Document Actions


Content Based Image Retrieval (CBIR) is a search technology that could aid medical diagnosis by retrieving and presenting earlier reported cases that are related to the one being diagnosed. The diagnosis of interstitial lung diseases is one of the problems addressed.

Content Based Image Retrieval (CBIR) is a search paradigm that selects examples from an image collection with content that is similar to a query image. The large number of exams in Picture Archiving and Communication Systems (PACS) motivates the study of CBIR systems for Computer Aided Diagnosis (CAD), where given an undiagnosed exam, CBIR can be used to aid the diagnosis process by automatically retrieving previously reported relevant exams. This allows radiologists to compare cases and check fellow radiologists’ conclusions, which improves the diagnostics practice, particularly for inexperienced radiologists and complex diagnostic problems.

The objective of the Biomedical Imaging Lab is to design and implement a CBIR system that automatically collects medical information from radiology reports and learns the similarity and dissimilarity among them to help the specialists, providing similar cases that can help the diagnosis process.


One special clinical application is addressed in order to characterise and analyse the Interstitial Lung Disease (ILD). ILD is a set of more than 100 lung disorders that affect the lung interstitium. They cause progressive scarring of the lung tissue, involving breathing complications that could lead to respiratory failure in advanced stages. Some of these diseases are very difficult to differentiate, justifying the need of using advanced CBIR systems, with a learning strategy based on medical records, for guiding the metric learning in the Image Space.


José Ramos (INESC TEC), Isabel Ramos (FMUP), Rui Ramos (FMUP), Aurélio Campilho (INESC TEC).


Fundação para a Ciência e a Tecnologia (FCT).

Main Publications

  • Thessa T. J. P. Kockelkorn, Clara I. Sánchez, Jan C. Grutters, Rui Ramos, Pim A. de Jong, Max A. Viergever, José Ramos, Cornelia Schaefer-Prokop, Bram van Ginneken: Interactive classification of lung tissue in CT scans by combining prior and interactively obtained training data: A simulation study. ICPR 2012: 105-108. Link
  • José Ramos, Thessa T. J. P. Kockelkorn, Bram van Ginneken, Max A. Viergever, Rui Ramos, Aurélio Campilho: Supervised Content Based Image Retrieval Using Radiology Reports. ICIAR (2) 2012: 249-258. Link
  • José Ramos, Thessa Kockelkorn, Bram van Ginneken, Max A. Viergever, Jan Grutters, Rui Ramos, and Aurélio Campilho: Learning Interstitial Lung Diseases CT Patterns from Reports Keywords. IWPIA 2013: 21-32. Link


PhD. Thesis

  • José Ricardo Ferreira de Castro Ramos, Content based image retrieval as a computer aided diagnosis tool for radiologists. Doctoral Degree on Electrical and Computer Engineering, FEUP, (Supervisor: Aurélio Campilho), in process.


Master Thesis

  • Christophe Rodrigues da Silva, Categorização de imagens e pesquisa de base de dados através de exemplos. Master degree on Electrical and Computer Engineering, FEUP.  (Supervisor: Aurélio Campilho), 2009.


C-BER +351 22 209 4000