header advert
Orthopaedic Proceedings Logo

Receive monthly Table of Contents alerts from Orthopaedic Proceedings

Comprehensive article alerts can be set up and managed through your account settings

View my account settings

Visit Orthopaedic Proceedings at:

Loading...

Loading...

Full Access

Oncology

THE DIAGNOSIS OF PRIMARY LONG BONE TUMOURS BY EXPERT SYSTEMS

The European Musculo-Skeletal Oncology Society (EMSOS)



Abstract

Aim

Improving the quality of clinical and radiologic differential diagnosis of intramedullary tumours of long bones.

Methods

A database includes clinical and radiologic (X-ray, CT and MRI methods) signs of 106 patients with osteosarcoma (n = 44), chondrosarcoma (n = 31) and giant cell tumour (n = 31). Multivariate analysis of clinical and radiologic characteristics and developing informative set of criteria (decision rule) for the differential diagnosis of osteosarcoma, chondrosarcoma and giant cell tumour were provided with program «ASTA».

Results

Before examination in Blokhin Oncology Research Centre in 70% of the osteosarcomas and chondrosarcomas and 60% of GCTs the size of the tumour was more than 8 cm. The reason of the late patients' admission to a specialized medical department is inaccurate diagnosis of these tumours. In our study diagnostic accuracy of the differential diagnosis of osteosarcoma, chondrosarcoma and GCT was 89% in case if the decision rules were based on 14 the most informative clinical and X-ray features, 84% if based on 14 clinical and CT features and 88% if based on 9 MRI features. The comparative analysis revealed a high accuracy in determination of these tumours by using decision rules developed on the basis of multivariate analysis of clinical and X-ray criteria.

Conclusion

The comparative accuracy of the developed differential diagnostic criteria (decision rules) of clinical and X-ray, clinical and CT and MRI features proved high informative of each method. The diagnostic accuracy of clinical and X-ray decision rule (89%) exceeded the diagnostic accuracy of radiologist's examination before (62%) and after (83%) admission to Oncology Centre. It proves the necessity for further development and practical application of diagnostic expert systems.