压缩感知让MRI扫描提速8倍
推荐指数 68.0 NO. 011 · 2026.05.16
发布2026/05/15Score65Comments23
为什么值得看
斯坦福统计学家Donoho将高维几何中的压缩感知理论应用于MRI重建，通过随机欠采样配合凸优化，把扫描时间从1小时压缩到8分钟。这是纯数学理论直接颠覆医疗硬件的罕见案例，对做信号处理、医疗AI的工程师有范式参考价值。
编辑判断
压缩感知的核心洞察是反直觉的：不是先采样再压缩，而是直接采集压缩后的信息。Donoho团队证明，只要信号本身是稀疏的（医学图像在特定变换域确实如此），随机采样后的重建问题可以转化为一个凸优化问题，用L1范数最小化稳定求解。
这个思路在2017年之后已经渗透进更广泛的AI基础设施。现在的扩散模型、神经辐射场（NeRF）的隐式重建，本质上都在借用同一套"欠采样+先验约束"的框架。做AI for Science的团队如果还在用传统全采样思路处理高维数据，可能直接损失一个数量级的效率。
实际落地层面，GE、Siemens的MRI设备已经内置了基于压缩感知的加速协议，但开源实现（如BART工具箱）和商用算法的差距仍然很大，这是医疗AI创业的一个被低估的切入点。
社区反馈
意见分歧 20 条评论
核心争论：压缩感知的实际价值 vs 数学基础研究ROI的争议
You can do a lot better than this if you redefine the problem from directly generating images with certain contrasts to maximizing information gain, even with weak magnets. They've since basically run out of money and are on life support, but Q Bio [0] had that tech working years ago, able to quickl
I remember one of my diploma students continued with discrete tomography as PhD, topic "Binary Tomography by Iterating Linear Programs" and I found it super interesting to get down the number of shots and at the same time increasing the accuracy a lot.
It's a nice review but the end reads like a funding pitch. The most important Mathematicians like donoho and Tao in the US seem to currently experience budget cuts and start to address the public.