Case Study
Al algorithms development to speed up the diagnosis of rare lung diseases (Idiopathic pulmonary fibrosis, IPF).
Use Case Pathology: AI can help to speed up the work of the pathologist
Slide to see the complete image

In its first version, the ‘Komombo’s doctor assistant solution' will analyse and provide insights to help in the diagnosis of the rare disease “idiopathic pulmonary fibrosis”.  


The utility is focused on enabling Computer Vision capabilities. Komombo will assist in spotting shapes, patterns and the relative location of objects on an image, etc.

 

For this reason, ‘Komombo’s doctor assistant’ is set to be useful to any medical practitioner that analyzes images to better understand the anatomy of either individual patients or population groups. 


Pathologists, radiologists, epidemiologists, biomedical engineers, data scientists, medical researchers, and clinicians would benefit from our platform.  


In this specific case,  Komombo looks for fiber density, group locators -as rare samples are located in a specific part of the organ-, grouping of granulomes, and detect unhealthy lungs through an X-ray. 

 

As a result of the analysis, the user gets an annotated image. This means that the original image is shown with marks that are highlighting the areas where the AI engine has spotted the signs that lead to a diagnosis of the illness.  

 

Over time, Komombo will assist in processing medical images from several sources such as Computed Tomography (CT), Magnetic Resonance Imaging (MRI) scanner, Computerized Axial Tomography (CT Scan), X-Ray and ultrasound images, biopsies, etc. 

Komombo makes possible to detect early changes in lung tissue before the disease progresses and to apply treatment as soon as possible, ultimately improving the quality of life of patients and/or increasing their survival.


This annotated input will then be read by the specialist, who doesn’t need to spend a huge amount of time searching for traces of the illness anymore, but instead analyses the traces spotted by ‘Komombo’s doctor assistant’.  


Along with the annotated image, a degree of confidence on the answer will be shown; so that, the expert doctor can decide whether to complement the human diagnosis with what the ‘assistant Komombo’ has found. The doctor finally validates the information given by Komombo to make the final diagnosis.