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

Abstract

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

Lung Segmentation and Characterization in COVID-19 Patients for Assessing Pulmonary Thromboembolism: An Approach Based on Deep Learning and Radiomics

“…Several authors have already dealt with the problem of COVID-19 detection from CT scans [12,17,18,21,22] or focused on the task of lung and lesion segmentation [13,[15][16][17][20][21][22]. In our cohort, all of the patients were positive to COVID-19, since the objective was not that of realizing another algorithm for COVID-19 diagnosis.…”

“…Several recent studies addressed the task of lung parenchyma and lesion segmentation from CT scans in COVID-19 patients [12][13][14][15][16][17][18][19][20][21][22]. Wu et al proposed a joint classification and segmentation (JCS) system in order to address the issue of performing real-time and explainable COVID-19 diagnosis from CT scans [12].…”

“…The approach considered by the authors incorporates the following steps: thresholding to extract the lung region; image enhancement exploiting harmony search optimization and Otsu thresholding; image segmentation and region of interest (ROI) extraction from the binary image to compute the level of severity [16]. Ter-Sarkisov et al proposed a one-shot model by exploiting Mask R-CNN [23] functionalities for the COVID-19 classification and lesion segmentation in chest CT scans utilizing the affinity between lesion mask features [17]. Zhao et al built an open-source dataset composed of 349 COVID-19 CT images, coming from 216 patients, and 463 non-COVID-19 CT images.…”