Interpretable survival prediction for colorectal cancer using deep learning.

Interpretable survival prediction for colorectal cancer using deep learning.
复制标题

DOI:
10.1038/s41746-021-00427-2
复制
发表时间:
2021-04-19
影响因子:
15.2
通讯作者:
Mermel CH
Mermel CH
中科院分区:
医学1区
文献类型:
--
作者:
Wulczyn E;Steiner DF;Moran M;Plass M;Reihs R;Tan F;Flament-Auvigne I;Brown T;Regitnig P;Chen PC;Hegde N;Sadhwani A;MacDonald R;Ayalew B;Corrado GS;Peng LH;Tse D;Müller H;Xu Z;Liu Y;Stumpe MC;Zatloukal K;Mermel CH

文献摘要

参考文献

被引文献

相似文献

从基于深度学习的预后组织病理学模型中得出可解释的预后特征仍然是一个挑战。在这项研究中,我们开发了一个深度学习系统 (DLS),使用 3652 个病例(27,300 张幻灯片)来预测 II 期和 III 期结直肠癌的疾病特异性生存率。当对分别包含 1239 例(9340 张幻灯片)和 738 例(7140 张幻灯片)的两个验证数据集进行评估时,DLS 的 5 年疾病特异性生存 AUC 分别为 0.70(95% CI:0.66-0.73)和 0.69(95% CI:0.64-0.72),并且为一组九个临床病理特征增加了显着的预测价值。为了解释 DLS,我们探索了不同的人类可解释特征解释 DLS 分数差异的能力。我们观察到,T 类、N 类和等级等临床病理特征解释了 DLS 评分的一小部分方差(两个验证集中的 R2 = 18%)。接下来,我们通过对基于深度学习的图像相似性模型的嵌入进行聚类来生成人类可解释的组织学特征,并表明它们解释了大部分方差(R2 为 73-80%)。此外,与高 DLS 分数最密切相关的聚类衍生特征在孤立时也具有很高的预后作用。由于具有独特的视觉外观(邻近脂肪组织的分化差的肿瘤细胞簇),注释者以 87.0-95.5% 的准确度识别了该特征。我们的方法可用于解释预后深度学习模型的预测,并揭示潜在的新颖预后特征,人们可以可靠地识别这些特征以进行未来的验证研究。
Deriving interpretable prognostic features from deep-learning-based prognostic histopathology models remains a challenge. In this study, we developed a deep learning system (DLS) for predicting disease-specific survival for stage II and III colorectal cancer using 3652 cases (27,300 slides). When evaluated on two validation datasets containing 1239 cases (9340 slides) and 738 cases (7140 slides), respectively, the DLS achieved a 5-year disease-specific survival AUC of 0.70 (95% CI: 0.66–0.73) and 0.69 (95% CI: 0.64–0.72), and added significant predictive value to a set of nine clinicopathologic features. To interpret the DLS, we explored the ability of different human-interpretable features to explain the variance in DLS scores. We observed that clinicopathologic features such as T-category, N-category, and grade explained a small fraction of the variance in DLS scores (R2 = 18% in both validation sets). Next, we generated human-interpretable histologic features by clustering embeddings from a deep-learning-based image-similarity model and showed that they explained the majority of the variance (R2 of 73–80%). Furthermore, the clustering-derived feature most strongly associated with high DLS scores was also highly prognostic in isolation. With a distinct visual appearance (poorly differentiated tumor cell clusters adjacent to adipose tissue), this feature was identified by annotators with 87.0–95.5% accuracy. Our approach can be used to explain predictions from a prognostic deep learning model and uncover potentially-novel prognostic features that can be reliably identified by people for future validation studies.
DOI: 10.1186/gb-2006-7-10-r100
发表时间: 2006
期刊: Genome biology
影响因子: 12.3
作者:
Carpenter AE;Jones TR;Lamprecht MR;Clarke C;Kang IH;Friman O;Guertin DA;Chang JH;Lindquist RA;Moffat J;Golland P;Sabatini DM
通讯作者: Sabatini DM
DOI: 10.1038/s41598-018-21758-3
发表时间: 2018-02-21
期刊: Scientific reports
影响因子: 4.6
作者:
Bychkov D;Linder N;Turkki R;Nordling S;Kovanen PE;Verrill C;Walliander M;Lundin M;Haglund C;Lundin J
通讯作者: Lundin J
DOI: 10.1148/radiographics.20.2.g00mc15399
发表时间: 2000-03-01
期刊: RADIOGRAPHICS
影响因子: 5.5
作者:
Horton, KM;Corl, FM;Fishman, EK
通讯作者: Fishman, EK
DOI: 10.1038/s41571-019-0252-y
发表时间: 2019-11-01
影响因子: 78.8
作者:
Bera, Kaustav;Schalper, Kurt A.;Madabhushi, Anant
通讯作者: Madabhushi, Anant
DOI: 10.1016/s0140-6736(07)61866-2
发表时间: 2007-12-15
期刊: LANCET
影响因子: 168.9
作者:
Gray, Richard;Barnwell, Jennifer;Kerr, David J.
通讯作者: Kerr, David J.