Computer-aided classification of anomalies in anatomical structures

A classifier for classifying a candidate anomaly of interest in a virtual anatomical structure, the classifier comprising:

• means for calculating at least one characteristic chosen from the group consisting of the following characteristics associated with the anomaly:
• a characteristic of a neck of the anomaly,
• a normalized wall thickness of the anomaly, and
• a similarity between the anomaly and a digital template;
• and
• means for classifying the anomaly based on the at least one characteristic.

A machine comprising:

• one or more microprocessors coupled to memory;
• wherein the one or more microprocessors are programmed to process a digital representation of an anomaly in a colon by:
• in a digital representation of the colon comprising the digital representation of the anomaly, matching a digital template to the digital representation of the anomaly, wherein the matching generates a similarity score, and the matching matches without regard to particular intensity values;
• and
• based at least on the matching, classifying the digital representation of the anomaly, wherein the classifying comprises feeding the similarity score to a classifier, which considers at least the similarity score when classifying the digital representation of the anomaly;
• wherein the digital template represents a hemispherical idealized anomaly of interest, a round plateau idealized anomaly of interest, or an idealized classic polyp model.