Abstract
Because current methods of imaging prostate cancer are inadequate, biopsies cannot be guided effectively and treatment cannot be planned, targeted or monitored optimally. Therefore, our research is aimed at ultrasonically characterizing cancerous prostate tissue so that we can image it more effectively and thereby provide improved means of detecting, treating, and monitoring prostate cancer. We base our characterization methods on spectrum analysis of radiofrequency (RF) ultrasonic (US) echo signals combined with clinical variables such as prostate-specific antigen (PSA). We classify parameters of US spectra (USS) using artificial neural networks (ANNs), and express classification efficacy using relative-operating- characteristic (ROC) methods. The US methods produce ROC- curve areas of 0.84 compared to 0.64 for conventional methods. We generate lookup tables (LUTs) that translate spectral parameters and clinical variables directly to pixel values in tissue-type images (TTIs) that show cancerous regions in 2- or 3- D. Based on encouraging results obtained by others using magnetic-resonance spectroscopy (MRS) for identifying cancerous prostate lesions, we are investigating means of combining MRS with USS methods to produce multimodality TTIs.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have