Fraunhofer Portugal AICOS EDoF Dataset

The Fraunhofer Portugal AICOS EDoF Dataset was produced within the TAMI project and is composed of images of microscopic fields of view (FOV) of Liquid-based Cervical Cytology (LBC) samples. A total of 15 LBC samples were supplied by the Pathology Services from Hospital Fernando Fonseca and the Portuguese Oncology Institute of Porto. For each LBC sample, a set of images were obtained using a version of µSmartScope [1,2] prototype adapted to the cervical cytology use case [3,4].

µSmartScope is a portable 3D-printed prototype based on a smartphone developed by Fraunhofer AICOS that allows the fully automatic acquisition of microscopic images. The acquisition is done using the automated focus approach described in [1], where for each FOV, all images in the precise phase of the approach are stored as well as an indication of the image with the best focus metric (standard deviation of the Tenenbaum gradient [5]).

The dataset is divided into two partitions: the first partition contains the raw data with 5 images per stack without any pre-processing; the second partition contains the pre-processed data according to the best workflow proposed in [6], and also the respective EDoF generated using Complex Wavelets method for the fusion of the microscopy Images. The images from the second partition were pre-processed using chromatic, static, and elastic alignment; they can be used to fully replicate the work in [6].

The size of every image present in this database is 960x720 pixels. For this work, 144 EDoFs images generated with 5 aligned images per stack are used (where the central image is the one with the best focus metric (C). Figure 1 shows an example of a stack and the respective EDoF.

If you find this dataset useful for your research, please cite as: T. Albuquerque, L. Rosado, R. Cruz, M. J. M. Vasconcelos, T. Oliveira J. S. Cardoso, Rethinking Low-Cost Microscopy Workflow: Image Enhancement using Deep Based Extended Depth of Field Methods, In Intelligent Systems with Applications (Elsevier), 2023. https://doi.org/10.1016/j.iswa.2022.200170.