Back

Detecting adrenal lesions on 3D CT scans using a 2.5D deep learning model

Poon, S. T.; Hanna, F. W.; Lemarchand, F.; George, C.; Clark, A.; Lea, S.; Coleman, C.; Sollazzo, G.

2023-02-24 radiology and imaging
10.1101/2023.02.22.23286184 medRxiv
Show abstract

Many cases of adrenal lesions, known as adrenal incidentalomas, are discovered incidentally on CT scans performed for other medical conditions. Whilst they are largely benign, these lesions can be secretory and/or malignant. Therefore, early investigation is crucial to promptly and efficiently manage those requiring intervention whilst to reassuring the remaining majority in a timely manner. Traditionally, the detection of adrenal lesions on CT scans relies on manual analysis by radiologists, which can be time-consuming and unsystematic. Using AI and deep learning, we examined whether or not applying these technology can augment the detection of adrenal incidentalomas in CT scans. We developed a 2.5D deep learning model to perform image classification on 3D CT scans of patients to classify between lesion and healthy adrenal glands. When tested on an independent test set, our 2.5D model obtained an AUC of the ROC curve of 0.95, and a classification sensitivity of 0.86, and specificity of 0.89. These results suggest that deep learning may be a promising tool for detecting adrenal lesions and improving patient care.

Matching journals

The top 8 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.