Inobitec DICOM Viewer

Chapter 11. Lung Damage Assessment with the Help of Artificial Intelligence


Functionality is available in a separate module which is activated in the Pro edition for an extra fee

The purpose of the module described in this chapter is to give the users (medical staff) an opportunity to obtain diagnostic information concerning lung damage resulting from COVID-19 from independent external sources with the help of artificial intelligence. The work of the AI is based on neural networks and with a high degree of probability detects abnormalities helping the doctor to establish a correct diagnosis. The studies are carried out on the basis of anonymized data. Automated analysis provides an opportunity to save the time required for establishing a diagnosis and assessing the lung damage.

The results of the assessment of lung parenchyma damage (on the whole, as well as for the right/left lung or each lobe of the lung separately) are presented as % of the corresponding lung volume (on the whole, as well as for the right/left lung or each lobe of the lung separately).

PIC Attention! By contrast with the method of assessment on the scale of lung parenchyma involvement in pathological processes, which is convenient for visual assessment, we do not calculate the score, which is then converted into %. It must be taken into account by the specialists interpreting the results, as when you use a scoring system, the resulting figures are normally greater than when you calculate the damage directly.
PIC The lung damage assessment carried out with the help of artificial intelligence does not constitute a diagnosis and must be confirmed by a specialist.

As a result of the assessment, you will get:

  • a report in the form of a PDF document;

  • an image depicting the abnormalities marked by different colors.