Detection and Classification of Acute Lymphoblastic Leukemia (ALL) from Blood Smear Microscopic Images
Krisha Khandelwal1, Supriya Lande1, Aishwarya Mirashi1, Pooja Kohok2
1Students, Dept. of Computer Engineering, Pune Institute of Computer Technology, Pune, India.
2Faculty, Dept. of Computer Engineering, Pune Institute of Computer Technology, Pune, India.
Abstract—White blood cells (WBC) which are also known as leukocytes, are a crucial part of the human body’s immune system. Leukaemia is a haematological disorder that starts from the bone marrow and affects white blood cells (WBCs) due to mutations in DNA causing blood cell production to become out of control. This prevents the development of healthy blood cells. Acute lymphocytic leukemia (ALL) is a kind of cancer where the bone marrow forms many more lymphocytes. ALL can further be classified into L1, L2 and L3 based on the structure of nucleus and cytoplasm. Automatic detection of leukemia or detection of blood cancer is a tough job and it is very much essential in healthcare centres. ALL identification and interpretation using peripheral blood smear (PBS) pictures plays an important role in early identification in screening and treatment. Conventionally, the method was achieved manually by a skilled technician during a considerable amount of time however the examination of those PBS pictures by laboratory users typically contains diagnostic errors owing to the nonspecific nature of ALL signs and symptoms that usually results in misdiagnosis. So, to achieve more accurate classification results we use powerful segmentation and deep learning techniques to train the model on these images. Firstly, we pre-process the images to apply segmentation methods on them. The proposed model will attempt to eradicate the probability of errors in the manual process and it can be used as a supporting analysis tool for pathologists. So, various Artificial Intelligence-based ALL classification methods and approaches are analysed in a well-defined manner with advantages and disadvantages.
Keywords—Acute Lymphoblastic Leukemia, Classification, Detection, Deep Learning Techniques.
