Radiomics-based machine learning differentiates “ground-glass” opacities due to COVID-19 from acute non-COVID-19 lung disease

Abstract

Ground-glass opacities (GGOs) are a non-specific high-resolution computed tomography (HRCT) finding tipically observed in early Coronavirus disesase 19 (COVID-19) pneumonia. However, GGOs are also seen in other acute lung diseases, thus making challenging the differential diagnosis. To this aim, we investigated the performance of a radiomics-based machine learning method to discriminate GGOs due to COVID-19 from those due to other acute lung diseases. Two sets of patients were included: a first set of 28 patients (COVID) diagnosed with COVID-19 infection confirmed by real-time polymerase chain reaction (RT-PCR) between March and April 2020 having (a) baseline HRCT at hospital admission and (b) predominant GGOs pattern on HRCT; a second set of 30 patients (nCOVID) showing (a) predominant GGOs pattern on HRCT performed between August 2019 and April 2020 and (b) availability of final diagnosis. Two readers independently segmented GGOs on HRCTs using a semi-automated approach, and radiomics features were extracted using a standard open source software (PyRadiomics). Partial least square (PLS) regression was used as the multivariate machine-learning algorithm. A leave-one-out nested cross-validation was implemented. PLS β-weights of radiomics features, including the 5% features with the largest β-weights in magnitude (top 5%), were obtained. The diagnostic performance of the radiomics model was assessed through receiver operating characteristic (ROC) analysis. The Youden’s test assessed sensitivity and specificity of the classification. A null hypothesis probability threshold of 5% was chosen (p < 0.05). The predictive model delivered an AUC of 0.868 (Youden’s index = 0.68, sensitivity = 93%, specificity 75%, p = 4.2 × 10–7). Of the seven features included in the top 5% features, five were texture-related. A radiomics-based machine learning signature showed the potential to accurately differentiate GGOs due to COVID-19 pneumonia from those due to other acute lung diseases. Most of the discriminant radiomics features were texture-related. This approach may assist clinician to adopt the appropriate management early, while improving the triage of patients.

3D CT-Inclusive Deep-Learning Model to Predict Mortality, ICU Admittance, and Intubation in COVID-19 Patients

“…The need of rapid diagnosis and effective allocation of the available resources to guarantee the best treatment possible [ 5 ] opens the way for taking advantage of predictive models to optimize patient care in this critical scenario. As a consequence, numerous AI-powered predictive models have populated the literature in the past 2 years [ 6 , 13 , 14 ]. However, strong reproducibility and generalizability across different patient populations and different centers are needed to translate theoretical models into the clinical practice.…”

“…Disease severity on chest CT correlated with the clinical status of COVID-19 patients and was successfully used to predict short-term progression in recent studies [ 11 , 12 ]. Chest CT-based DL models demonstrated high accuracy in differentiating COVID-19 from community-acquired pneumonia and non-COVID-19-related ground-glass opacities [ 13 , 14 ].…”

Sansoninto-induced Lung Injury

“…Pulmonary infiltration unresolved by antibiotics, a normal or slightly elevated leukocyte count and CRP level, an elevated KL-6 level, and a decreased percent volume capacity on a pulmonary function test all support the diagnosis ( 5 ). HRCT showed non-segmental ground-glass attenuation, which indicates viral pneumonia or immune-mediated pneumonitis ( 6 ). As the patient was treated in the midst of the COVID-19 pandemic, we excluded SARS-CoV-2 infection by PCR first; a simultaneous bronchoscopic evaluation was not performed due to reluctance, as it was a high-risk procedure with the potential for transmission.…”

COVID-DSNet: A novel deep convolutional neural network for detection of coronavirus (SARS-CoV-2) cases from CT and Chest X-Ray images

“…Hypoxemia is the predominant finding of respiratory failure in patients with severe pneumonia due to SARS-CoV-2, while hypercapnia respiratory failure is rare [20] . In the result of pneumonia (pulmonary edema) caused by SARS-CoV-2, foggy structures are seen in the lung density; this unclear image in the lung is called ground-glass opacities (GGO) [21] . GGO does not occur solely as a result of SARS-CoV-2 infection.…”