At a time when Artificial Intelligence(AI) is growing in importance in the field of medicine, a study by researchers from the German Center for Neurodegenerative Diseases (DZNE) and University of Bonn, shows that Artificial Intelligence(AI) is capable of detecting with high precision one of the most common forms of cancer—acute myeloid leukaemia (AML)— proving that the technology can aid in providing improved healthcare facilities.
The study illustrates the potential of AI-based detection in supporting standard methods of cancer diagnosis and the quicker commencement of therapy. "We aimed to investigate the potential on the basis of a specific example," said Prof. Joachim Schultze, a research group leader at the DZNE, in a statement.
What is AML?
It is a form of leukaemia that is associated with the build-up of pathologically mutated bone marrow cells. These cells invariably enter the bloodstream, and both, healthy and tumour cells simultaneously exist in it. The initial symptoms can be as simple as a severe cold. Without timely diagnosis and treatment, AML can result in death in a matter of weeks.
The fingerprint of Gene functions: Transcriptome
Transcriptome can be considered as a type of fingerprint of gene activity. It is important to note that not all genes in all the cells of the body are activated or 'switched on', and the activation occurs only under specific conditions.
This can be seen in the profiles of gene activity, which makes the study of transcriptome crucial. "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," said Schultze
The research focussed on AML and studied data acquired from cells found in blood samples; encompassing thousands of genes found in them. Data from over 12,000 blood samples, which came from 105 different studies, were analysed. Nearly 4,100 of these samples were from those diagnosed with AML, while the others were collected from healthy individuals or those with other diseases.
Talking about the data used, Schultze said, "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."
High rate of accuracy
Machine Learning (ML) algorithms received data inputs such as the origin of the sample from a patient with AML, or from a healthy individual. The algorithms then looked at the transcriptome in the blood samples for disease-specific patterns, and categorised them into samples with and without samples.
"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," said Schultze.
Potential for future applications
If the method sees practical application, it can aid conventional diagnostics and reduce costs. A simple blood test could reveal the existence of AML in the analysed sample. "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," said Schultze.
Stressing on the life-threatening risk of AML, its ability to slip diagnosis, and the need to enable physicians in the timely detection of the study, Schultze added, "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."