Acute Myeloid Leukemia Fought With New AI Tool

Acute myeloid leukemia (AML) is a rare and aggressive cancer with a high recurrence rate, requiring personalized treatment due to its impact on both blood and bone marrow.
Oncologists need samples from both to determine its genetic profile and best treatment options, often leading to delays in starting lifesaving therapies.
Kiran Vanaja, an assistant research professor at Northeastern University, developed an AI tool to expedite diagnosis and treatment planning for AML.
The AI tool maps genetic mutations within a patient’s cancer cells and employs a neural network to suggest targeted drugs and predict resistance.
Vanaja uses a LEGO analogy to explain how his tool can handle the complexity of gene combinations in cancer cells, enabling swift treatment suggestions in a single night.
The research team trained the model using genotype and phenotype data from AML patients to improve its accuracy in modeling cancer cells’ genetic expressions.
The neural network tool aims to significantly reduce the time from diagnosis to treatment and the researchers are exploring its application to other types of cancers.