Kai Rakovic

Dr. Kai Rakovic

University of Glasgow, Glasgow, Glasgow City, UK
orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 February 2023 in Science Advances
Verified
Journal Article
Closed access
ARAlvaro Román-Fernández
Alvaro Román-Fernández
MAMohammed Ahmed
Mohammed Ahmed
FKFernanda G. Kugeratski
Fernanda G. Kugeratski
JAJayanthi Anand
Jayanthi Anand
ESEmma Sandilands
Emma Sandilands
DBDavid M. Bryant
David M. Bryant
The glycocalyx component and sialomucin podocalyxin (PODXL) is required for normal tissue development by promoting apical membranes to form between cells, triggering lumen formation. Elevated PODXL expression is also associated with metastasis and poor clinical outcome in multiple tumor types. How PODXL presents this duality in effect remains unknown. We identify an unexpected function of PODXL as a decoy receptor for galectin-3 (GAL3), whereby the PODXL-GAL3 interaction releases GAL3 repression of integrin-based invasion. Differential cortical targeting of PODXL, regulated by ubiquitination, is the molecular mechanism controlling alternate fates. Both PODXL high and low surface levels occur in parallel subpopulations within cancer cells. Orthotopic intraprostatic xenograft of PODXL-manipulated cells or those with different surface levels of PODXL define that this axis controls metastasis in vivo. Clinically, interplay between PODXL-GAL3 stratifies prostate cancer patients with poor outcome. Our studies define the molecular mechanisms and context in which PODXL promotes invasion and metastasis.

13 May 2022 in Diagnostics
Verified
Journal Article
Open access
PDF file
KRKai Rakovic
Kai Rakovic
RCRichard Colling
Richard Colling
LBLisa Browning
Lisa Browning
MDMonica Dolton
Monica Dolton
MHMargaret R. Horton
Margaret R. Horton
CVClare Verrill
Clare Verrill
There has been particular interest in the deployment of digital pathology (DP) and artificial intelligence (AI) in the diagnosis of prostate cancer, but little is known about the views of the public on their use. Prostate Cancer UK supporters were invited to an online survey which included quantitative and qualitative questions exploring views on the use of DP and AI in histopathological assessment. A total of 1276 responses to the survey were analysed (response rate 12.5%). Most respondents were supportive of DP (87%, 1113/1276) and of testing AI in clinical practice as a diagnostic adjunct (83%, 1058/1276). Respondents saw DP as potentially increasing workflow efficiency, facilitating research, education/training and fostering clinical discussions between clinician and patient. Some respondents raised concerns regarding data security, reliability and the need for human oversight. Among those who were unsure about AI, information was requested regarding its performance and others wanted to defer the decision to use it to an expert. Although most are in favour of its use, some are unsure, and their concerns could be addressed with more information or better communication. A small minority (<1%) are not in favour of the testing of the use of AI in histopathology for reasons which are not easily addressed.

7 October 2025 in Nature Communications
Verified
Journal Article
Open access
PDF file
FSFarzaneh Seyedshahi
Farzaneh Seyedshahi
KRKai Rakovic
Kai Rakovic
NPNicolas Poulain
Nicolas Poulain
AQAdalberto Claudio Quiros
Adalberto Claudio Quiros
IPIan R. Powley
Ian R. Powley
JQJohn Le Quesne
John Le Quesne
Mesothelioma is a highly lethal and poorly biologically understood disease which presents diagnostic challenges due to its morphological complexity. This study uses self-supervised AI (Artificial Intelligence) to map the histomorphological landscape of the disease. The resulting atlas consists of recurrent patterns identified from 3446 Hematoxylin and Eosin (H&E) stained images scanned from resected tumour slides. These patterns generate highly interpretable predictions, achieving state-of-the-art performance with 0.65 concordance index (c-index) for outcomes and 88% AUC in subtyping. Their clinical relevance is endorsed by comprehensive human pathological assessment. Furthermore, we characterise the molecular underpinnings of these diverse, meaningful, predictive patterns. Our approach both improves diagnosis and deepens our understanding of mesothelioma biology, highlighting the power of this self-learning method in clinical applications and scientific discovery.

