Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Machine learning-enabled label-free SERS for microbial sensing: Toward robust, generalizable, and deployable workflows.

Talanta·2026
Same author

Machine learning algorithms in the estimation of sex from 3DCT-generated cranial and pelvic measurements.

International journal of legal medicine·2026
Same author

Epidemiology and Prevention of Lung Cancer in Europe: A Review of Burden, Disparities, and the Path to Equitable Early Detection.

Acta medica academica·2026
Same author

The prognostic and decision-making impact of KELIM CA-125 in patients with advanced ovarian cancer: A real-world study from Bosnia and Herzegovina.

European journal of obstetrics, gynecology, and reproductive biology·2026
Same author

Machine Learning-Based Classification of Gliomas and Tumor Grades with SHAP-Guided Feature Interpretation.

Genes·2026
Same author

C-reactive protein-to-lymphocyte ratio and perforation in pediatric appendicitis.

Surgery today·2026

Related Experiment Video

Updated: May 24, 2025

Semi-automatic PD-L1 Characterization and Enumeration of Circulating Tumor Cells from Non-small Cell Lung Cancer Patients by Immunofluorescence
10:29

Semi-automatic PD-L1 Characterization and Enumeration of Circulating Tumor Cells from Non-small Cell Lung Cancer Patients by Immunofluorescence

Published on: August 14, 2019

10.5K

A novel deep learning framework for automatic scoring of PD-L1 expression in non-small cell lung cancer.

Saidul Kabir1, Muhammad E H Chowdhury2, Rusab Sarmun1

  • 1Department of Electrical and Electronic Engineering, University of Dhaka, Dhaka, Bangladesh.

Biomolecules & Biomedicine
|March 4, 2025
PubMed
Summary

This study introduces an automated deep learning framework to precisely evaluate programmed death-ligand 1 (PD-L1) expression in non-small cell lung cancer (NSCLC) using whole slide images. The AI model accurately predicts Tumor Proportion Score (TPS), improving immunotherapy eligibility assessment.

More Related Videos

Using 22C3 Anti-PD-L1 Antibody Concentrate on Biopsy and Cytology Samples from Non-small Cell Lung Cancer Patients
07:43

Using 22C3 Anti-PD-L1 Antibody Concentrate on Biopsy and Cytology Samples from Non-small Cell Lung Cancer Patients

Published on: September 25, 2018

10.5K
Industrialized, Artificial Intelligence-guided Laser Microdissection for Microscaled Proteomic Analysis of the Tumor Microenvironment
13:01

Industrialized, Artificial Intelligence-guided Laser Microdissection for Microscaled Proteomic Analysis of the Tumor Microenvironment

Published on: June 3, 2022

3.6K

Related Experiment Videos

Last Updated: May 24, 2025

Semi-automatic PD-L1 Characterization and Enumeration of Circulating Tumor Cells from Non-small Cell Lung Cancer Patients by Immunofluorescence
10:29

Semi-automatic PD-L1 Characterization and Enumeration of Circulating Tumor Cells from Non-small Cell Lung Cancer Patients by Immunofluorescence

Published on: August 14, 2019

10.5K
Using 22C3 Anti-PD-L1 Antibody Concentrate on Biopsy and Cytology Samples from Non-small Cell Lung Cancer Patients
07:43

Using 22C3 Anti-PD-L1 Antibody Concentrate on Biopsy and Cytology Samples from Non-small Cell Lung Cancer Patients

Published on: September 25, 2018

10.5K
Industrialized, Artificial Intelligence-guided Laser Microdissection for Microscaled Proteomic Analysis of the Tumor Microenvironment
13:01

Industrialized, Artificial Intelligence-guided Laser Microdissection for Microscaled Proteomic Analysis of the Tumor Microenvironment

Published on: June 3, 2022

3.6K

Area of Science:

  • Computational pathology
  • Artificial intelligence in oncology
  • Biomarker quantification

Background:

  • Programmed death-ligand 1 (PD-L1) expression is a critical predictive marker for anti-PD-1/PD-L1 immunotherapy in non-small cell lung cancer (NSCLC).
  • Current assessment of PD-L1 expression via immunohistochemistry (IHC) and Tumor Proportion Score (TPS) evaluation by pathologists can be subjective and time-consuming.
  • Automating TPS evaluation is crucial for enhancing accuracy, consistency, and efficiency in determining patient eligibility for immunotherapy.

Purpose of the Study:

  • To develop and validate a novel automated deep learning framework for accurate PD-L1 expression and TPS evaluation in NSCLC whole slide images (WSIs).
  • To improve the precision and consistency of TPS assessment, thereby optimizing patient selection for immunotherapy.
  • To reduce the workload on pathologists and potentially lower the cost of biomarker assessment.

Main Methods:

  • A deep learning framework was developed, involving tumor patch identification, tumor area segmentation, and cell nuclei detection using models like EfficientNet, Inception, Vision Transformer, UNet, DeepLabV3+, and StarDist.
  • A hybrid human-machine approach was used for annotating a dataset of 66 NSCLC WSIs.
  • The framework estimated TPS based on the ratio of positively stained to total viable tumor cells, categorizing PD-L1 expression into negative (<1%), low (1-49%), and high (≥50%) levels.

Main Results:

  • The Vision Transformer model achieved a high F1-score of 97.54% for classification, while the modified DeepLabV3+ model attained a Dice Similarity Coefficient of 83.47% for segmentation.
  • The framework's predicted TPS showed a strong correlation (0.9635) with pathologist assessments.
  • The automated three-level PD-L1 expression classification achieved an F1-score of 93.89%.

Conclusions:

  • The proposed deep learning framework demonstrates promising performance for automated PD-L1 TPS evaluation in NSCLC.
  • This automated approach has the potential to provide clinically significant results more efficiently and cost-effectively.
  • The framework offers a viable tool to enhance the accuracy and consistency of PD-L1 biomarker assessment for immunotherapy selection.