MULTILEVEL PRECONDITIONING FOR 3D LARGE-SCALE SOFT-FIELD MEDICAL APPLICATIONS MODELLING

摘要: Soft-field imaging methods, such as Optical Tomography (OT) and Electrical Impedance (EIT) have significant potential for medical they are non-invasive, portable inexpensive. Possible clinical applications include epilepsy monitoring, cerebral stroke differentiation screening breast cancer. Recent advances in data acquisition instrumentation image reconstruction algorithms raise the requirement to handle multiple large datasets from detailed large-scale geometric descriptions of biological objects. Thus, a major bottleneck lies processing number linear equations that result Finite-Element formulation soft-field problems. Common numerical tools not suited problems, therefore alternative approaches required. We propose facilitation an innovative multi-level inverse-based incomplete LU preconditioning approach improve computational efficiency EIT OT system matrices. This combines static reordering scaling, controlled growth inverse triangular factors, approximation Schur-complement scheme. Comparison with conventional factorisation provided speed improvement up 11 times preconditioner setup time, 12 solution runtime models. In addition, new monopolar current sources is introduced. Current sinks represented by combinations compact basis. Only corresponding solutions processed. These serve basis construction entire excitation pattern. exploits information content given optimal manner avoids redundant computation.