Department of Electronic Engineering, Fudan University, Shanghai, China.
Department of Gastroenterology, Changhai Hospital, Shanghai, China.
Technol Health Care. 2022;30(S1):47-59. doi: 10.3233/THC-228005.
Automated diagnosis of gastrointestinal stromal tumors' (GISTs) cancerization is an effective way to improve the clinical diagnostic accuracy and reduce possible risks of biopsy. Although deep convolutional neural networks (DCNNs) have proven to be very effective in many image classification problems, there is still a lack of studies on endoscopic ultrasound (EUS) images of GISTs. It remains a substantial challenge mainly due to the data distribution bias of multi-center images, the significant inter-class similarity and intra-class variation, and the insufficiency of training data.
The study aims to classify GISTs into higher-risk and lower-risk categories.
Firstly, a novel multi-scale image normalization block is designed to perform same-size and same-resolution resizing on the input data in a parallel manner. A dilated mask is used to obtain a more accurate interested region. Then, we construct a multi-way feature extraction and fusion block to extract distinguishable features. A ResNet-50 model built based on transfer learning is utilized as a powerful feature extractor for tumors' textural features. The tumor size features and the patient demographic features are also extracted respectively. Finally, a robust XGBoost classifier is trained on all features.
Experimental results show that our proposed method achieves the AUC score of 0.844, which is superior to the clinical diagnosis performance.
Therefore, the results have provided a solid baseline to encourage further researches in this field.
自动化诊断胃肠道间质瘤（GISTs）癌变是提高临床诊断准确性和降低活检风险的有效方法。虽然深度卷积神经网络（DCNN）在许多图像分类问题中已被证明非常有效，但对于 GISTs 的内镜超声（EUS）图像仍然缺乏研究。这主要是由于多中心图像的数据分布偏差、类间相似度高和类内变异性大以及训练数据不足所导致的。
本研究旨在将 GISTs 分为高风险和低风险类别。
首先，设计了一种新颖的多尺度图像归一化块，以并行方式对输入数据进行相同大小和相同分辨率的调整。使用扩张掩模获得更准确的感兴趣区域。然后，构建了一种多通道特征提取和融合块来提取可区分的特征。基于迁移学习构建的 ResNet-50 模型被用作肿瘤纹理特征的强大特征提取器。同时还分别提取了肿瘤大小特征和患者人口统计学特征。最后，在所有特征上训练稳健的 XGBoost 分类器。
实验结果表明，我们提出的方法的 AUC 得分为 0.844，优于临床诊断性能。
因此，这些结果为该领域的进一步研究提供了坚实的基础。
