Machine Learning Driven Alloy Property Analysis for Optimized Bio Medical Applications

Within the context of the present era, biomaterials are of utmost importance. In recent decades, breakthroughs in medical and material processing fields have led to the development of numerous biomaterials with features that suits to wide variety of applications. The Biomaterials are mostly utilized in orthopaedic treatment, Odontology, delivering drugs, tissue engineering, and cardiac devices. A range of implant materials has been utilized throughout different fields of medical sciences that is based on the specific needs of each application. The Titanium based alloy combinations pulls many of the researcher's attention mainly because of its bio compatible properties that suits for various bio implantation applications. Hence, in this research paper, the Ti-Al-Nb-Mo alloy is considered and their properties are analysed using the machine learning based classifier and optimization can be done with the well-known efficient Particle-Swarm Optimization (PSO) algorithm. The suggested alloy is subjected to analyse the mechanical properties using the X-ray diffraction analysis (XRD) and Transmission Electron Microscopy (TEM) tests and then the model is trained with the Random Forest classifier that analyses the physical properties such as young's modulus and tensile strength of the proposed alloy with various ratios and then their features are optimized for better accuracy with proposed PSO. The results demonstrates that the proposed machine learning model with the integration of RF and PSO shows the better prediction accuracy of alloy properties that in turn reduces the complicated experimental testing and finds the suitable ratio of proposed alloy for the specific bio implant applications.