Clinical and Technical Publications using Niramai Technology
Updated March 2026 · All peer-reviewed
Note: Soft copies may be subject to publisher copyright and are shared for academic use only – please do not redistribute.
Note: Niramai holds 40 granted patents covering thermal imaging and AI. Ideas in these publications are protected; reuse is highly discouraged.
Peer-reviewed full papers in journals
Adaptive Density-Informed AI Framework for Improving Breast Cancer Detection Using Mammography and Thermal Imaging
Kakileti ST, Govindaraju B, Sampangi S, Manjunath G. Innovative Practice in Breast Health, Apr 2026.
Multimodal AI that adapts to breast density, combining mammography with thermal imaging.
Niramai: An AI Solution to Save Lives
Koning R, Roche MP, Dey K. Harvard Business School Case 725-439, Mar 2025.
Harvard case study on Niramai’s mission, technology, and growth.
A real-world evaluation of an innovative AI tool for population-level breast cancer screening
Adapa K, Gupta A, Singh S, et al. Nature npj Digital Medicine, 8, 2 (2025).
Independent real-world evaluation of Thermalytix at the population level, published in a Nature portfolio journal.
Generative AI Approaches for Synthesizing High-Fidelity Breast Thermal Images
AIIIMA 2024 · Springer Nature LNCS.
New generative methods for producing realistic thermal images for AI training.
Thermal Radiomics for Early Detection of Diabetic Foot Ulcers Using Infrared Thermography
AIIIMA 2024 · Springer Nature LNCS.
Extending Niramai’s thermal radiomics work beyond breast – to diabetic foot ulcers.
Quantitative insights from breast cancer screening of 100,000 women in India using an AI-based tool
Preprint (Research Square), Aug 2024.
Insights from one of the largest AI-based breast cancer screening cohorts to date.
Evaluating Clinical Efficacy of Thermalytix as a Breast Cancer Screening Tool in Young Women
Journal of Young Women’s Breast Cancer and Health, 1(1&2):58-65, Dec 2024.
Efficacy of Thermalytix in younger women – a population where mammography sensitivity is limited.
Scaling Niramai: Disrupting Breast Cancer Detection Using AI
Harvard Business Publications, Jan 2023.
Business case on scaling Niramai’s AI-powered screening.
A prospective evaluation of breast thermography enhanced by ML for screening in a general population
Frontiers in Artificial Intelligence, Jan 2023.
Blinded, IEC-approved prospective comparison of Thermalytix vs mammography on 459 women at Max Hospital, Delhi.
AI-enhanced thermal breast imaging in the diagnosis of invasive breast cancer: 2 case reports
European Journal of Medical Case Reports, 2023.
Two symptomatic patients identified by Thermalytix and confirmed as IDC.
3D-BreastNet: A Self-supervised Deep Learning Network for Reconstruction of 3D Breast Surface from 2D Thermal Images
MICCAI Workshop on AI over Infrared Images for Medical Applications, 2023.
Self-supervised model for reconstructing 3D breast surface from 2D thermal images.
Feasibility and outcomes of AI-enhanced breast thermography (Thermalytix) at primary health centres in South India
International Journal of Community Medicine and Public Health, Oct 2022.
Analysis of 6,935 women screened with Thermalytix in BBMP referral hospitals.
Automated vascular analysis of breast thermograms with interpretable features
Journal of Medical Imaging (Bellingham), July 2022.
Blood vessel segmentation and classification used in Thermalytix to produce interpretable reports for radiologists.
Multicentric study evaluating Thermalytix vs standard screening modalities in subjects with suspected breast cancer
BMJ Open, 11(10):e052098, Oct 2021. doi:10.1136/bmjopen-2021-052098
Multisite prospective diagnostic-accuracy study of Thermalytix vs standard imaging in symptomatic patients (early algorithm version, 2017).
A Systematic Study to Evaluate the Benefit of using a Visualization Tool for Breast Thermography
Thermology International (American Academy of Thermology), May 2021.
