The scientific community has joined forces to mitigate the scope of the current COVID-19 pandemic. The early identification of the disease, as well as the evaluation of its evolution is a primary task for the timely application of medical protocols. The use of medical images of the chest provides valuable information to specialists. Specifically, chest X-ray images have been the focus of many investigations that apply artificial intelligence techniques for the automatic classification of this disease. The results achieved to date on the subject are promising. However, some results of these investigations contain errors that must be corrected to obtain appropriate models for clinical use. This research discusses some of the problems found in the current scientific literature on the application of artificial intelligence techniques in the automatic classification of COVID-19. It is evident that in most of the reviewed works an incorrect evaluation protocol is applied, which leads to overestimating the results.

Since the outbreak of the COVID-19 pandemic, computer vision researchers have been working on automatic identification of this disease using radiological images. The results achieved by automatic classification methods far exceed those of human specialists, with sensitivity as high as 100% being reported. However, prestigious radiology societies have stated that the use of this type of imaging alone is not recommended as a diagnostic method. According to some experts the patterns presented in these images are unspecific and subtle, overlapping with other viral pneumonias. This report seeks to evaluate the analysis the robustness and generalizability of different approaches using artificial intelligence, deep learning and computer vision to identify COVID-19 using chest X-rays images. We also seek to alert researchers and reviewers to the issue of “shortcut learning”. Recommendations are presented to identify whether COVID-19 automatic classification models are being affected by shortcut learning. Firstly, papers using explainable artificial intelligence methods are reviewed. The results of applying external validation sets are evaluated to determine the generalizability of these methods. Finally, studies that apply traditional computer vision methods to perform the same task are considered. It is evident that using the whole chest X-Ray image or the bounding box of the lungs, the image regions that contribute most to the classification appear outside of the lung region, something that is not likely possible. In addition, although the investigations that evaluated their models on data sets external to the training set, the effectiveness of these models decreased significantly, it may provide a more realistic representation as how the model will perform in the clinic. The results indicate that, so far, the existing models often involve shortcut learning, which makes their use less appropriate in the clinical setting.

No abstract

El vertiginoso crecimiento y la precisión de las técnicas de Inteligencia Artificial (AI, del inglés Artificial Intelligence) permiten analizar grandes volúmenes de datos de forma rápida y eficiente. En ese sentido, la aplicación de técnicas de reconocimiento facial en sistemas de seguridad (video-vigilancia) no quedan exentas y resultan convenientes pues asistirían al desempeño humano en las labores de observación, interpretación y etiquetado de imágenes en tiempo real, a la vez que funcionarían como un sistema de alerta o alarma ante la presencia de intrusos. La implementación de estos sistemas puede llevarse a cabo con hardware relativamente barato y aprovechando la capacidad de procesamiento del clúster big data de la Universidad Central “Marta Abreu” de Las Villas (UCLV). Con la puesta en práctica del proyecto se ofrecen soluciones a las problemáticas identificadas en la dirección de informatización asociadas a la gestión de cuentas por parte de los usuarios y aplicaciones futuras relacionadas con la detección de personal en áreas de interés. Con la implementación se pretenden dos posibles contribuciones: en primer lugar, se ha de diseñar un procedimiento capaz de ensamblar un conjunto de datos a gran escala, minimizando al mismo tiempo la cantidad de anotaciones manuales involucradas. Este procedimiento se ha de desarrollar para caras, pero evidentemente es adecuado para otras clases de objetos, así como para tareas específicas. La segunda contribución ha de ser mostrar que una Red Neuronal Convolucional (CNN, del inglés Convolutional Neural Network), profunda con la formación adecuada, puede lograr resultados comparables a los del estado de la técnica.

PurposeThe development of a robust model for automatic identification of COVID-19 based on chest x-rays has been a widely addressed topic over the last couple of years; however, the scarcity of good quality images sets, and their limited size, have proven to be an important obstacle to obtain reliable models. In fact, models proposed so far have suffered from over-fitting erroneous features instead of learning lung features, a phenomenon known as shortcut learning. In this research, a new image classification methodology is proposed that attempts to mitigate this problem. Methods To this end, annotation by expert radiologists of a set of images was performed. The lung region was then segmented and a new classification strategy based on a patch partitioning that improves the resolution of the convolution neural network is proposed. In addition, a set of native images, used as an external evaluation set, is released. ResultsThe best results were obtained for the 6-patch splitting variant with 0.887 accuracy, 0.85 recall and 0.848 F1score on the external validation set. ConclusionThe results show that the proposed new strategy maintains similar values between internal and external validation, which gives our model generalization power, making it available for use in hospital settings.