Dataset IN is missing a sample in the splits

nx23.jpg from the New Delhi Dataset A is, according to the publisher of the dataset, a non-TB sample that belongs to the training set.

However, according to André and a radiologist, it is an adult with Miliar TB.

It seems like this sample was excluded from the protocols. Might want to have a closer look into this sample and decide on if this should be included in the protocol as TB-positive.