SMILE-100 visualization tool with automated hotspot annotation as an aid to thermographers.
Observational Study to Evaluate Clinical Efficacy of Thermalytix in Symptomatic and Asymptomatic Women
JCO Global Oncology (ASCO), 6:1472-1480, Oct 2020.
Systematic study of 470 participants comparing Thermalytix with mammography, ultrasound, and biopsy.
Personalized risk prediction for breast cancer pre-screening using AI and thermal radiomics
Artificial Intelligence in Medicine (Elsevier), 105:101854, 2020.
Core thermal radiomics framework underlying the Thermalytix algorithm.
Estimation of Breast Cancer Incidence Risk with Incomplete or Inaccurate Information
Asian Pacific Journal of Cancer Prevention, 21(8):2307-2313, 2020.
Deep learning model computing 1-year breast cancer risk from non-imaging questionnaire data.
Exploring deep learning networks for tumour segmentation in infrared images
Quantitative InfraRed Thermography Journal, 17(3):153-168, Jul 2020.
Deep learning approach for detecting hotspots that may represent potential tumours.
Robust Estimation of Breast Cancer Incidence Risk in Presence of Incomplete or Inaccurate Information
Asian Pacific Journal of Cancer Prevention, 21(8):2307-2313, 2020.
ML-based risk assessment without thermal images.
Thermalytix – using AI to save lives
XRDS: Crossroads, ACM Magazine for Students, 26(3):38-41, 2020.
Accessible overview of Niramai Thermalytix in ACM’s student magazine.
Cascaded CNN for View-Independent Breast Segmentation in Thermal Images
IEEE EMBC, 2019.
Segmentation model used in Thermalytix for the breast ROI in thermal images.
Extraction of medically interpretable features for classification of malignancy in breast thermography
IEEE EMBC, 1062-1065, 2016.
First version of the Thermalytix algorithm; semantic features driving the ML classifier.
Automated blood vessel extraction in two-dimensional breast thermography
IEEE ICIP, 380-384, Sep 2016.
Algorithm for extracting blood-vessel structure from breast thermal images.
Automatic determination of hormone receptor status in breast cancer using thermography
MICCAI, 636-643, Oct 2016.
ML to determine hormonal status of breast cancer from thermal images alone.
Semi-automated breast cancer tumor detection with thermographic video imaging
IEEE EMBC, Milan, Aug 2015.
Early version of the algorithm using image-processing techniques.
Initial evaluation of human-supervised automated breast cancer screening using thermography
QIRT Asia, Mamallapuram, Jul 2015.
First lab-level evaluation of the ML algorithm on thermal images.
Method for classifying cancerous and normal regions in breast thermography for small-size tumors
QIRT Asia, Mamallapuram, Jul 2015.
Very early version of the classification technique.
Advances in breast thermography
InTech · New Perspectives in Breast Imaging, 2017:91. doi:10.5772/intechopen.69198
Robust Data-Driven Region-of-Interest Segmentation for Breast Thermography
Handbook of Dynamic Data Driven Applications Systems, Vol. 2, 2023.
Presentations at international conferences (abstracts & posters)
Comparing User Experience Between AI-powered Thermal Test and Mammography
Clinical Cancer Research, 32(4_Suppl):PS5-10-15, Feb 2026 (SABCS).
Single-center study comparing patient experience between Thermalytix and mammography.
AI-based Thermalytix for management of Breast Pain in Primary Care
SABCS 2024, San Antonio, Texas, Dec 2024.
The Punjab breast cancer AI-digital project: Implementing Artificial Intelligence-based triaging at scale for population-level screening of breast cancer
European Breast Cancer Conference, Milan, Italy, Mar 2024.
Field study on 15,069 participants – cancer detection rate 3-4.5x that of CBE.
Real-world evaluation of thermogram + CBE in community-based breast cancer screening
Annals of Oncology, 34:S950, Oct 2023.
Multi-modal AI imaging for improving breast cancer screening across breast densities: a pilot study
European Breast Cancer Conference, Milan, Italy, Mar 2023.
