帕金森病早期诊断受主观判断、资源依赖及技术局限制约。“诊断笔”创新集成磁弹性笔尖与铁磁流体墨水，实时捕捉书写运动信号，经一维维卷积神经网络（1D CNN）分析，临床试点准确率达96.22% 。该技术低成本、便携，适配基层医疗与远程监测场景，可推动大规模筛查及治疗方案动态调整，缩小医疗资源差距。虽存在样本量有限、纵向监测能力不足等问题，但为神经退行性疾病数字化诊断提供新路径。未来可融合语言、步态等多模态数据，优化硬件与算法，完善伦理机制，助力全球健康公平。
Early diagnosis of Parkinson’s disease (PD) suffers from subjectivity, resource dependence, and technical limitations. The "diagnosis pen" innovatively integrates a magnetoelastic tip and ferrofluid ink to capture writing motion signals in real - time, which are then analyzed by a 1D convolutional neural network (1D CNN). Pilot clinical tests show an accuracy of 96.22%. Cost - effective and portable, this technology is suitable for primary care and remote monitoring scenarios, enabling large - scale screening and dynamic adjustment of treatment plans to narrow the gap in medical resources. Although there are issues such as limited sample size and insufficient longitudinal monitoring capabilities, it provides a new approach to the digital diagnosis of neurodegenerative diseases. In the future, it can integrate multimodal data such as language and gait, optimize hardware and algorithms, and improve the ethical mechanism to promote global health equity.
