Yogananda CGB,  Pinho MC,  Reddy D,  Holcomb J,  Wagner BC,  Ellingson BM,  Cloughesy TF,  Raymond C,  Oughourlian T,  Hagiwara A,  Wang C,  To M-S,  Bhardwaj S,  Chong C,  Agzarian M,  Falcão AX,  Martins SB,  Teixeira BCA,  Sprenger F,  Menotti D,  Lucio DR,  LaMontagne P,  Marcus D,  Wiestler B,  Kofler F,  Ezhov I,  Metz M,  Jain R,  Lee M,  Lui YW,  McKinley R,  Slotboom J,  Radojewski P,  Meier R,  Wiest R,  Murcia D,  Fu E,  Haas R,  Thompson J,  Ormond DR,  Badve C,  Sloan AE,  Vadmal V,  Waite K,  Colen RR,  Pei L,  Ak M,  Srinivasan A,  Bapuraj JR,  Rao A,  Wang N,  Yoshiaki O,  Moritani T,  Turk S,  Lee J,  Prabhudesai S,  Morón F,  Mandel J,  Kamnitsas K,  Glocker B,  Dixon LVM,  Williams M,  Zampakis P,  Panagiotopoulos V,  Tsiganos P,  Alexiou S,  Haliassos I,  Zacharaki EI,  Moustakas K,  Kalogeropoulou C,  Kardamakis DM,  Choi YS,  Lee S-K,  Chang JH,  Ahn SS,  Luo B,  Poisson L,  Wen N,  Tiwari P,  Verma R,  Bareja R,  Yadav I,  Chen J,  Kumar N,  Smits M,  van der Voort SR,  Alafandi A,  Incekara F,  Wijnenga MMJ,  Kapsas G,  Gahrmann R,  Schouten JW,  Dubbink HJ,  Vincent AJPE,  van den Bent MJ,  French PJ,  Klein S,  Yuan Y,  Sharma S,  Tseng T-C,  Adabi S,  Niclou SP,  Keunen O,  Hau A-C,  Vallières M,  Fortin D,  Lepage M,  Landman B,  Ramadass K,  Xu K,  Chotai S,  Chambless LB,  Mistry A,  Thompson RC,  Gusev Y,  Bhuvaneshwar K,  Sayah A,  Bencheqroun C,  Belouali A,  Madhavan S,  Booth TC,  Chelliah A,  Modat M,  Shuaib H,  Dragos C,  Abayazeed A,  Kolodziej K,  Hill M,  Abbassy A,  Gamal S,  Mekhaimar M,  Qayati M,  Reyes M,  Park JE,  Yun J,  Kim HS,  Mahajan A,  Muzi M,  Benson S,  Beets-Tan RGH,  Teuwen J,  Herrera-Trujillo A,  Trujillo M,  Escobar W,  Abello A,  Bernal J,  Gómez J,  Choi J,  Baek S,  Kim Y,  Ismael H,  Allen B,  Buatti JM,  Kotrotsou A,  Li H,  Weiss T,  Weller M,  Bink A,  Pouymayou B,  Shaykh HF,  Saltz J,  Prasanna P,  Shrestha S,  Mani KM,  Payne D,  Kurc T,  Pelaez E,  Franco-Maldonado H,  Loayza F,  Quevedo S,  Guevara P,  Torche E,  Mendoza C,  Vera F,  Ríos E,  López E,  Velastin SA,  Ogbole G,  Soneye M,  Oyekunle D,  Odafe-Oyibotha O,  Osobu B,  Shu'aibu M,  Dorcas A,  Dako F,  Simpson AL,  Hamghalam M,  Peoples JJ,  Hu R,  Tran A,  Cutler D,  Moraes FY,  Boss MA,  Gimpel J,  Veettil DK,  Schmidt K,  Bialecki B,  Marella S,  Price C,  Cimino L,  Apgar C,  Shah P,  Menze B,  Barnholtz-Sloan JS,  Martin J,  Bakas S,  et al. (2023)
A Review of the Metrics Used to Assess Auto-Contouring Systems in Radiotherapy.
Mackay K,  Bernstein D,  Glocker B,  Kamnitsas K,  Taylor A,  et al. (2023)
Context Label Learning: Improving Background Class Representations in Semantic Segmentation.
Li Z,  Kamnitsas K,  Ouyang C,  Chen C,  Glocker B,  et al. (2023)
Joint Optimization of Class-Specific Training- and Test-Time Data Augmentation in Segmentation.
Li Z,  Kamnitsas K,  Dou Q,  Qin C,  Glocker B,  et al. (2023)