Page 23 - ESAM-1-4

Engineering Science in

Additive Manufacturing

REVIEW ARTICLE
Machine learning-driven additive manufacturing

of biomedical metals: A review of forward
prediction, inverse optimization, and quality

control

1
3
Yi Mao 1 , Deyu Jiang 1 , Uglov Vladimir 2 , Zhou Jing *, and Liqiang Wang *
¹State Key Laboratory of Metal Matrix Composites, School of Material Science and Engineering,
Shanghai Jiao Tong University, Shanghai, China
²Laboratory of NanoElectroMagnetics, Institute for Nuclear Problems, Belarusian State University,
Minsk, Belarus
³Department of Anatomy, Youjiang Medical University for Nationalities, Baise, Guangxi, China

Abstract

Additive manufacturing (AM) for biomedical metals presents revolutionary
opportunities for producing personalized, complex structured biomedical
components. However, the high nonlinearity and complexity of the manufacturing
*Corresponding authors:
Liqiang Wang process pose significant challenges to the performance consistency of biomedical
(wang_liqiang@sjtu.edu.cn) metals. Traditional trial-and-error approaches and experience-based optimization
Zhou Jing methods are increasingly inadequate for meeting the demands of high-reliability
(zhoujing_goplay@163.com)
medical applications. In recent years, machine learning (ML) has emerged as a
Citation: Mao Y, Jiang D, powerful data-driven tool, deeply integrating into every stage of AM for biomedical
Vladimir U, Jing Z, Wang L.
Machine learning-driven additive metals and providing a driving force for its intelligent transformation and upgrading.
manufacturing of biomedical metals: This review outlines three key applications of ML in biomedical metal AM: at
A review of forward prediction, the property prediction stage, ML enables forward prediction of performance
inverse optimization, and quality
control. Eng Sci Add Manuf. characteristics by establishing precise mapping relationships between process
2025;1(4):025440031. parameters and macrostructure quality, microstructure, and mechanical/functional
doi: 10.36922/ESAM025440031 properties; at the process optimization level, ML-driven inverse optimization
Received: October 29, 2025 algorithms efficiently navigate high-dimensional parameter spaces to achieve both
single-objective perfection and multi-objective balancing; at the quality monitoring
Revised: November 24, 2025
and control level, ML enables real-time diagnosis of manufacturing defects and
Accepted: November 26, 2025 even closed-loop adaptive control by integrating multiple in situ sensor data. This
Published online: December 5, review explores how ML can facilitate the biomedical metals during the AM process
2025 and outlines its future development toward fully integrated intelligent design and
Copyright: © 2025 Author(s). manufacturing processes.
This is an Open-Access article
distributed under the terms of the
Creative Commons Attribution Keywords: Machine learning; Additive manufacturing; Biomedical metals; Forward
License, permitting distribution, prediction; Inverse optimization; Quality control and monitoring
and reproduction in any medium,
provided the original work is
properly cited.
Publisher’s Note: AccScience
Publishing remains neutral with 1. Introduction
regard to jurisdictional claims in 1 2
published maps and institutional Biomedical metals, such as titanium and its alloys, cobalt-chromium alloys,
4
3
affiliations. biodegradable magnesium/zinc alloys, and medical-grade stainless steel, are widely
Volume 1 Issue 4 (2025) 1 doi: 10.36922/ESAM025440031