Back

Liver Echinococcosis Lesion Classification Tool by Deep Learning: development, deployment, and validations

Huang, L.; Xi, X.; Chen, Z.; Zeng, Y.

2022-01-30 infectious diseases
10.1101/2022.01.27.22269985 medRxiv
Show abstract

The endemic of Echinococcosis imposed heavy disease burden in some areas. The sonography for Echinococcosis lesions was essential to disease diagnosis and managements. Especially the biological typing of lesions was key to disease treatments. We used deep-learning tools to help sonographer to classify the lesion types. The model achieved 85%(302/376) accuracy, in contrast to senior sonographer achieved 72%(61/85) accuracy. The accuracy of AI model was higher than senior sonographer (p-value=0.01), could be a feasible method to help sonographer in remote area.

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.