Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

Christoffer Nellåker

Christoffer Nellåker

Christoffer Nellåker

Associate Professor

Computational phenotyping, medical image analysis

I am Associate Professor based in the Big Data Institute and part of the Nuffield Department of Women's and Reproductive Health. With my background in neurovirology, computational biology, machine learning and computer vision research, my groups focus resides firmly in translational interdisciplinary research.

Computational phenotyping

My research interests all fall within the wider scope of analysing high information content data (such as image data) to extract biologically meaningful information to aid understanding of basic biological processes and genetic associations, or direct impact on clinical pathways.

Histology image analysis

In collaboration with the Prof. Cecilia Lindgren group, and with research and clinical collaborators around the world we investigate histology based phenotyping of placenta and metabolic health relevant tissues.

Placenta is a severely understudied organ, considering its pivotal role in gestational development of a healthy fetus. A significant proportion of pregnancy and birth complications are known to have some etiological or associated changes in placenta structure or function. Commonly, placenta is sent for histopathological assessment after a severe adverse outcome during pregnancy - however the potential value of this tissue imaging is lacking due to perinatal pathologists' extreme workloads and relatively few quantitative metrics able to be analysed. Utilising a heirarchical bottom up approach we systematically analyse large microscopy imaging of heamatoxylin and eosin stained placenta tissue sections. By building up tissue phenotype signatures from cells, to tissue we can get robust signatures of morphology and structure relevant to organ health.

Adipose tissue is a key endocrine and energy storage organ for metabolic health and disease. By employing computer vision analysis of histological sections we extract phenoytpe metrics associated to disease states and genetic variants.


Antimicrobial resistance testing

Antimicrobial resistance has been identified as one of the largest threats to public health developing around the world. We are researching methods to speed up and improve clinical pathways for targeting the right microbes with the right antibiotics.

Through an ongoing collaboration with Department of Physics and the Nuffield Department of Clinical Medicine, grown under the Oxford Martin Programme on Antimicrobial Resistance Testing. Within this project we are looking to enable direct testing of clinical samples, using ultrasensitive microscopy tests, sophisticated image analysis and machine learning, hugely speeding up the process by which clinicians obtain the information they need.


Facial phenotyping for rare disease

Rare diseases are individually rare but collectively very common, however the clinical pathway to acurately diagnose a rare disease can be very long. We are researching using deep learning image analysis of facial images to help make diagnoses faster and more accurate.

It is estimated that one in seventeen people have some type of rare disease. Clinicians frequently look for characteristic changes in facial features to help find a diagnosis. Within the group we are translating the latest developments in computer vision and computational biology to aid diagnosis of rare diseases. The work is a collaborative effort to apply the latest techniques from facial recognition research for disease phenotyping. The aim is to bring this to clinical use to help narrow down the search for a correct diagnosis and to be used together with genome sequencing to identify mutations causing disease.