Projects

ALL Detection Based on Deep Learning

  • Acute Lymphoblastic Leukemia detection based on adaptive unsharpening and Deep Learning: [GitHub]
  • Histopathological transfer learning for Acute Lymphoblastic Leukemia detection: [GitHub]
  • ALLNet: Acute Lymphoblastic Leukemia detection using lightweight convolutional networks: [GitHub]
  • ALL-IDB Patches: Whole slide imaging for Acute Lymphoblastic Leukemia detection using Deep Learning: [GitHub]
  • DL4ALL: Multi-task cross-dataset transfer learning for Acute Lymphoblastic Leukemia detection: [GitHub]
  • A decision support system for Acute Lymphoblastic Leukemia detection based on Explainable Artificial Intelligence: [GitHub]

Acute Lymphoblastic (or Lymphocytic) Leukemia (ALL) is a disease that affects the blood cells, can spread rapidly throughout the body, and may result in fatal consequences if not detected at an early stage. One of the techniques routinely used to diagnose ALL consists in analyzing White Blood Cells (WBC) present in peripheral blood samples to look for malformations or abnormalities. Such malformations may be an indicator of lymphoblasts, which naturally occur in the bone marrow. However, an elevated number of WBCs with lymphoblast characteristics may be a sign of ALL.

Traditionally, the analysis of the WBC morphology is performed manually by an expert pathologist, who looks at the blood cells and estimates the concentration of lymphoblasts present in peripheral blood. Such process, being extremely repetitive and time-consuming, may lead to fatigue, with the consequence that the pathologist could miss important information correlated with the presence of ALL.

To overcome the disadvantages of a manual inspection process, Computer Aided Diagnosis (CAD) systems are being increasingly researched: such systems are often based on image processing and machine learning and, by automatically detecting lymphoblasts, can help the pathologist in performing a preliminary screening of the blood samples. Among CAD systems, recent methods are increasingly considering the use of machine learning approaches based on Deep Learning (DL) and Convolutional Neural Networks (CNN), due to their high accuracy in several fields, including medical imaging. In particular, CNNs have the ability of automatically learning data representations, without the need for a handcrafted feature extraction step, with the consequence that CAD systems based on CNNs may be designed with limited knowledge of the application scenario.



Project page: https://iebil.di.unimi.it/cnnALL/index.htm