AI Tool Accelerates Diagnosis and Treatment Strategies for Acute Myeloid Leukemia
A new artificial intelligence tool developed by researchers at Northeastern University is poised to dramatically reduce the time it takes to diagnose and determine treatment plans for acute myeloid leukemia (AML), a rare and aggressive cancer. The tool, recently patented, offers a potential lifeline for patients facing a disease with a high recurrence rate and limited one-size-fits-all treatment options.
Understanding the Challenges of AML
Acute myeloid leukemia, or AML, affects individuals of all ages. According to Kiran Vanaja, an assistant research professor in bioengineering at Northeastern University, AML presents significant challenges due to its high recurrence rate and the absence of a universal treatment approach . Diagnosing AML requires analyzing samples of both blood and bone marrow to understand the disease’s unique genetic makeup and identify the most appropriate treatment. This process can often take a month or more, a critical delay given the cancer’s aggressive nature and a median survival rate of less than five years after initial diagnosis .
How the AI Tool Works
The AI tool developed by Vanaja’s team at Northeastern’s Roux Institute in Portland, Maine, aims to compress this timeline to as little as a single night. The platform first maps the genetic mutations driving a patient’s AML. It then utilizes a neural network – a computational model inspired by the human brain – to suggest potential drugs and predict the likelihood of drug resistance .
Vanaja explains that understanding the interplay between a cell’s genetic code (genotype) and its observable characteristics (phenotype) is crucial. Cancer cells, in their attempt to survive, undergo significant internal changes, creating a disconnect between genotype and phenotype. The neural network is designed to untangle these complex relationships, considering the vast number of possible gene combinations.
The Power of Deep Learning
The tool leverages deep learning, a type of artificial intelligence, to analyze data from thousands of cells from approximately a dozen patients, combined with data from existing scientific studies. This allows the model to accurately identify patterns and predict treatment responses . The AI’s ability to rapidly process and analyze these complex genetic interactions is what sets it apart from traditional methods like gene and RNA sequencing.
Beyond AML: Future Applications
While initially focused on AML, Vanaja believes the underlying AI network has broader applications. Its core function – connecting genotype to phenotype – is applicable to a wide range of diseases, including solid tumors. Ongoing research will focus on expanding the model’s capabilities and validating its predictions with real-world clinical data .
Researchers are also exploring strategies to prioritize drug combinations for relapsed acute myeloid leukemia using machine learning .