You are here

  • Home
  • Online First
  • Lymphoma triage from H&E using AI for improved clinical management

Article Text

Article menu
Download PDFPDF
Original research
Lymphoma triage from H&E using AI for improved clinical management
  1. Anna Maria Tsakiroglou1,
  2. Chris M Bacon2,
  3. Daniel Shingleton3,
  4. Gabrielle Slavin4,
  5. Prokopios Vogiatzis5,
  6. Richard Byers1,6,
  7. Christopher Carey2,7,
  8. Martin Fergie1,8
  1. 1Spotlight Pathology Ltd, Manchester, UK
  2. 2Translational and Clinical Research Institute, Newcastle University, Newcastle upon Tyne, UK
  3. 3Division of Laboratory Medicine, Manchester University NHS Foundation Trust, Manchester, UK
  4. 4Department of Cellular Pathology, Manchester University NHS Foundation Trust, Manchester, UK
  5. 5Cellular Pathology Department, Southend University Hospital, Mid & South Essex NHS Trust, Essex, UK
  6. 6Department of Histopathology, Manchester University NHS Foundation Trust, Manchester, UK
  7. 7Haematology, Newcastle Upon Tyne Hospitals NHS Foundation Trust, Newcastle Upon Tyne, UK
  8. 8School of Health Sciences, Faculty of Medicine, Biology and Health, The University of Manchester, Manchester, UK
  1. Correspondence to Dr Martin Fergie, School of Health Sciences, Faculty of Medicine, Biology and Health, The University of Manchester, Manchester, UK; martin.fergie{at}spotlightpathology.com

Abstract

Aims In routine diagnosis of lymphoma, initial non-specialist triage is carried out when the sample is biopsied to determine if referral to specialised haematopathology services is needed. This places a heavy burden on pathology services, causes delays and often results in over-referral of benign cases. We aimed to develop an automated triage system using artificial intelligence (AI) to enable more accurate and rapid referral of cases, thereby addressing these issues.

Methods A retrospective dataset of H&E-stained whole slide images (WSI) of lymph nodes was taken from Newcastle University Hospital (302 cases) and Manchester Royal Infirmary Hospital (339 cases) with approximately equal representation of the 3 most prevalent lymphoma subtypes: follicular lymphoma, diffuse large B-cell and classic Hodgkin’s lymphoma, as well as reactive controls. A subset (80%) of the data was used for training, a further validation subset (10%) for model selection and a final non-overlapping test subset (10%) for clinical evaluation.

Results AI triage achieved multiclass accuracy of 0.828±0.041 and overall accuracy of 0.932±0.024 when discriminating between reactive and malignant cases. Its ability to detect lymphoma was equivalent to that of two haematopathologists (0.925, 0.950) and higher than a non-specialist pathologist (0.75) repeating the same task. To aid explainability, the AI tool also provides uncertainty estimation and attention heatmaps.

Conclusions Automated triage using AI holds great promise in contributing to the accurate and timely diagnosis of lymphoma, ultimately benefiting patient care and outcomes.

  • LYMPHOMA
  • Image Processing, Computer-Assisted
  • Machine Learning
  • Histology

Data availability statement

No data are available. The software used in the current study is Spotlight Pathology proprietary IP and cannot be shared. The Manchester and Newcastle datasets cannot be shared. The CAMELYON 17 dataset is publicly available from https://camelyon17.grand-challenge.org/.

https://doi.org/10.1136/jcp-2023-209186

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Data availability statement

    No data are available. The software used in the current study is Spotlight Pathology proprietary IP and cannot be shared. The Manchester and Newcastle datasets cannot be shared. The CAMELYON 17 dataset is publicly available from https://camelyon17.grand-challenge.org/.

    View Full Text

    Footnotes

    • Handling editor Runjan Chetty.

    • RB, CC and MF contributed equally.

    • Contributors MF, RB, AMT and CC were responsible for study conception and design. MF, RB, DS, GS and PV acquired data. AMT analysed data. AMT and MF were responsible for AI algorithm design and implementation. AMT, MF, RB, GS, PV, CC and CMB interpreted data. AMT drafted the manuscript. All the authors reviewed manuscript and gave final approval. MF, RB and CC contributed equally. MF is responsible for the overall content as the guarantor.

    • Funding This study was funded by Spotlight Pathology.

    • Competing interests AMT is an employee of Spotlight Pathology; MF is CEO and cofounder of Spotlight Pa-thology; RB is CMO and cofounder of Spotlight Pathology.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.