Tumor heterogeneity is a key reason for the failure of therapeutic approaches and the failure to control patients' disease. Because of the random nature of cancer gene mutations, the pathogenesis of each patient is not exactly the same, so whether a relatively single traditional anti-cancer therapy can match the patient's condition becomes an indeterminate probability problem. With deeper understanding of cancer development mechanisms and the development of artificial intelligence, precision medicine is increasingly likely to help people solve the problem of tumor heterogeneity. We propose specific solutions for precision medicine from several perspectives. First, we extract information from medical images of patients and determine tumor types through Radiomics technology, then use genetic analysis technology to analyze exosomes released from tumor cells contained in patients' blood to obtain tumor tissue trends, and use tumor tissues obtained from patients' puncture biopsies to obtain immune infiltration of tumor tissues. Through these three perspectives, we can obtain information about the tumor itself, Including molecular typing of tumors and prognosis of patients after neoadjuvant chemotherapy, as well as its future trend and the body's response to the tumor tissue. Through the above three aspects, we can analyze the tumor situation to determine the most suitable treatment plan for the patient.
Keywords: Radiogenomics, Pattern classification, Tumour detection and diagnosis, MRI