The present invention provides a nano-sliver-fatty acid (nano-Ag-FA) impregnated hydrophobic porous paper-based substrate for surface-enhanced Raman scattering  (SERS). The nano-Ag-FA impregnated paper as SERS substrate enhances the sensitivity and specificity of Raman scattering enabling precise identification of  characteristic metabolic Raman peaks indicative of cancer. Further, the integration of the 1D CNN-LSTM model based system enables automated analysis and  generation of prevalence scores based on Raman spectral patterns, streamlining the diagnostic process. The invention is able to differentiate between non-cancerous  individuals and those with prevalent cancer classes i.e., breast, lung, or larynx cancers by leveraging the unique acquisition of enhanced Raman spectral patterns from  plasma samples obtained from blood, offering a non-invasive and efficient means of cancer detection based on Nano-Ag-FA-coated paper substrates. Through  meticulous validation of well-resolved Raman peaks, particularly focusing on amino acids-associated proteins, lipids, nucleic acids, and DNAs, the substrate ensures  accurate discrimination between cancerous and non-cancerous samples.