Machine Learning
Algorithms and Machine Learning are parts of artificial intelligence that have increasing influence throughout the healthcare sector

News from our sponsor: Sectra Denmark

Algorithms and Machine Learning are parts of artificial intelligence that have increasing influence throughout the healthcare sector, among other things within radiology and pathology. We face a number of challenges, both now and in the future, that intelligent use of data can help solve. The population lives longer, and more people are affected by diseases associated with old age. With the sharp rise in medical imagery and the lack of radiologists and pathologists, it is necessary to find ways to work differently to meet the high standards.

"With Machine Learning, we have the opportunity to optimize among other things cancer treatment," says Sune Mark Henriksen, Managing Director of Sectra Denmark." Many believe that the purpose of introducing Machine Learning is to stop educating, for example, radiologists and eventually replacing them. At Sectra, we prefer to support the radiologists in their daily work with Machine Learning."

But what do radiologists say about the prospect of a machine doing a part of the work for them? Radiology is moving more and more towards quantitative imaging; it is easy to imagine a future where more and more types of intervention are judged by a set of national and international accepted criteria. According to a survey conducted in Scandinavia, Benelux and the United States, radiologists are very positive towards the idea that Machine Learning can support their work.

Statement: Automatic characterization and scoring of lesions according to internationally accepted criteria would be valuable to me.

In this case, it is important to consider how automatically generated results are presented and used: For discovery (in support of the radiologist), for diagnosis (auto-generated diagnoses) or something in between. "We believe it is important that the results are presented in such a way that they provide support for the radiologist, who herself makes the final assessment. A Machine Learning application must support the work of the radiologist, do not give the final result," says Sune Mark Henriksen, and continues: "This emphasizes the importance of working closely with the doctors, to ensure that the developed and unrolled Machine Learning applications solve issues that makes sense for those who use the systems, and that it is integrated into their daily workstations in a way that supports their work rather than restricting it."

The Machine Learning area is in rapid development and the last thing that the customer wants is to be locked to one platform/supplier. By choosing a supplier neutral solution, the customer can choose to integrate exactly the Machine Learning application they prefer.

In the future we will see a development of doctors' tools. They will have a larger toolbox and can expect better tools at the same time. The future will also offer the opportunity to choose and deselect services as required. "Today, there is a lot of hype about Machine Learning and it will certainly provide many opportunities, but we cannot predict the future and it will take time," says Sune Mark Henriksen.

Fact Box: Machine Learning:

  • Machine Learning is a form of artificial intelligence
  • Machine Learning can be used for pattern recognition in large amounts of data
  • Sectra is the Nordic region's largest supplier of image management systems for the healthcare sector, such as RIS/PACS/VNA, etc.