140 likes | 274 Views
A FUZZY EXPERT SYSTEM. DESIGN FOR DIAGNOSIS OF. PROSTATE CANCER. B.HEMAKUMAR. B.Tech, P.G.D.D.I. DEPARTMENT OF BIOMEDICAL SIGNAL PROCESSING & INSTRUMENTATION SASTRA TANJORE. Why fuzzy about Prostate cancer. Of all the Cancers in men, Prostate cancer is one of the most common.
E N D
A FUZZY EXPERT SYSTEM DESIGN FOR DIAGNOSIS OF PROSTATE CANCER
B.HEMAKUMAR B.Tech, P.G.D.D.I. DEPARTMENT OF BIOMEDICAL SIGNAL PROCESSING & INSTRUMENTATION SASTRA TANJORE
Why fuzzy about Prostate cancer • Of all the Cancers in men, Prostate cancer is one of the most common. • Prostate cancer is also one of the most difficult cancers where early diagnosis is very often difficult to make. Hence decisions regarding management is still not standardized. • We have made a FUZZY EXPERT SYSTEM which gives to the user the patient possibility ratio of the prostate cancer.
MATERIALS & METHODS • The patients who had symptoms of prostatic disease were taken up for the study • Patients were randomly selected and were subjected to transabdominal ultrasound (Using 3.5 MHZ Probe) for the study of different zones. Images were obtained in transverse and longitudinal planes • The prostate echo texture, and volume were imaged
MATERIALS & METHODS contd... • The four parameters (Prostate Volume, Echo Texture, Total Acid Phosphatase, Prostate fraction of the Acid Phosphatase) were used as input and Prostate cancer Risk (PCR) was determined using fuzzy expert systems
Sample rules.. • If (ET is N) and (PAP is L) and (TAP is L) and (PV is L) then (PCR is VL) • If (ET is N) and (PAP is VL) and (TAP is VL) and (PV is VL) then (PCR is N) • If (ET is N) and (PAP is VH) and (TAP is VH) and (PV is VH) then (PCR is VH)
TAP (K.A.units) PAP (K.A.units) PV (cc) ET Literature (%) FES (%) 2.47 2.24 1.77 1.59 1.24 1.54 1.45 1.26 1.14 1.04 87.3 80.7 58.9 48.7 34.8 D I N N N 75 –100 50 – 75 25 – 50 0 – 25 NIL 88.8 62.6 37.5 12.4 NIL Comparison of FES & literature..
References • Ismail SARITAS, Novruz ALLAHVERDI and Ibrahim Unal SERT, “AFuzzy Expert System Design for Diagnosis of Prostate Cancer”, International Conference on Computer Systems and Technologies, 2003. • Lorenz A., Blum M., Ermert H., and Senge Th., “Comparison of Different Neuro-Fuzzy Classification Systems for the Detection ofProstate Cancer in Ultrasonic Images”, http://www.lp-it.de/neuro-fuzzy- classification.pdf
References contd... • Zadeh, L. A. (1965). "Fuzzy sets,“ Information and Control, 8:338-353 • 4. http://www.mdsdx.com - Lab news, spring issue, May 2001 • 5. Timothy J.Ross , “ Fuzzy logic with Engineering Applications”, Mc Graw – Hill, Inc., 1997.