English Türkçe
Covid 19 and Pneumonia AI Diagnostic Test

Usage of MantisCOVID artificial intelligence algorithm that can evaluate on Chest radiography images:

Note: The MantisCOVID artificial intelligence model provided here is trained using the data sets below. This study is currently in a research phase and is not yet designed as ready-to-manufacture models (not for direct clinical diagnostic purposes), the algorithm will be constantly trained as new data is available. MantisCOVID is not recommended for individual diagnosis, this application has been developed to assist physicians in local health institutions.

Upload the chest radiograph image to the system and check the artificial intelligence output. If you are sure of the disease of the person with Chest Radiography image you have uploaded, you can correct the disease, select it and give feedback. Artificial intelligence will be trained and updated at regular intervals.

Reference datasets: COVID-19 image data collection
RSNA Pneumonia Detection Challenge

Start Image Analysis.

By filling out the form, you can send a request for review, then click on the link to be displayed on the interface to access the review screen and review the results of MantisCOVID analysis. Since the personal information you provide is information that will prove your physician expertise, your feedback will help us keep the dataset up to date. Total number of images used in the data set:

8851

Normal

99

COVID-19 Patient

9579

Zatüre Patient

Test results carried out with 150 patient data not included in the training set: (01.04.2020)
There is no specification for pneumonia type. The specificity values for Covid-19 diagnosis are high and low sensitivity values will be increased by updating the dataset.

Send