AI in the Industry – Saving Lifes in Medical Technology

Today we would like to introduce another part of the „AI in Industry“ series. This time, however, it is not so much about numbers and profit margins, but rather about a far more important good in our lives: health. AI can support the diagnosis and treatment of diseases and thus speed up the healing process. This requires high amounts of data for a valid statement and high accuracy, which can be achieved using data mining, for example.

Data mining is the process of analyzing large amounts of data from different perspectives and summarizing it into valuable information. Data mining plays an important role in the field of information and communication technology. The medical technology industry generates huge complex data about patients, hospital resources, diseases, diagnosis methods, electronic patients record, etc.

Data analysis is a critical factor for medical diagnosis. Artificial neural networks (ANNs) have proven suitable for the diagnosis of various diseases. However, ANNs are trained by a suitable database of example cases. For this reason, the quality of training and diagnosis results strongly depend on the database used. Database quality becomes a challenge. Besides, the noise in the data can not only impact the normal uncertainty of the measured data but also can impact secondary factors, such as the coexistence of more than one disease. It cannot be predicted unless crossover effects are taken into account when building the training database.

With an appropriate selection of data, tumors or signs of disease can then be detected that may not be detectable by humans. This leads to earlier treatment and thus to higher chances of success and milder disease progression. Even in the current pandemic, such AI applications can enable stronger containment!