Posted in | News | Medical Robotics

Study Shows Artificial Intelligence can Detect Acute Myeloid Leukemia

In a proof-of-concept study, scientists from the German Center for Neurodegenerative Diseases (DZNE) and the University of Bonn Artificial have shown that artificial intelligence has the ability to detect acute myeloid leukemia (AML)—one of the most common forms of blood cancer-—in a highly reliable manner.

Their method is based on analyzing the gene activity of blood cells. When applied in practice, this method could support traditional diagnostics and probably expedite the start of therapy. The study outcomes have been reported in the iScience journal.

Artificial intelligence is a much-discussed topic in medicine, especially in the field of diagnostics. “We aimed to investigate the potential on the basis of a specific example,” explained Prof. Joachim Schultze, a research group leader at the DZNE and Head of the Department for Genomics and Immunoregulation at the LIMES Institute of the University of Bonn.

Because this requires large amounts of data, we evaluated data on the gene activity of blood cells. Numerous studies have been carried out on this topic and the results are available through databases. Thus, there is an enormous data pool. We have collected virtually everything that is currently available.

Prof. Joachim Schultze, Research Group Leader, DZNE

Fingerprint of Gene Activity

The focus of Schultze and his team was on the “transcriptome,” a specific type of fingerprint of gene activity. In each cell, based on its condition, only specific genes are “switched on” actually, which can be observed in their gene activity profiles. Precisely, such data—extracted from cells in blood samples and including several thousand genes—were studies as part of the present study.

The transcriptome holds important information about the condition of cells. However, classical diagnostics is based on different data. We therefore wanted to find out what an analysis of the transcriptome can achieve using artificial intelligence, that is to say trainable algorithms,” stated Schultze, who is a member of the Bonn-based “ImmunoSensation” cluster of excellence. “In the long term, we intend to apply this approach to further topics, in particular in the field of dementia.”

The focus of this study was on AML. In the absence of sufficient treatment, leukemia of this form causes death within a few weeks. AML involves the proliferation of pathologically modified bone marrow cells, which can eventually make their way into the bloodstream.

In the end, both tumor cells and healthy cells start drifting in the blood. All these cells have characteristic gene activity patterns, which were all taken into account in the analysis. Data from over 12,000 blood samples—which were obtained from 105 different studies—were considered. So far, this is the largest dataset for a metastudy on AML.

About 4,100 of these blood samples were extracted from individuals diagnosed with AML, and the rest were derived from individuals suffering from other diseases or from healthy individuals.

High Hit Rate

Parts of this data set were fed to their algorithms by the researchers. The input included information related to whether or not a sample was taken from an AML patient.

The algorithms then searched the transcriptome for disease-specific patterns. This is a largely automated process. It’s called machine learning,” stated Schultze. On the basis of this pattern recognition, more data was analyzed and categorized by the algorithms, that is, it was classified into samples without AML and with AML.

Of course, we knew the classification as it was listed in the original data, but the software did not. We then checked the hit rate. It was above 99 percent for some of the applied methods. In fact, we tested various methods from the repertoire of machine learning and artificial intelligence. There was actually one algorithm that was particularly good, but the others were close behind.

Prof. Joachim Schultze, Research Group Leader, DZNE

Application in Practice?

According to Schultze, when this approach is put to use, it could enable traditional diagnostics and help save costs. “In principle, a blood sample taken by the family doctor and sent to a laboratory for analysis could suffice. I guess that the cost would be less than 50 euros.” A range of techniques is included in classical AML diagnostics.

The cost of a few of these is a few hundred euros per run, stated Schultze. “However, we have not yet developed a workable test. We have only shown that the approach works in principle. So we have laid the groundwork for developing a test.” Schultze stressed that in the future, AML diagnosis will continue to necessitate expert physicians.

The aim is to provide the experts with a tool that supports them in their diagnosis. In addition, many patients go through a real odyssey until they finally end up with a specialist and get a diagnosis.” This is because, in the early stages, AML symptoms could resemble those of severe cold. But AML is a life-threatening condition that must be treated as early as possible.

With a blood test, as it seems possible on the basis of our study, it is conceivable that the family doctor would already clarify a suspicion of AML. And when the suspicion is confirmed, the patient is referred to a specialist. Possibly, the diagnosis would then happen earlier than it does now and therapy could start earlier.

Prof. Joachim Schultze, Research Group Leader, DZNE


Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type