Classification of Brain tumor
12 ビュー (過去 30 日間)
I am working on a project of Brain tumor detection. I have extracted the tumor using k means clustering, can anyone tell me how can i classify the tumor as benign or malignant, or calculate the stage of tumor depending upon the features like area, solidity etc. during searching i have found about Knnclassify, can any one tell me how can i use it
回答 (3 件)
Image Analyst 2012 年 7 月 10 日
Although I've worked in radiology, I'm not a radiologist. I suggest you ask one to see if there are any visual clues that can distinguish a benign tumor from a malignant one. Or ask a pathologist if you're looking at biopsies instead of radiological images.
When I worked in radiology they didn't even like software to claim to make diagnoses (usurping their job). Software engineers were also reluctant to do so because of liability concerns. Would you want your software to claim it was benign, only to be sued when it turned out the patient died because it was actually malignant?
Ryan 2012 年 7 月 14 日
In addition to IA's answer, this seems to be a question requesting a novel idea of how to differentiate tumor types accurately using current MRI technology. Honestly, if I knew how to do that I would not tell you how to do it and would instead write the program myself and copyright it.
Walter Roberson 2012 年 7 月 22 日
The people I worked with found that it was much more effective to make classifications based upon MRS (Magnetic Resonance Spectroscopy) rather than MRI (Magnetic Resonance Imaging). We pretty much dropped automated MRI classification somewhere around a dozen years ago. The clues we were able to dig out of the relative chemical composition (MRS) often found signature chemical differences that could not be visually distinguished.