Abstract: Leukemia can be cured if it is detected and treated at the early stage. The need for automation of leukemia detection arises since current methods involve manual examination of the blood smear as the first step toward diagnosis. This consumes more time and also its accuracy is greatly dependent on operator's ability. So this paper has reviewed commonly used techniques in the prevention of leukemia detection with good accuracy rate. In addition to this various metrics has also been presented to measure the working model capability for leukemia detection.

Keywords: Leukemia, Medical Imaging, Classification, Image Processing, Metrics, Survival Rate.