Boundary detection of retinoblastoma tumors with neural networks

Consistent and accurate measurement of retinoblastoma tumors is of importan t clinical value for treatment management. This paper presents an algorithm for the determination of retinoblastoma (RB) tumor to assist in the determ ination of tumor volume changes throughout treatment periods. The result of the development of a neural network approach for the analysis of three-dim ensional ultrasound images shows that it is possible to identify retinoblas toma tumors and accurately determine the front and back boundary of the tum or. The algorithm used was a soft competitive learning network with two inp uts. The outputs of the network identify the eye, the tumor, and the back o f the eye. (C) 2001 Elsevier Science Ltd. All rights reserved.