181 women – adding Thermalytix to mammography interpretation improved sensitivity by 15%.
Evaluating Clinical Performance of Thermalytix in Women with Dense Breasts
76th IRIA Conference, 2024.
Cohort study of 717 women in dense-breast population.
A prospective study comparing performance of Thermalytix with Diffusion-Weighted MRI
76th IRIA Conference, 2024.
Efficacy of Thermalytix compared to standard imaging in detecting early breast cancer (T1, T2)
3rd Indian Cancer Congress, Nov 2023.
Retrospective analysis of T1/T2 lesion detection across prior clinical trials.
Novel clinically effective, appropriate, and affordable breast cancer screening solutions for resource-constrained settings: a scoping review
3rd Indian Cancer Congress, Nov 2023.
Health Technology Assessment on Thermalytix: A Novel Breast Cancer Screening Technology
Value in Health, 25(12):S317-S318.
Analyzing the performance of Thermalytix, an AI-based breast cancer screening solution, in a community setting
SABCS 2022, San Antonio, Texas, Dec 2022.
6,935 women screened with Thermalytix across 22 BBMP hospitals.
An Automated Risk Stratification System for Breast Cancer Screening using Thermalytix
Kakileti ST, Madhu HJ, Bansal R, et al. SABCS 2022, Dec 2022.
B-Score (generated by Thermalytix) for risk-stratified screening.
Real-world evidence from a breast-screening pilot for the underprivileged: BBMP hospitals
European Breast Cancer Conference, Barcelona, Spain, Nov 2022.
6,935 women screened with Thermalytix in 22 BBMP hospitals.
MaThAI: A multimodal imaging modality combining Mammography and Thermalytix
European Breast Cancer Conference, Barcelona, Spain, Nov 2022.
Multimodal approach using B-Score to prioritize mammography scans for early malignancy detection.
Evaluating screening performance of Thermalytix by comparing breast lesion sizes with mammography
UK Imaging and Oncology (UKIO) 2022, Liverpool, UK.
Post-hoc analysis of 470 subjects – Thermalytix correlated with mammography for small lesion detection.
Performance of AI-based breast cancer screening in a community setting: a real-world evaluation study
The Lancet Oncology, 23:S20, Jul 2022.
Funnel analysis of 13,933 women screened with Thermalytix in community settings.
AI for early-stage breast cancer screening
ASCO Breakthrough Summit, Bangkok, Oct 2019. Top-50 presentation.
Observational study on 769 participants vs mammography, ultrasound, and biopsy.
A prospective study of an AI-based breast cancer screening solution for resource-constrained settings
ASCO Annual Meeting 2021.
Thermalytix vs mammography in 459 participants.
Prospective evaluation of an affordable, portable AI modality for breast cancer screening across breast densities
2022 ASCO Annual Meeting.
Thermal Radiomics for Detection of Axillary Lymph Node Metastases: A Pilot Study
Global Breast Cancer Conference (GBCC 10), Seoul, Apr 2021.
Correlation of Thermalytix in detecting a breast lesion as benign, malignant, or normal
SABCS 2020, San Antonio, Texas, Dec 2020 (PS2-44).
Pilot comparing B-Score with BI-RADS and histopathology.
A Non-Radiation Based Screening to Detect Dense Breasts
SABCS 2019, San Antonio, Texas, Dec 2019.
Detecting breast density from thermal images alone.
AI over thermal images for radiation-free breast cancer screening
SABCS 2018, Dec 2018.
First poster comparing Thermalytix with mammography on 247 participants – benefit on BI-RADS 0 mammograms.
Machine Learning over Thermal Images for Accurate Breast Cancer Screening
Int. Conf. on Advances in Breast Cancer Treatments, Kyoto, Japan, Apr 2018.
AI-based Thermalytix vs manual thermography.
Thermalytix: An Advanced AI-Based Solution for Non-Contact Breast Screening
Int. Conf. on Breast Cancer Management, Amsterdam, Feb 2018.
First pilot comparing Thermalytix with mammography on 147 participants.