分析测试学报2025，Vol.44Issue(6)：1208-1218,1226,12.DOI:10.12452/j.fxcsxb.240929423
基于机器学习的质谱成像法庭科学应用进展
Application Advances in Forensic Science for Machine Learning-based Mass Spectrometry Imaging
摘要
Abstract
Given the continuous evolution of criminal tactics,the field of forensic science urgently requires a technology capable of analyzing both the morphology and composition of physical evidence.Mass spectrometry imaging(MSI)technology,integrated with machine learning,offers a highly sen-sitive,specific,and extensive analysis method that is nearly non-destructive.It excels in the exami-nation of fingerprints,documents,and physicochemical evidence by providing the potential to ex-tract critical chemical information from rich datasets.This paper reviews the current state of research on MSI data analysis,including the tracing of the origin of residues,the analysis of the duration of residue presence,and image enhancement techniques.And this review highlights the integration of MSI with machine learning algorithms for the forensic analysis of various types of evidence.The syner-gy between MSI's ability to generate detailed chemical images and machine learning's data processing capabilities effectively addresses complex forensic challenges.The paper underscores the importance of this technology in enhancing the extraction of evidential information,thereby supporting more ac-curate and efficient investigations in legal proceedings.Despite the hurdles,such as operational com-plexity and the interpretability of algorithms,the future of forensic science,empowered by MSI and machine learning,appears poised to deliver transformative solutions to real-world forensic problems.关键词
质谱成像/法庭科学/机器学习/多元变量分析Key words
mass spectrometry imaging/forensic science/machine learning/multivariate analysis分类
化学化工引用本文复制引用
张硕,满瀚泽,郭冲,赵雅彬,李展平..基于机器学习的质谱成像法庭科学应用进展[J].分析测试学报,2025,44(6):1208-1218,1226,12.基金项目
中国人民公安大学刑事科学技术双一流创新研究专项(2023SYL06) （2023SYL06）
