Back

Using Mobile Phones For Computer Aided Detection Of Anomalies On Chest Radiographs: Proof Of Concept

Mitini-Nkhoma, S. C.

2019-11-15 radiology and imaging
10.1101/19012146 medRxiv
Show abstract

Mastering the art of interpreting chest radiographs can be a long and daunting process. We trained a neural network to classify photographs of chest radiographs into 6 groups. The trained model is hosted on a webserver. The model receives and processes photographs of chest radiographs taken using mobile phones. The system then informs the user the most probable diagnosis. Such systems could prove useful in low resource settings which do not have enough highly trained medical personnel and cannot afford the new generation of CAD based radiology equipment.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.