Computer Aided Detection and Diagnosis in Medical Imaging: A Review of Clinical and Educational Applications

Publication TypeConference Proceedings
Año de publicación2016
AuthorsHernández-Rodríguez, J, Cabrero-Fraile, FJ, Rodríguez-Conde, MJ, Gómez-Llorente, PLuis
Nombre de la ConferenciaTEEM '16 Proceedings of the Fourth International Conference on Technological Ecosystems for Enhancing Multiculturality
Fecha de publicación11/2016
PublisherACM New York, NY, USA
Conference LocationSalamanca, Spain — November 02 - 04, 2016
Keywordsartificial neural networks, cancer detection, computed tomography, computer aided detection, computer aided diagnosis, convolutional neural networks, deep learning, image database, image processing, mammography, Medical education, medical imaging, radiography, radiology, radiology training

Computer Aided Detection and Diagnosis, known as CAD, is a fundamental tool for assisting radiologists in the image interpretation task. Since the 1980s, its popularity has grown, and it has become an important area of research in computer science, generating a great amount of published papers and many software systems. A wide variety of computer algorithms, such as image processing techniques, classification systems and artificial intelligence approaches, such as Artificial Neural Networks (ANNs), have been used in this field. CAD applications have extended to an increasing number of image modalities, as for example radiography, ultrasound, computed tomography (CT), mammography, magnetic resonance imaging (MRI), and positron emission tomography (PET) and to a great number of diseases, such as lung cancer, breast cancer, colon cancer, Alzheimer's disease, spine cancer, aneurysm, interstitial lung diseases... In this paper, a review of the fundamentals of some of these CAD systems, with special emphasis on those schemes based on ANNs, will be conducted. An analysis of methods and results obtained in clinical practice and a report of the advantages derived from its use (mainly focused in mammography, thoracic and colonic CT) will also be presented. In the last section, a compilation of CAD papers related with medical education and specialists' training reveals that this area has not been developed by many researchers. CAD schemes with educational purposes have been created for learning patterns in error making by radiology trainees, identifying regions where false positive detection is more probable, and predicting case difficulty for improving the learning process. The use of Computer Aided Education related with CAD software and Content-based image retrieval techniques have also been utilized as methods for enhancing learning.