“If cancer is suspected at an early stage, radiological imaging examination procedures such as magnetic resonance imaging (MRI) or computed tomography (CT) scanning are often used. Today, only a subsequent biopsy can determine whether abnormal changes are actually associated with a malignant tumor. This involves taking a tissue sample from the affected area and examining it histopathologically in the laboratory. The suspected medical condition can thus be further clarified, confirmed and, ultimately, the appropriate treatment decision can be taken,” explains PD Dr. Dieter Fedders, Chief Physician of Diagnostic and Interventional Radiology at Klinikum Chemnitz gGmbH.

Biopsies are not only time-consuming and resource-intensive for the healthcare system, but they are also an uncomfortable procedure for patients because of the uncertainty involved. Waiting for the results to be confirmed over a period of several days can cause extreme psychological stress.

In order to reduce this waiting time and minimize unnecessary biopsies in the interest of the patient as well as the healthcare system, Fraunhofer ENAS and the hospital in Chemnitz are joining forces: They will work closely together to combine radiological imaging with AI methods to advance individualized cancer diagnostics.

 

Using radiomics to detect cancer

Imaging techniques are essential to modern tumor diagnostics. These techniques can visualize suspicious tissue changes and detect serious illnesses with high accuracy today. Although, this requires skilled medical professionals with a trained eye and years of experience in identifying indications of illnesses in radiological images. However, even the most experienced radiologists often have difficulty detecting very slight changes that are only marginally distinguishable from the surrounding and maybe healthy tissue.

Sophisticated algorithms have the potential to effectively support diagnostics by providing a second opinion: Established AI-based methods enable the analysis of huge amounts of data and the identification of patterns within the data. At present, however, it is often only image data that is analyzed, while other parameters, such as patient and laboratory data, are not considered.

The so called radiomics method, a combination of the words radiology and genomics, attempts to close this gap. This involves extracting detailed image features, such as texture, shape and intensity characteristics, from tumors based on radiological findings, such as CT or MRI images. By comparing the results with laboratory medical findings and their molecular or genetic information, a more reliable and non-invasive diagnosis can be made. By combining all available information, it is possible to precisely differentiate between malignant and benign structures, thus improving the accuracy and reliability of diagnoses – without the need for surgery or biopsy.

“At Fraunhofer ENAS, we have extensive experience in developing sophisticated machine learning methods and AI models. While using mathematical methods in radiological imaging cannot replace medical professionals, it can assist them in their diagnoses with a fast approach that is gentle on patients,” explains Dr. Mario Baum, head of the “Health Systems” department at Fraunhofer ENAS, who initiated the collaboration with the Klinikum Chemnitz gGmbH.

In addition, the radiomics method can help to predict treatment outcomes, which can be utilized to plan individualized treatments. “The analysis of image data in combination with sociological parameters, such as age, gender and place of residence, as well as clinical values, such as underlying medical conditions and survival time, can indicate how well a patient might respond to certain treatments. Based on this information, personalized medicine can be used to create a tailored therapy that is optimally suited to the patient and improve the success of the treatment,” states PD Dr. Dieter Fedders.

Reducing the data to the parameters that actually have a decisive influence on the course of the disease or the prognosis of the treatment, the so-called dimensional reduction, as well as evaluating the interpretability of the results is one of the central research questions that the researchers will address in the course of their collaboration.

In addition to identifying cancers such as lung or breast cancer, medical professionals and researchers hope that radiomics will help them detect other tissue diseases more quickly in the future. This should also make it easier to detect pulmonary embolism, in which blood vessels in the lungs can become blocked, posing a life-threatening risk.

 

Medical phantoms optimize imaging techniques

A second focus of the collaboration between the two Chemnitz-based institutions in the field of radiology is the development of medical phantoms. These substitute models artificially mimic the human anatomy and make it easier for doctors to simulate imaging procedures under realistic conditions without living models before medical technology is approved and thus used on humans.

Fraunhofer ENAS is working together with the Klinikum Chemnitz gGmbH on precisely this kind of dynamic test phantom for computed tomography (CT) to simulate physiological processes in the human body, such as blood circulation.

For example, in order to visualize vascular diseases on CT images even more precisely and conclusively, patients are often administered a contrast agent before the examination. In order to inject this dose in the most optimal way that is gentle on patients, a phantom is to be created that artificially mimics the human vascular system and the distribution of the contrast agent. Fraunhofer ENAS is contributing its expertise to the prototype production of microfluidic and fluidic systems, as well as to their design, implementation and characterization. In addition, integrated sensors will be used to measure both the flow of the blood substitute fluid and the contrast agent, as well as the pressure in the phantom. The results of the collaboration will provide medical professionals with model-based data for optimizing treatment and reducing the use of contrast agents in the future.

“Our close cooperation with the Chemnitz Hospital enables us to develop sophisticated solutions that are directly aligned with medical needs. As a result, our technologies meet specific real-world requirements and will benefit patients at their bedside in the future, improving their quality of life. Early, reliable identification of benign and malignant tissue allows immediate, tailored treatment to be administered in line with a personalized medical approach. With Chemnitz Hospital, we have a strong regional partner at our side with whom we can achieve a milestone in modern diagnostics,” adds the Fraunhofer scientist.

 

A focus on promoting young professionals

Both institutions are also interested in working closely together in the future to support young scientists. Students shall be given the opportunity to gain practical experience in the field of medical technology and clinical research during their studies or to conduct research at the Klinikum Chemnitz gGmbH and Fraunhofer ENAS as part of their thesis.

If you are interested in working with us to find solutions to medical technology challenges, please contact us today.