Digital Biomarkers for Oncology Group, German Cancer Research Centre (DKFZ), Heidelberg, Germany.
Skin Cancer Unit, German Cancer Research Center (DKFZ), Heidelberg, Germany; Department of Dermatology, Venereology and Allergology, University Medical Center Mannheim, Ruprecht-Karl University of Heidelberg, Mannheim, Germany; DKFZ Hector Cancer Institute at the University Medical Center Mannheim, Mannheim, Germany.
Eur J Cancer. 2023 Apr;183:131-138. doi: 10.1016/j.ejca.2023.01.021. Epub 2023 Feb 4.
In machine learning, multimodal classifiers can provide more generalised performance than unimodal classifiers. In clinical practice, physicians usually also rely on a range of information from different examinations for diagnosis. In this study, we used BRAF mutation status prediction in melanoma as a model system to analyse the contribution of different data types in a combined classifier because BRAF status can be determined accurately by sequencing as the current gold standard, thus nearly eliminating label noise.
We trained a deep learning-based classifier by combining individually trained random forests of image, clinical and methylation data to predict BRAF-V600 mutation status in primary and metastatic melanomas of The Cancer Genome Atlas cohort.
With our multimodal approach, we achieved an area under the receiver operating characteristic curve of 0.80, whereas the individual classifiers yielded areas under the receiver operating characteristic curve of 0.63 (histopathologic image data), 0.66 (clinical data) and 0.66 (methylation data) on an independent data set.
Our combined approach can predict BRAF status to some extent by identifying BRAF-V600 specific patterns at the histologic, clinical and epigenetic levels. The multimodal classifiers have improved generalisability in predicting BRAF mutation status.
在机器学习中，多模态分类器可以提供比单模态分类器更具概括性的性能。在临床实践中，医生通常也依赖于来自不同检查的一系列信息来进行诊断。在这项研究中，我们使用黑色素瘤中的 BRAF 突变状态预测作为模型系统，来分析组合分类器中不同数据类型的贡献，因为 BRAF 状态可以通过测序作为当前的金标准准确确定，从而几乎消除了标签噪声。
我们通过组合单独训练的图像、临床和甲基化数据的随机森林，来训练一个基于深度学习的分类器，以预测癌症基因组图谱队列中原发性和转移性黑色素瘤中的 BRAF-V600 突变状态。
通过我们的多模态方法，我们在独立数据集上获得了 0.80 的接收器操作特征曲线下面积，而个体分类器则在 0.63（组织病理学图像数据）、0.66（临床数据）和 0.66（甲基化数据）的接收器操作特征曲线下面积。
我们的组合方法可以通过在组织学、临床和表观遗传水平上识别 BRAF-V600 特异性模式，在一定程度上预测 BRAF 状态。多模态分类器在预测 BRAF 突变状态方面提高了通用性。
