One Shot Model For The Prediction of COVID-19 and Lesions Segmentation In Chest

One Shot Model For The Prediction of COVID-19 and Lesions Segmentation In Chest CT Scans Through The Affinity Among Lesion Mask Features

We introduce a model that segments lesions and predicts COVID-19 from chest CT scans through the derivationof an affinity matrix between lesion masks. The novelty of the methodology is based on the computation of theaffinity between the lesion masks’ features extracted from the image.First, a batch of vectorized lesion masks isconstructed. Then, the model learns the parameters of the affinity matrix that captures the relationship between featuresin each vector. Finally, the affinity is expressed as a single vector of pre-defined length. Without any complicated datamanipulation, class balancing tricks, and using only a fraction of the training data, we achieve a 91.74% COVID-19sensitivity, 85.35% common pneumonia sensitivity, 97.26% true negative rate and 91.94% F1-score. Ablation studies showthat the method can quickly generalize to new datasets. All source code, models and results are publicly available onhttps://github.com/AlexTS1980/COVID-Affinity-Model.

通过病变面罩特征之间的亲和力预测胸部CT扫描中COVID-19和病变分割的单发模型

我们引入了一种模型，该模型可以通过对病变面罩之间的亲和矩阵进行推导，对病变进行分割并通过胸部CT扫描预测COVID-19。该方法的新颖性基于对从图像中提取的病变蒙版特征之间的亲和力的计算。.. 首先，构建一批矢量化病变蒙版。然后，模型学习捕获每个向量中特征之间关系的亲和力矩阵的参数。最后，亲和力表示为预定义长度的单个向量。无需任何复杂的数据处理，类平衡技巧，仅使用训练数据的一小部分，我们就可以达到91.74％的COVID-19敏感性，85.35％的普通肺炎敏感性，97.26％的真实阴性率和91.94％的F1评分。消融研究表明，该方法可以快速推广到新的数据集。所有源代码，模型和结果均可在https://github.com/AlexTS1980/COVID-Affinity-Model上公开获得。 （阅读更多）