3-D tissue typing of prostate tissue based on spectral parameters

ernest j feleppa,william r fair,t liu,neil e fleshner,victor e reuter,andrew kalisz,michael s cookson,a rosado
DOI: https://doi.org/10.1109/ULTSYM.1995.495769
1995-01-01
Abstract:Prostate cancer is the highest-incidence and second-leading cancer killer of men in the US. Its diagnosis now relies virtually exclusively on biopsy, but because of the poor specificity and sensitivity of conventional ultrasound many unnecessary biopsies are performed and many justified biopsies fail to sample cancer that is present. Ultrasonic spectrum analysis of radio-frequency (rf) echo signals may offer a means of improving the sensitivity and specificity of ultrasonic means of guiding biopsies. These improvements could reduce unnecessary biopsies, and particularly, false-negative biopsies by improving biopsy guidance. Our results are encouraging for differentiating cancerous from non-cancerous tissue, they typically produce ROC curve areas of approximately 0.9, which represents sensitivity and specificity superior to other imaging means of determining whether biopsy is required
What problem does this paper attempt to address?