生物与医学2019arxiv.org
使用深度学习方法在乳腺钼靶影像中检测乳腺癌
Breast cancer detection in mammography using deep learning approach
进阶 · 约 30 分钟读完前置知识：深度学习基础、卷积神经网络、乳腺钼靶影像
导读
这篇论文研究如何用深度学习方法在乳腺钼靶影像中检测乳腺癌。从标题看，核心是让模型从钼靶图像中识别癌症相关征象；具体性能与数据设置需以论文为准。
HN 关注点：AI 在放射科的实际能力与局限，以及乳腺钼靶筛查本身是否值得大规模投入。
HN 评论精选
这几条评论里：支持 1 质疑 1 补充 1 亲历 1
亲历放射科医生强调钼靶是筛查而非确诊，AI 讨论常忽略临床流程与筛查效用争议。
作为放射科医生，我指出这类 AI 放射科讨论常忽略几点：钼靶不是孤立解读，它通常是一长串检查的第一步，更准确说是决定是否需要活检的筛查，而非癌症诊断；钼靶的总体诊断表现并不算好，所以常用超声、断层合成和 MRI 作为支持或替代；钼靶筛查的整体效用也存在争议。放射科医生比任何人都希望提高敏感性和特异性。那些鼓吹“放射学亟待颠覆”或“AI 超越放射科医生”的人，显然没进过诊所，但向 VC 或管理层推销时可能很喜欢这套话术。
neuro_image3查看英文原文
There are several key points that get left out in AI radiology conversations such as this one: 1) Mammograms are not interpreted in a vacuum. In fact mammograms are usually the first in a long line of tests before a breast cancer or other diagnosis is ultimately made. In fact, it's probably more accurate to refer to mammography as a screening exam for which patients need a biopsy rather than a diagnostic test for cancer (there are rare exceptions, but overall this point holds). 2) Speaking frankly as a radiologist myself, tests like mammograms aren't even that good in terms of overall diagnosis. Thats why ultrasound, tomosynthesis and MRI are often used as supporting evidence and/or alternative exams. 3) There is controversy over the overall utility of mammograms, particularly in the screening context. Radiologists more than anyone would like the sensitivity and specificity of these studies to be higher. It strikes me that the people that push these "radiology is ripe for disruption" or "AI outperforms radiologists" hyperbolic arguments are clearly people that have never seen the inside of a clinic. I'm sure they love this rhetoric though when pitching to VCs or sitting around the conference table coming up with 'breakthrough ideas' to turn into power-points for the other administrators.
支持放射科因可远程化和大量标注数据，最易受 AI 与外包冲击。
在主要专科中，放射科似乎最可能被显著颠覆。首先它可以远程完成，若监管放松，外国放射科医生可能以低得多的成本读片。其次，深度学习在这类任务上能快速进步，因为已有大量标注数据集：每张钼靶拍完后就已被标为正常或异常。
RcouF1uZ4gsC查看英文原文
Of the major specialties, it seems that radiology is the most in danger of significant disruption. First of all, it can be done remotely, so there is risk if the regulation is lightened that foreign radiologists will be allowed to read studies at much less cost. The other issue is that this is something that deep learning can rapidly progress in given there are already a plethora of labeled data sets. For example, every mammogram that is taken has already been labeled normal or abnormal.
补充乐观预测 AI 影像将扩展到全身筛查和神经退行性疾病早期发现。
我想这只是冰山一角。不久后我们可能每隔几年做一次全身 3D 扫描，从所有癌症、心脏病、胃肠问题，到痤疮和头皮屑这样平常的问题，都由算法基于健康与患病身体部位的累积图像数据库来诊断。真正的希望是能看进大脑，在症状出现前多年发现 CTE 和阿尔茨海默病。
mikece查看英文原文
I imagine this is just scratching the surface and before long we'll be doing full-body, 3D scans every few years and everything from cancers (all of them) to heart disease, to gastero-intestinal issues, to things even as mundane as acne and dandruff will be diagnosed by algorithms pulling on a cumulative database of images of healthy and diseased body parts. The real hope is to be able to see into the brain and pick up things like CTE and Alzheimers years before symptoms manifest.
质疑质疑钼靶筛查价值，主张 AI 转向 ctDNA 和 3D 超声等新模态。
与其在钼靶这块“挤血”——多年巨额投入仍未改善乳腺癌患者寿命——AI/ML 应更广地看循环肿瘤 DNA 和 3D 超声等模态。
statesdj查看英文原文
Instead of trying to squeeze blood from the mammography stone that has failed to improve longevity in breast cancer patients despite enormous investment over many years, AI/ML need to take a broader perspective and look at modalities like circulating tumor DNA and 3-D ultrasound
相关资源
从这篇论文所有 HN 讨论帖的全部评论里整理出来的链接。
历次讨论
- 2019Hacker News 讨论帖190 分 · 116 评
同领域还值得读
CRISPR 在活体动物中消除 HIV-1 感染
CRISPR eliminates HIV-1 infection in live animals
报告利用 CRISPR/Cas9 基因编辑技术在活体动物中清除 HIV-1 感染：研究人员通过腺病毒载体递送特定的 Cas9 系统，从被感染细胞中切除嵌入的 HIV 病毒 DNA。
生物与医学2017 · 1,455 分 · 239 评入门 · 约 30 分钟编辑精选保护区飞行昆虫生物量27年下降75%
75% decline over 27 years in total flying insect biomass in protected areas
这篇发表于 PLOS ONE 的研究通过在德国保护区长达 27 年的昆虫生物量监测，发现飞行昆虫总量下降了约 75%。
生物与医学2017 · 901 分 · 432 评入门 · 约 45 分钟编辑精选为什么大多数已发表的研究发现是错的
Why Most Published Research Findings Are False (2005)
这篇 2005 年发表于 PLOS Medicine 的经典论文由 John Ioannidis 撰写，论证了在统计功效不足、研究偏倚和多重检验等因素下，大多数已发表的医学研究发…
生物与医学2022 · 388 分 · 351 评入门 · 约 60 分钟编辑精选人类免疫系统的物理接线图
A physical wiring diagram for the human immune system
这篇 Nature 论文系统绘制了人类免疫系统的“物理接线图”：通过大规模实验鉴定免疫细胞表面受体之间的相互作用（包括约 30 种此前未知的受体互作），并结合定量计算模型模拟免疫…
生物与医学2022 · 342 分 · 49 评专业 · 约 90 分钟编辑精选脂肪组织在减重后仍保留肥胖的表观遗传记忆
Adipose tissue retains an epigenetic memory of obesity after weight loss
这篇 Nature 论文研究了肥胖对脂肪组织的长期影响：即使减重成功，脂肪细胞仍保留着肥胖时期的表观遗传记忆，可能解释为什么减重后容易反弹。
生物与医学2025 · 263 分 · 327 评进阶 · 约 45 分钟编辑精选