Artificial intelligence and infrared imaging can automatically classify tumors and are faster than traditional methods. Scientists from the Ruhr University Bochum announced the successful application of the technology in a press release.
Advances in therapeutic options in recent years have significantly improved the chances of a cure for colorectal cancer patients. However, new approaches such as immunotherapies require accurate diagnosis so that they can be tailored specifically to the individual.
Researchers at the Center for Protein Diagnostics (PRODI) use artificial intelligence in combination with infrared imaging to optimally tailor colorectal cancer therapy to individual patients. The automated method can supplement existing pathological analyses. The team led by Professor Klaus Gerwert reported on it in January in the “European Journal of Cancer”.
Fighting brain tumors with an implant and electrical stimulation
Many types of cancer have become increasingly treatable over the years. Unfortunately, this does not apply to brain cancer.
Human tissue
The PRODI team has developed this new method in recent years. The so-called label-free infrared (IR) imaging measures the genomic and protein composition of the examined tissue. This information is decoded and displayed as images using artificial intelligence. To this end, the researchers use image analysis methods from the field of deep learning.
In collaboration with clinical partners, the PRODI team was able to demonstrate that analyzing certain neural networks makes it possible to accurately determine the relevant parameters in colorectal cancer. It is a standardized, user-independent, automated process and a classification of the tumor can be made within an hour.
Effectiveness of the therapy
In classical diagnostics, the so-called microsatellite status (how much abnormality there is in the DNA of the cancer cells) is determined by immunostaining of various proteins or by DNA analysis. “15 to 20 percent of patients with colorectal cancer show microsatellite instability in the tumor tissue,” says Professor Andrea Tannapfel, head of the Institute of Pathology at Ruhr University. “This instability is a positive biomarker indicating that immunotherapy will be effective.”
With the ever-improving therapeutic options, the rapid and uncomplicated determination of such biomarkers is also becoming increasingly important. Neuronal networks were adapted, optimized and trained at PRODI on the basis of IR-microscopic data in order to arrive at a diagnosis. Unlike immunostaining, this approach does not require dyes and is significantly faster than DNA analysis. “We were able to demonstrate that the accuracy of IR imaging for determining microsatellite status is close to the most commonly used method in the clinic, immunostaining,” says PhD candidate Stephanie Schörner. “Through further development and optimization of the method, we expect it to become even more accurate,” adds Dr. Frederik Grosserüschkamp.
