Abstract:BACKGROUND AND OBJECTIVE: This paper aims to overview multidimensional mining algorithms in relation to Magnetic Resonance Imaging (MRI) radiogenomics for computer aided detection and diagnosis of breast tumours. The work also aims to address a new problem in radiogenomics mining: how to combine structural radiomics information with non-structural genomics information for improving the accuracy and efficacy of Neoadjuvant Chemotherapy (NAC).METHODS: This requires the automated extraction of parameters from non-structural breast radiomics data, and finding feature vectors with diagnostic value, which then are combined with genomics data. In order to address the problem of weakly labelled tumour images, a Generative Adiversarial Networks (GAN) based deep learning strategy is proposed for the classification of tumour types; this has significant potential for providing accurate real-time identification of tumorous regions from MRI scans. In order to efficiently integrate in a deep learning framework different features from radiogenomics datasets at multiple spatio-temporal resolutions, pyramid structured and multi-scale densely connected U-Nets are proposed. A bidirectional gated recurrent unit (BiGRU) combined with an attention based deep learning approach is also proposed.RESULTS: The aim is to accurately predict NAC responses by combining imaging and genomic datasets. The approaches discussed incorporate some of the latest developments in of current signal processing and artificial intelligence and have significant potential in advancing and provide a development platform for future cutting-edge biomedical radiogenomics analysis.CONCLUSIONS: The association of genotypic and phenotypic features is at the core of the emergent field of Precision Medicine. It makes use of advances in biomedical big data analysis, which enables the correlation between disease-associated phenotypic characteristics, genetics polymorphism and gene activation to be revealed.

Non-invasive cancer diagnosis theories and technologies based on nuclear magnetic resonance

Nuclear magnetic resonance spectroscopy analysis and cancer feature extraction technologies

Analysis Of Nuclear Magnetic Resonance Spectroscopy Based On Wavelet Transform And Pattern Recognition

Data Processing Technology of Nuclear Magnetic Resonance for Cancer Clinical Diagnosis

Application and Progress of Magnetic Resonance Imaging in Diagnosis of Pediatric Tumors

Current Applications and Future Development of Magnetic Resonance Fingerprinting in Diagnosis, Characterization, and Response Monitoring in Cancer

Wavelet-based Denoised and Feature Extraction of NMR Spectroscopy Based on Pattern Recognition

NMF in MR Spectroscopy

Radiolabeled PET/MRI Nanoparticles for Tumor Imaging

Quantitative Magnetic Resonance Imaging Biomarkers in Oncological Clinical Trials: Current Techniques and Standardization Challenges

Feature Analysis and De-noising of MRS Data Based on Pattern Recognition and Wavelet Transform

Optical – Magnetic probe for evaluating cancer therapy

Frontier Technology The Development Status of Nano-magnetic Induction Therapy of Tumor

Structural Defect‐Enabled Magnetic Neutrality Nanoprobes for Ultra‐High‐Field Magnetic Resonance Imaging of Isolated Tumor Cells In Vivo

An Update on MR Spectroscopy in Cancer Management: Advances in Instrumentation, Acquisition, and Analysis

MRI Radiogenomics in Precision Oncology: New Diagnosis and Treatment Method

Magnetic Resonance Imaging (MRI) and MR Spectroscopic Methods in Understanding Breast Cancer Biology and Metabolism

Magnetic Resonance-Guided Cancer Therapy Radiomics and Machine Learning Models for Response Prediction

MRI radiogenomics for intelligent diagnosis of breast tumors and accurate prediction of neoadjuvant chemotherapy responses-a review

Magnetic Resonance Spectroscopy for Cervical Cancer: Review and Potential Prognostic Applications

Magnetic Resonance Imaging: Progresses and Perspective