A digital labelled image dataset from the Microbiology Laboratory of the Drassanes-Vall d'Hebron International Health and Infectious Diseases Centre. The dataset contains 2571 digital labelled images of clinical thick blood smear samples. Thick blood smear samples were employed for malaria diagnosis, by the identification of the Plasmodium parasite. Labels were applied identifying leukocytes, ring stage Plasmodium trophozoites and mature Plasmodium trophozoites. Thick blood smear images were acquired at 1000x magnification (10x ocular and 100x immersive oil objective lens). A Leica ICC50W integrated digital microscopy camera (5.0 MP) and the digital camera of a Samsung Galaxy S20 (64 MP, 0.8 μm, f/2.0, OIS) smartphone device were employed for image acquisition. Image format files are .jpg in RGB color space and labels format files are .txt. Images and labels are organized in three different folders for convolutional neural networks training and validation. The “train” folder contains 80% of images (2059 images), the “validation” folder contains 15% of images (387 images), and the “test” folder contains 5% of images (125 images). This study was conducted in accordance with the Declaration of Helsinki and approved by the Clinical Research Ethics Committee (CEIm) of the Vall d’Hebron University Hospital/Vall d’Hebron Research Institute with reference number PR(AG)40/2023.

Annotation App (Python), 1.0