Artificial neural network in predicting cancer based on infrared spectroscopy
In this work, we present a Real-Time (RT), on-site, machine-learning-based methodology for identifying human cancers. The presented approach is reliable, effective, cost-effective, and non-invasive method, which is based on Fourier Transform Infrared (FTIR) spectroscopy—a vibrational method with the ability to detect changes as a result of molecular vibration bonds using Infrared (IR) radiation in human tissues and cells. Medical IR Optical System (IROS) is a tabletop device for real-time tissue diagnosis that utilizes FTIR spectroscopy and the Attenuated Total Reflectance (ATR) principle to accurately diagnose the tissue. The combined device and method were used for RT diagnosis and characterization of normal and pathological tissues ex vivo/in vitro. The solution methodology is to apply Machine Learning (ML) classifier that can be used to differentiate between cancer, normal, and other pathologies. Excellent results were achieved by applying feedforward backpropagation Artificial Neural Network (ANN) with supervised learning classification on 76 wet samples. ANN method shows a high performance to classify; overall, 98.7% (75/76 biopsies) of the predictions are correctly classified and 1.3% (1/76 biopsies) is wrong classification.