16 October 2025 in bioRxiv
Verified
Preprint
Open access
LFLucas Farndale
Lucas Farndale
KRKai Rakovic
Kai Rakovic
IPIan R. Powley
Ian R. Powley
PHPhimmada Hatthakarnkul
Phimmada Hatthakarnkul
FSFarzaneh Seyedshahi
Farzaneh Seyedshahi
KYKe Yuan
Ke Yuan
Modern spatial proteomic methods, such as multiplex immunofluorescence (mIF) imaging, offer a data-rich view of spatial biology in intact tissues. However, interpreting its complexity is a major bottleneck, limiting its potential for biological discovery and clinical translation. Current computational methods often rely solely on segmentation-based analysis that discards crucial morphological information and are limited to testing pre-defined hypotheses. Here, we introduce a self-supervised learning (SSL) framework, HyPERSTAC, which integrates SSL and segmentation-based analysis to enable hypothesis-agnostic, context-aware discovery of biomarkers directly from mIF images. Our approach extracts rich feature representations that capture holistic architectural patterns, which integrate cellular morphology, marker interactions, and microenvironmental context without human supervision. Applying HyPERSTAC to close to 400,000 mIF tissue images from two cancer cohorts totalling over 1,800 patients, we show that SSL-derived features complement existing spatial analyses. Using HyPERSTAC to integrate SSL-derived features with segmentation-based neighbourhood analysis or the established spatial frameworks TRACERx-PHLEX and MuSpAn increased prognostic concordance in held-out patients. HyPERSTAC improves survival prediction for four SSL training methods, and two state-of-the-art foundation models. The method autonomously identified previously unknown and potentially clinically actionable biological patterns. In lung adenocarcinoma, these include a Ki67-mediated immune evasion phenotype, a subcellular pattern of GLB1 expression which aligns with low-grade EGFR -driven tumours, and distinct modes of tumour-immune interaction in PD-L1 + patients. We also find a regulatory T-cell-mediated immuno-suppressive environment promoting tumour budding in colorectal carcinoma. Our work shows that SSL can complement segmentation-based analysis, establishing HyPERSTAC as a powerful, scalable, explainable, and unbiased platform to decode tissue ecosystems without pre-defined hypotheses. This transforms high-plex imaging from a hypothesis-testing tool into a hypothesis-generating engine that can accelerate the discovery of next-generation spatial biomarkers. We release two large image datasets from LATTICe-A , creating the largest publicly available multiplex immunofluorescence patient cohort, with two panels totalling almost 6,000 images from 1,000 patients and accompanying survival data. This creates a new benchmark for the community and enables further biomarker discovery from these data.

3 December 2025 in PLOS ONE
Verified
Journal Article
Open access
PDF file
RBRachel L. Baird
Rachel L. Baird
DMDavid Mason
David Mason
KRKai Rakovic
Kai Rakovic
FBFiona Ballantyne
Fiona Ballantyne
IPIan R. Powley
Ian R. Powley
JQJohn Le Quesne
John Le Quesne
Multiplexed methods for the detection of protein expression generate extremely data-rich images of intact tissue sections. These images are invaluable for the quantification and analysis of complex biology and biomarker development. However, their interpretation presents a considerable analytical challenge. Cell segmentation from images is a key bottleneck and a major focus of research activity in artificial intelligence. Most current methods depend initially on the use of a nuclear counterstain to identify nuclear boundaries, which is a relatively straightforward task. The cellular boundary is then assigned either by expansion of the nuclear outline, or by the use of membrane or cytoplasm-specific stains to delineate cell boundaries, or by some combination of the two. The task is critical, as inaccurate segmentation leads to information loss and data contamination from neighbouring cells. Increasingly sophisticated methods are being developed to address these issues, but each has its own shortcomings. We present an alternative method which is inspired by the fact that the assignation of a cellular phenotype ‘by eye’ does not depend upon the accurate identification of cell boundaries. We present an easy-to-use deep learning-based cellular phenotyping method which leverages this human capacity to assign phenotypes without segmenting the entire cell, and which can accurately phenotype cells based on nuclear segmentation alone. Using human ground truth annotations of entire cellular regions, we developed a classifier leveraging the U-Net architecture within a commercially available deep learning image analysis platform, but the principle is transferrable to any deep-learning framework. Crucially, training requires only a single example of each compartmental stain (nuclear/cytoplasmic/membranous). The resulting algorithm assigns class identities to cells with nuclear labelling alone, without the need for whole cell expansion. The method is highly novel, broadly generalisable, and comparable in accuracy to intensity-based phenotyping methods, bridging the gap between inaccurate cellular segmentation and accurate phenotype generation.

13 January 2025 in bioRxiv
Verified
Preprint
Open access
PDF file
RPRachel Louise Pennie
Rachel Louise Pennie
DMDavid Mason
David Mason
KRKai Rakovic
Kai Rakovic
FBFiona Ballantyne
Fiona Ballantyne
IPIan R Powley
Ian R Powley
JQJohn Le Quesne
John Le Quesne
Multiplexed methods for the detection of protein expression generate extremely data-rich images of intact tissue sections. These images are invaluable for the quantification and analysis of complex biology and biomarker development. However, their interpretation presents a considerable analytical challenge. Cell segmentation from images is a key bottleneck and a major focus of research activity in artificial intelligence. Most current methods depend initially on the use of a nuclear counterstain to identify nuclear boundaries, which is a relatively straightforward task. The cellular boundary is then assigned either by expansion of the nuclear outline, or by the use of membrane or cytoplasm-specific stains to delineate cell boundaries, or by some combination of the two. The task is critical, as inaccurate segmentation leads to information loss and data contamination from neighbouring cells. Increasingly sophisticated methods are being developed to address these issues, but each has its own shortcomings. We present an alternative method which is inspired by the fact that the assignation of a cellular phenotype ‘by eye’ does not depend upon the accurate identification of cell boundaries. We present an easy-to-use deep learning-based cellular phenotyping method which leverages this human capacity to assign phenotypes without segmenting the entire cell, and which can accurately phenotype cells based on nuclear segmentation alone. Using human ground truth annotations of entire cellular regions, we developed a classifier leveraging the U-Net architecture within a commercially available deep learning image analysis platform, but the principle is transferrable to any deep-learning framework. Crucially, training requires only a single example of each compartmental stain (nuclear/cytoplasmic/membranous). The resulting algorithm assigns class identities to cells with nuclear labelling alone, without the need for whole cell expansion. The method is highly novel, broadly generalisable, and comparable in accuracy to intensity-based phenotyping methods, bridging the gap between inaccurate cellular segmentation and accurate phenotype generation.