结合CNN和Transformer病变信号引导的蜂窝肺CT图像识别

Combination of CNN and Transformer for Lesion-Guided Honeycomb Lung CT Image Recognition

蜂窝肺是多种肺部疾病晚期的CT影像表现,其影像特征表现为多样化的囊状病变,呈现蜂窝样外观。现有的计算机辅助诊断方法,难以有效解决蜂窝肺病灶形态不一、位置不同所导致的识别准确率不高的问题。因此,提出一种结合CNN和Transformer病变信号引导下的蜂窝肺CT影像识别模型。采用多尺度信息增强模块丰富CNN获取的不同尺度下特征的空间信息与通道信息,同时使用病变信号生成模块以强化病变特征表达;利用Transformer获取特征的长距离依赖信息以弥补CNN在提取全局信息方面的缺陷;引入多头交叉注意力机制融合特征信息获得分类结果。实验结果表明,该模型在蜂窝肺和COVID-CT数据集中分别取得了99.67%、97.08%的精确度,对比其他模型具有更加精准的识别结果,验证了模型的有效性和泛化性。

Honeycomb lung is a CT imaging manifestation of various advanced lung diseases,characterized by diverse cystic lesions presenting a honeycomblike appearance.Existing computeraided diagnosis methods struggle to effectively address the low identification accuracy caused by the varied morphology and different locations of cellular lung lesions.Therefore,a combined CNN and Transformer model guided by lesion signals is proposed for cellular lung CT image recognition.In this model,a multiscale information enhancement module is first employed to enrich the spatial and channel information of features obtained by CNN at different scales.Simultaneously,a lesion signal generation module is used to strengthen the expression of lesion features.Subsequently,Transformer is utilized to capture longrange dependency information of features,compensating for the deficiency of CNN in extracting global information.Finally,a multihead crossattention mechanism is introduced to fuse feature information and obtain classification results.Experimental results demonstrate that the proposed model achieves accuracies of 99.67%and 97.08%on the honeycomb lung and COVIDCT dataset,respectively.It outperforms other models,providing more precise recognition results and validating the effectiveness and generalization of the model.