New technology can predict which patients will achieve remission and which will relapse.
Investigators have developed the first computer machine-learning model with the potential to accurately predict which patients with acute myelogenous leukemia (AML) will achieve remission and which will relapse.
In a study published in the IEEE Transactions on Biomedical Engineering, investigators trained the computer using bone marrow data and medical histories of patients with AML, as well as blood data from healthy individuals.
“It’s pretty straightforward to teach a computer to recognize AML, once you develop a robust algorithm, and in previous work we did it with almost 100% accuracy,” said senior author Murat Dundar. “What was challenging was to go beyond that work and teach the computer to accurately predict the direction of change in disease progression in AML patients, interpreting new data to predict the unknown: which new AML patients will go into remission and which will relapse.”
The investigators used cases in which the computer had no information and evaluated them using the algorithm by applying knowledge about similar cases in the database, according to the study.
The computer was then able to predict remission with 100% accuracy, while relapse was accurately predicted in 90% of relevant cases.
“As the input, our computation system employs data from flow cytometry, a widely utilized technology that can rapidly provide detailed characteristics of single cells in samples such as blood or bone marrow,” said first author Bartek Rajwa. “Traditionally, the results of flow cytometry analyses are evaluated by highly trained human experts rather than by machine-learning algorithms. But computers are often better at extracting knowledge from complex data than humans.”
The authors noted that automated measurement and monitoring of patient responses to AML treatment is crucial for objective evaluation of the status of disease progression, as well as for timely assessment of treatment strategies.
“Machine learning is not about modeling data,” Dundar said. “It’s about extracting knowledge from the data you have so you can build a powerful, intuitive tool that can make predictions about future data that the computer has not previously seen—–the machine is learning, not memorizing––and that’s what we did.”