In recent years, there has been a large increase in the number of companies offering analytical services, some of which are based on AI/machine learning, to assist in the interpretation of radiological images. Medical Connections has been able to help several such companies with the transition from academic research to salable products which can integrate into real-world hospital systems.
The key to such integration is appropriate use of DICOM and occasionally other standards to communicate with the existing hospital Radiology Information System (RIS) and the Picture Archiving and Communication System (PACS). Typically, this involves three discrete steps:
- Identify completed and suitable studies ready for analysis via communication with either the RIS or PACS
- Obtain the image data from the PACS using the DICOM retrieval protocols, including negotiation of compression and (where necessary) decompression of the images to give raw pixel data.
- Create DICOM “objects” (typically images, structured reports or encapsulated PDF documents), and use DICOM to push those to the PACS (and possibly notify the RIS of their existence)
By virtue of its extensive DICOM experience, Medical Connections is well-placed to assist companies in any of these stages, and we are happy to offer the following services, all of which have already been used successfully by existing customers in this field:
- DicomObjects toolkit to simplify the programming of DICOM (available in COM & .NET versions). Latest versions of DicomObjects support both traditional DICOM and the newer DicomWeb transactions.
- Consultancy on which DICOM services to use and how best to use them
- Testing of the conformity of the analytics service and the objects it creates to the relevant standards, and their robustness to unusual but legal data (e.g. with different character sets)
If you would like to discuss what we can offer in this area, please contact firstname.lastname@example.org