Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Precision lung cancer classification using convolutional neural networks with enhanced image pre-processing and model

Subrata Sinha¹, Saurav Mali¹, Sanchaita Rajkhowa²

¹Department of Computational Sciences, Brainware University, Kolkata, West Bengal, 700125, India.

Scientific Reports

|July 6, 2026

Summary

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

Integrative Transcriptomic Analysis Identifies Hypoxia-responsive Cell Cycle Hub Genes as Prognostic Markers in Glioblastoma.

Annals of neurosciences·2026

Same author

Integrative subtractive genomics and molecular dynamics-based approach for drug repurposing against female genital tuberculosis.

Molecular diversity·2026

Same author

Frequency and Distribution of Lymphomas in Northwestern India: A Retrospective Analysis of 923 Cases Using the Latest World Health Organization Classification 5<sup>th</sup> Edition.

Indian journal of hematology & blood transfusion : an official journal of Indian Society of Hematology and Blood Transfusion·2026

Same author

Precision targeting of BTK in chronic lymphocytic leukemia: computational insights into structural dynamics and the influence of mutations on BTK inhibitors through QM and MM studies.

Journal of biomolecular structure & dynamics·2026

Same author

Effectiveness of BBV152 vaccine against SARS-CoV-2 infections, hospitalizations and deaths among healthcare workers in the setting of high delta variant transmission in New Delhi, India.

The National medical journal of India·2026

Same author

Integrative Metabolomic, Network Pharmacology, and Experimental Evidence for <i>Lepidium sativum</i> Seed Extract as a Natural Modulator of Pulmonary Fibrosis via the <i>ncNRFR/Let-7d</i> Regulatory Pathway.

Pharmaceuticals (Basel, Switzerland)·2025

This summary is machine-generated.

This study developed a Convolutional Neural Network (CNN) model for lung cancer subtype classification using histopathology images. The AI model achieved 98.59% accuracy, overcoming staining variability for improved diagnosis.

Area of Science:

Computational pathology
Artificial intelligence in oncology
Digital pathology

Background:

Lung cancer is a leading cause of global mortality, demanding accurate and timely diagnosis.
Traditional histopathological analysis of Hematoxylin and Eosin (H&E)-stained slides is the gold standard but faces challenges due to staining variability.
Variability in staining techniques can hinder manual interpretation and impact diagnostic consistency.

Purpose of the Study:

To design and develop a Convolutional Neural Network (CNN) model for classifying lung carcinoma subtypes.
To address diagnostic challenges posed by staining variability in histopathological images.
To improve the robustness and accuracy of lung cancer diagnosis through automated analysis.

Main Methods:

Keywords:

Convolutional neural network Lung cancer detection Machine learning Particle swarm optimization Reinhard color normalisation

Related Experiment Videos

Utilized the publicly available LC25000 dataset of lung histopathological images.

Applied Reinhard color normalization and Gaussian filtering for pre-processing to mitigate staining variability and noise.

Employed Particle Swarm Optimization (PSO) for hyperparameter tuning to develop a multi-scale CNN architecture.

Main Results:

The optimized CNN model achieved a high classification accuracy of 98.59% across three categories (two non-small cell lung carcinoma subtypes and one benign class).
Pre-processing techniques, particularly color normalization, significantly enhanced classification consistency and accuracy.
The developed model demonstrated strong diagnostic performance and improved resilience to variations in staining.

Conclusions:

The developed CNN model offers a robust and accurate solution for classifying lung carcinoma subtypes.
Color normalization techniques are crucial for improving the reliability of automated histopathological image analysis.
The model's successful deployment as an Android application facilitates potential real-time clinical application.