For certain diseases, early detection greatly increases a patient’s chances of survival. This is true for acute leukemia and lymphoma: the sooner they are diagnosed and treatment begins, the more likely it is that the patient will overcome the disease. To achieve this objective, the Hospital Clínic of Barcelona and the Universitat Politècnica de Catalunya (UPC) are working together on the development of CellSimatic, an imaging tool for the detection of severe diseases of the blood.
Morphological comparison of cell types
CellSimatic detects atypical cells by digitally processing images of peripheral blood. The procedure is very straightforward: a sample of the patient’s blood is taken and a drop stained with colouring agents is observed under a microscope, which is connected to a camera in order to record high-resolution images. Then, a software program compares the image against a database and immediately detects whether the cells are anomalous and, if so, what kind they are.
The method will be easy to use, but it requires complex development combining biomedical, computer and mathematical expertise. To automate the detection of malignant cells, it is first necessary to design software that can classify images using artificial intelligence techniques. And to do the classification, there is a prior need to catalogue all cell types in a database that includes images and morphological features for each type. Descriptions will have to be detailed enough to distinguish between different cell types, using data such as the colour, eccentricity, perimeter, texture and area for all the cell types being studied.
The technique will also make it easier to monitor the evolution of the disease and detect any relapses after the completion of treatment
The tool will speed up the diagnosis of diseases of the blood and, as a result, treatment will get started sooner. The technique will also make it easier to monitor the evolution of the disease and detect any relapses after the completion of treatment. Beyond the unquestionable benefits for patients, who will see improvement in their survival and prognosis, the technique will also have an impact on the healthcare system, reducing the number of costlier, more complex complementary studies carried out only in highly specialized labs.
A multidisciplinary team of researchers is developing the method under the leadership of Dr Anna Merino of the Biomedical Diagnostic Centre (CORE) at the Hospital Clínic, and Dr José Rodellar, professor of Applied Mathematics III at the UPC. The project was the recipient of the Innovation Award in the Technology category at the most recent MIHealth Forum, an international event on innovation and healthcare management held this past May in Barcelona. Thanks to a crowd-funding campaign, the project has now received money to acquire a portion of the material that is needed: a digital microscope camera and a software program for image management.