View Project

(i) A fast, non-invasive, real-time, label-free, and early diagnosis platform: Computer-Aided Diagnosis (CAD)-based detection technique utilizing Surface Enhanced Raman Spectroscopy (SERS) to differentiate the prevalence of three cancers: Breast, Lung, and Larynx.

(ii) Optimize the parameter for SERS substrate and spectral analysis to acquire spectra from patient blood plasma samples for differentiating the three types of cancer.

(iii) Establish a specially modified Artificial Intelligence (AI) based mathematical model with multilevel deep learning classifiers and multivariate data analysis to differentiate the SERS spectra based on the morphological changes of nucleic acids, carbohydrates, proteins, and lipids in blood samples.

The proposed system involves Surface-Enhanced Raman Spectroscopy-Embedded Artificial Intelligence (SERS-EAI), capable of identifying molecular fingerprints in body fluids, cells, and tissues. Limited work has been carried out in India for early cancer diagnosis using SERS in clinical scenarios. To address this gap, CSIR-National Institute for Interdisciplinary Science and Technology (NIIST) in collaboration with Olusium Technologies in Advanced Research (Olusium) initiated the development of a clinical diagnostic platform based on SERS and functionalized nanoparticle substrate, coupled with Embedded Artificial Intelligence. The pilot study was conducted in the first phase, and we observed the potential of this technology for ultrasensitive detection of various human cancer biomarkers, including breast, lung, head, and neck cancers along with healthy volunteers, with an overall sensitivity and specificity of 91.5% and 91%, achieving an overall accuracy of 91% based on different models.

The project aims (the second phase of the current project) for large-scale validation of this method with a larger number of samples (an estimated average of 200 samples per cancer class) collected at different geographic locations (multicentric) over three years. This validation will involve label-free SERS analysis of blood plasma or serum for the detection and classification of various cancers, utilizing the proprietary embedded AI technology. The goal is to implement the feasibility of this method for the early detection of prevalent cancers, including but not limited to Breast cancer, Head & neck cancer, and Lung cancer in India through a multi-centric clinical trial.

New Project

A validated Al algorithm utilizing Raman spectra (SERS) will be developed for early and accurate diagnosis of multiple cancers.

• Global classification i.e. the percentage (%) of the prevalence of Cancer vs Healthy.

• Individual classification of cancer samples i.e. percentage (%) of the prevalence of individual cancer category

• The whole diagnostic classification test will be performed within 30 minutes.

• Establish an AI-based hardware chip a well-classified platform that will be integrated into the Raman spectrometer for obtaining end readout of the cancer classification

• Documents for this diagnostic technique will be ready for submission to the Central Drugs Standard Control Organization (CDSCO) for approval.

• The SERS-AI technology is less invasive, less time-consuming, and built on a portable architecture, making it accessible and easy to use in remote and underserved areas. This technology can also reduce the burden on healthcare systems and providers by streamlining the cancer detection process and improving the efficiency of diagnosis, and it benefits the citizens w.r.t., their socio-economic status

• The development of this computer-aided diagnosis technology for early cancer detection can have a profound impact on society by improving the health outcomes and quality of life for cancer patients, reducing healthcare costs, and potentially saving lives.

• ln terms of future healthcare technologies and innovations, the focus of healthcare shifts from treating illness to sustaining wellness.

Partnership with healthcare Industry