عنوان مقاله
کاربرد منطق فازی در تشخیص کامپیوتری سرطان سینه: تحلیل لوبولاسیون
فهرست مطالب
چکیده
مقدمه
رسمی سازی با منطق فازی
درجات لوبولاریته و ریزلوبولاریته
نتیجه گیری
بخشی از مقاله
درجات لوبولاریته و ریزلوبولاریته
رادیولوژیست ها از یک شیوه غیر رسمی برای تعیین لوبولاریته و ریزلوبولاریته (لوبولاریته میکرو) توده استفاده می کنند. ب دو توده به تصویرکشیده شده در شکل 5 را در نظر می گیریم. توده اول دارای حرکات نوسانی عمیق و توده دوم دارای حرکات نوسانی سطحی می باشد.برای رسمی نمودن این وجه تمایز، از معیارهای مختلفی می توان استفاده نمود.
کلمات کلیدی:
Artificial Intelligence in Medicine 11 (1997) 75 – 85Fuzzy logic in computer-aided breast cancerdiagnosis: analysis of lobulation1Boris Kovalerchuka,b, Evangelos Triantaphylloua,b,*,James F. Ruizc, Jane ClaytondaDepartment of Industrial Engineering,Louisiana State Uni6ersity,3128CEBA Building,Baton Rouge,LA70803-6409,USAbStanley S.Scott Cancer Center,Louisiana State Uni6ersity Medical Center,2025Gra6ier Street,Suite622,New Orleans,LA70112,USAcDepartment of Radiology,Woman’s Hospital,9050Airline Highway,Baton Rouge,LA70895,USAdDepartment of Radiology,Louisiana State Uni6ersity Medical Center,1542Tulane A6e.,New Orleans,LA70112,USAReceived 31 October 1996; received in revised form 20 January 1997; accepted 15 February 1997AbstractThis paper illustrates how a fuzzy logic approach can be used to formalize terms in theAmerican College of Radiology (ACR) Breast Imaging Lexicon. In current practice, radiol-ogists make a relatively subjective determination for many terms from the lexicon related tobreast cancer diagnosis. Lobulation and microlobulation of nodules are two importantfeatures in the ACR lexicon. We offer an approach for formalizing the distinction of thesefeatures and also formalize the description of intermediate cases between lobulated andmicrolobulated masses. In this paper it is shown that fuzzy logic can be an effective tool indealing with this kind of problem. The proposed formalization creates a basis for the nextthree steps: (i) extended verification with blinded comparison studies, (ii) the automaticextraction of the related primitives from the image, and (iii) the detection of lobulated andmicrolobulated masses based on these primitives. © 1997 Elsevier Science B.V.Keywords:Fuzzy logic; Feature formalization; Breast cancer; Image recognition; Neuralnetworks* Corresponding author. E-mail: ietrian@lsuvm.sncc.lsu.edu1An early version of this paper was presented in the IEEE-FUZZY ’96 Conference in New Orleans,LA, USA, September 9 – 11, 1996.0933-3657:97:$17.00 © 1997 Elsevier Science B.V. All rights reserve
