Home About Us Praxis — Internships Studio — Hackathons Campus — Schools Healthcare Events & Milestones Courses Contact
Healthcare AI · Clinical Research

AI that sees what
the eye cannot.

Deep-learning research for early laryngeal cancer detection from narrow-band imaging — going beyond binary diagnosis to eight clinically distinct classes. Published in Springer, 2026, and covered across national media.

Read the paper →
8
Classes classified
2026
Springer publication
CE-NBI
Imaging modality
The key innovation

Beyond binary. Eight clinical classes.

Where existing systems classify tissue as merely benign or malignant, our model distinguishes eight distinct laryngeal conditions — enabling differential diagnosis at a granularity not possible before.

● Squamous Cell Carcinoma

The most aggressive malignant class.

● Carcinoma in Situ

Non-invasive carcinoma — a critical early-intervention target.

● Dysplasia Grade I

Mild dysplasia — the earliest precancerous change.

● Dysplasia Grade II

Moderate dysplasia — intermediate risk, needs monitoring.

● Dysplasia Grade III

Severe dysplasia — a high-risk precancerous condition.

● Papillomatosis

Benign laryngeal papilloma — viral aetiology, recurrent.

● Keratosis

Benign keratotic lesion — distinguished from early dysplasia.

● Healthy

Normal laryngeal mucosa — the baseline reference class.

Springer publication · IDSCS 2025

The paper.

Springer · LNNS Vol. 1945 · 26 May 2026

A Multiclass Classifier for the Differential Diagnosis of Laryngeal Carcinoma from Narrow-Band Imaging

Puneet Misra Mohd Usman Siddharth Chaurasia
DOI: 10.1007/978-3-032-24075-0_12 pp 124–136 ISBN: 978-3-032-24075-0

Abstract

Accurately distinguishing between laryngeal diseases is a difficult undertaking that frequently leads to diagnostic mistakes due to expert subjectivity. Most existing automated systems are limited to a binary classification of benign versus malignant tissue.

This study introduces a deep-learning framework performing fine-grained classification of Contact Endoscopy–Narrow Band Imaging (CE-NBI) into eight distinct classes, including squamous cell carcinoma, various grades of dysplasia, and papillomatosis. A U-shaped network augmented with a specialised attention block selectively focuses on the most discriminative features within endoscopic images.

The approach outperforms state-of-the-art techniques, and its efficient design makes it compatible with low-memory hardware — enabling deployment in district hospitals and resource-constrained clinical settings across India.

Laryngeal CancerContact Endoscopy CNNChannel Attention Spatial AttentionU-Net CE-NBISCC
Publication

Book: Data Science and Security
Series: LNNS, Vol. 1945
Publisher: Springer, Cham
Date: 26 May 2026

Clinical partner

Dr. Rakesh Srivastava, Chief Consultant, ENT Dept., Health City Vistaar Hospital, Gomtinagar.

Media coverage · 2026

National recognition.

Covered across leading English and Hindi publications.

Times of India · English

AI to feature in university course offerings

The research featured amid a university AI-in-education push, with new AI programmes launching from the 2026–27 session.

Dainik Jagran · Hindi

गले के कैंसर की जल्द पहचान के लिए एआई मॉडल

An early laryngeal-cancer detection AI developed by the research team's computer-science group.

Hindustan · Hindi

एआई मॉडल से गले के कैंसर की पहचान

Highlighted the clinical promise of NBI-based AI for early, biopsy-free diagnosis and treatment guidance.

Amrit Vichar · Hindi

नैरो बैंड इमेजिंग से गले के कैंसर की पहचान

Narrow-band imaging methodology covered in detail — the model identifies eight distinct laryngeal conditions.

Amar Ujala · Hindi

गले के कैंसर की पहचान करेगा नया एआई मॉडल

Coverage of clinical applicability — enabling timely treatment through precise early-stage identification.

Amar Ujala · Campus

हर छात्र को मिलेगा मेंटर

Coverage of academic mentorship programmes and PrajniX Labs' research partnerships.

Research capabilities

What we can do next.

Current capabilities available for institutional and clinical collaboration.

🔬

Endoscopic image classification

Attention-augmented U-Net validated on CE-NBI data, extensible to colonoscopy and bronchoscopy imaging.

🧬

Multi-class pathology detection

Fine-grained frameworks enabling differential diagnosis — beyond the binary benign/malignant limit of prior systems.

📋

Clinical NLP & report analysis

NLP over radiology reports and clinical records — extracting structured insight for decision support.

💾

Low-resource deployment

Architectures for low-memory clinical hardware — putting advanced diagnostics within reach of district hospitals.

Collaborate

Interested in this research?

We're seeking clinical collaborators, hospital partnerships, and research co-authors for multi-cancer screening and real-time surgical guidance.

View Publication ↗