Nuclear Magnetic Resonance in Metabolism and Clinical Medicine: Part 2: Present Scenario of Artificial Intelligence on NMR-Biochemical Correlation in Theranosis

Abstract

Combined efforts on Magnetic Resonance Imaging and Spectroscopy or NMR-biochemical correlation concept has achieved as established diagnostic tool in medical practice in clinical setting. Reviews and meta-analyses also indicate the great possibility of integrated multimodal multiparametric MRI and MRS using parallel receiver array coils in ultrahigh magnets  suitable for soft tissue imaging in decision making and monitoring disease progress towards personalized medicine. Recent guidelines suggest the urgent need of artificial intelligence for deep learning of brain, breast, prostate, liver, heart tissue biophysical and biochemical nature using digital ‘spectromics’ analysis along with other molecular imaging modalities. The fact of poor understanding of artificial intelligence and its use without validation poses several questions. The present opinion on high resolution MRI and MRS is shared on theranostic biomarkers of different soft tissues with recent reviews, government efforts and available literature on evidence based MR ‘spectromics’, artificial intelligence algorithms, clinical trials and meta-analyses to illustrate the MRI with MR spectroscopy as adjunct real-time, robust, fast tissue digital images with metabolic screening features to evaluate the patient disease and theranosis with available medical practices in clinical setting and their limitations.