Since the early 1980s when MRI imaging technology was put into clinical use, the number of MRI clinical tests has steadily increased by more than 10% every year. At the same time, exogenous MRI contrast agents have also been developed with the development of MRI technology. However, there are still challenges in the preparation of contrast agents for magnetic resonance imaging, such as how to prepare high-efficiency contrast agents with high stability and low biological toxicity. In order to study the contrast agent with simple preparation method, low cost, and good imaging effect, a magnetic resonance contrast agent was prepared by magnetic nanoparticles. By acting on magnetic resonance imaging detection method, and using polymer ligands to synthesize magnetic nanoparticles, experiments and tests of P(MA-alt-VAc) polymer ligand-modified magnetic nanoparticles were carried out. The experimental results showed that when nanoparticles containing different iron ion concentrations were incubated with DC 2.4 normal cells for 48 hours, the cell viability was still higher than 80% at concentrations up to 200 μm. It shows that the nanoparticle has high cell activity and good biological adaptability. The transverse relaxation ( r 2 ) value of the nanoparticles in aqueous solution at 37°C and 1.5 T magnetic field is 231.1 m-1 s-1, which is much higher than that of PTMP-PMAA ( r 2 = 35.1  mM-1 s-1), which is also more than five times the relaxation of SHU-555C ( r 2 = 44  mM-1 s-1). It shows that the nanoparticles prepared in this paper have good effect and can be used as a contrast agent in human brain for magnetic resonance imaging.

Bone age is an important metric to monitor children’s skeleton development in pediatrics. As the development of deep learning DL-based bone age prediction methods have achieved great success. However, it also faces the issue of huge computation overhead in deep features learning. Aiming at this problem, this paper proposes a new DL-based bone age assessment method based on the Tanner-Whitehouse method. This method extracts limited and useful regions for feature learning, then utilizes deep convolution layers to learn representative features in these interesting regions. Finally, to realize the fast computation speed and feature interaction, this paper proposes to use an extreme learning machine algorithm as the basic architecture in the final bone age assessment study. Experiments based on publicly available data validate the feasibility and effectiveness of the proposed method.