论文标题
使用卷积神经网络模型的鱼类和纳米菌图像的多模式登记
Multimodal registration of FISH and nanoSIMS images using convolutional neural network models
论文作者
论文摘要
纳米级次级离子质谱法（纳米sims）和荧光原位杂交（FISH）显微镜分别提供了微生物研究中靶向微生物群落的身份和细胞活性的高分辨率，多模式图像表示。尽管对微生物学家的重要性很重要，但鉴于这两个图像中的形态失真和背景噪声，鱼类和纳米图像的多模式登记都在挑战。在这项研究中，我们使用卷积神经网络（CNN）进行多尺度特征提取，用于计算最小转换成本特征匹配的形状上下文和薄板样条（TPS）模型，以用于鱼类和纳米IMIMS图像的多模式注册。使用标准指标对手动注册图像进行定量评估注册精度。尽管所有六个经过测试的CNN模型都表现良好，但根据大多数指标，RESNET18的表现都超过了VGG16，VGG19，GOGLENET和SHUFFLENET和RESNET101。这项研究证明了CNN在具有明显的背景噪声和形态扭曲的多模式图像的注册中的实用性。我们还显示通过二进化保留的聚合形状，是注册多模式微生物学相关图像的可靠特征。
Nanoscale secondary ion mass spectrometry (nanoSIMS) and fluorescence in situ hybridization (FISH) microscopy provide high-resolution, multimodal image representations of the identity and cell activity respectively of targeted microbial communities in microbiological research. Despite its importance to microbiologists, multimodal registration of FISH and nanoSIMS images is challenging given the morphological distortion and background noise in both images. In this study, we use convolutional neural networks (CNNs) for multiscale feature extraction, shape context for computation of the minimum transformation cost feature matching and the thin-plate spline (TPS) model for multimodal registration of the FISH and nanoSIMS images. Registration accuracy was quantitatively assessed against manually registered images, at both, the pixel and structural levels using standard metrics. Although all six tested CNN models performed well, ResNet18 was observed to outperform VGG16, VGG19, GoogLeNet and ShuffleNet and ResNet101 based on most metrics. This study demonstrates the utility of CNNs in the registration of multimodal images with significant background noise and morphology distortion. We also show aggregate shape, preserved by binarization, to be a robust feature for registering multimodal microbiology-related images.
