Salmanpour Mohammad R, Gorji Arman, Mousavi Amin, Fathi Jouzdani Ali, Sanati Nima, Maghsudi Mehdi, Leung Bonnie, Ho Cheryl, Yuan Ren, Rahmim Arman
BC Cancer Research Institute, Vancouver, BC V5Z 1L3, Canada.
Department of Radiology, University of British Columbia, Vancouver, BC V6T 1Z4, Canada.
Cancers (Basel). 2025 Jan 17;17(2):285. doi: 10.3390/cancers17020285.
This study explores a semi-supervised learning (SSL), pseudo-labeled strategy using diverse datasets such as head and neck cancer (HNCa) to enhance lung cancer (LCa) survival outcome predictions, analyzing handcrafted and deep radiomic features (HRF/DRF) from PET/CT scans with hybrid machine learning systems (HMLSs).
We collected 199 LCa patients with both PET and CT images, obtained from TCIA and our local database, alongside 408 HNCa PET/CT images from TCIA. We extracted 215 HRFs and 1024 DRFs by PySERA and a 3D autoencoder, respectively, within the ViSERA 1.0.0 software, from segmented primary tumors. The supervised strategy (SL) employed an HMLS-PCA connected with six classifiers on both HRFs and DRFs. The SSL strategy expanded the datasets by adding 408 pseudo-labeled HNCa cases (labeled by the Random Forest algorithm) to 199 LCa cases, using the same HMLS techniques. Furthermore, principal component analysis (PCA) linked with four survival prediction algorithms were utilized in the survival hazard ratio analysis.
The SSL strategy outperformed the SL method ( << 0.001), achieving an average accuracy of 0.85 ± 0.05 with DRFs from PET and PCA + Multi-Layer Perceptron (MLP), compared to 0.69 ± 0.06 for the SL strategy using DRFs from CT and PCA + Light Gradient Boosting (LGB). Additionally, PCA linked with Component-wise Gradient Boosting Survival Analysis on both HRFs and DRFs, as extracted from CT, had an average C-index of 0.80, with a log rank -value << 0.001, confirmed by external testing.
Shifting from HRFs and SL to DRFs and SSL strategies, particularly in contexts with limited data points, enabling CT or PET alone, can significantly achieve high predictive performance.
本研究探索一种半监督学习（SSL）伪标签策略，该策略使用头颈癌（HNCa）等多种数据集来增强肺癌（LCa）生存结果预测，通过混合机器学习系统（HMLS）分析PET/CT扫描中的手工制作和深度放射组学特征（HRF/DRF）。
我们从癌症成像存档（TCIA）和我们的本地数据库中收集了199例同时拥有PET和CT图像的LCa患者，以及来自TCIA的408例HNCa PET/CT图像。我们分别通过PySERA和一个3D自动编码器在ViSERA 1.0.0软件中从分割的原发性肿瘤中提取了215个HRF和1024个DRF。监督策略（SL）在HRF和DRF上采用了一个与六个分类器相连的HMLS-PCA。SSL策略通过使用相同的HMLS技术，将408例假标签HNCa病例（由随机森林算法标记）添加到199例LCa病例中，从而扩展了数据集。此外，在生存风险比分析中使用了与四种生存预测算法相关联的主成分分析（PCA）。
SSL策略优于SL方法（<<0.001），使用来自PET的DRF和PCA+多层感知器（MLP）时平均准确率达到0.85±0.05，而使用来自CT的DRF和PCA+轻梯度提升（LGB）的SL策略的平均准确率为0.69±0.06。此外，从CT中提取的HRF和DRF上与逐分量梯度提升生存分析相关联的PCA平均C指数为0.80，对数秩检验p值<<0.001，经外部测试得到证实。
从HRF和SL策略转向DRF和SSL策略，特别是在数据点有限、仅能使用CT或PET的情况下，能够显著实现较高的预测性能。
