PET/MRI Imaging: Concepts, Deep Learning, Clinical Trials and Applications in Clinical Imaging

Abstract

Cancers and events of malignancy are rampant. The increased incidence and cancer spread indicates it as public health concern.  Initially, tumor assessment and risk evaluation of cancer spread is done based on evidence-based histology, immunostaining and microimaging.  Now real-time, dynamic tumor physiochemical in vivo noninvasive smart deep learning based screening of tumor grades and decision-making based on imaging options are slowing to replace the long wait of immunostaining and needle biopsy reports in decision taking.  The PET/MRI is a new emerging hybrid imaging tool used in clinical settings for oncologic metastasis nodule imaging and assessment of cancer spread to other soft tissues and organs. The latest developments in deep learning for PET/MRI value in breast, prostate, colon, brain cancer oncologic applications show rapid interest among physicians for anatomical and functional information to make decision of chemo- or surgery intervention. The present paper shows the challenges with solutions and potentials of new innovations in hybrid PET/MRI vs PET/CT techniques using nano-carrier contrast agents in cancer tumor classification by artificial intelligence characterization of different tumor grades.