Ultrasonography of the hand

Abstract

Purpose: A joint project of the Departments of Hand Surgery and Radiology at the same hospital in Stockholm to place during eight years from 1989 to 1996. The aim of the project was to evaluate the possible benefits of diagnostic ultrasound in connection with specific clinical problems in hand surgery. Material: Originally a total of 281 patients was evaluated with ultrasound: one with an inflammatory disease, 68 with a history of trauma, and 212 which at clinical investigation or at ultrasound had a benign soft tissue tumour. They consisted of 63 ganglions, 22 glomus tumours, 38 haemangiomas, 24 lipomas, 34 nerve tumours, 24 localized villonodular synovitis, and 7 miscellaneous lesions. A number of 231 patients were retrospectively evaluated and 50 prospectively.

Method: In two investigations a Toshiba SSA-9OA ultrasound machine with a linear 7,5 MHz transducer was used. The remaining investigations were performed with an Acuson XP with a 7 MHz linear transducer. All but four examinations were performed by the same radiologist. In the 234 cases ultimately analysed the diagnoses were settled at operation and in the great majority of soft tissue tumours also by histopathological examination. In addition to basic statistics, a statistical model founded on odds quotients was set up to evaluate the different ultrasound characteristics. The model predicted the alternative diagnosis as first and second choice for each tumour and for the whole material.

Results: The investigation of trauma was focused on demonstrating the degree of dislocation of the ulnar collateral ligament of the first metacarpo-phalangeal joint. In 82% the ultrasound findings were in good agreement with those revealed at operation. Chi square analysis showed highly significant correlation between the ultrasound diagnosis and the operative findings. In one investigation 38 patients with clinically suspected ganglions were examined. Thirty-four of them were correctly diagnosed as ganglions at ultrasound and surgically confirmed. The sensitivity was 97% and the specificity 100%. In our retrospective investigation of five soft tissue tumour groups the statistical model predicted the correct diagnosis, with first and second alternatives added, for each tumour between 50 and 92 %, and for the whole material in 77 %. In our prospective investigation the reported ultrasound diagnosis was first correlated to histopathology, giving a sensitivity of 84%. Secondly, in the statistical model the correct diagnosis with first and second alternatives was predicted in 56-100% for each tumour and for the whole material in 63%. Compiling our retrospective investigation of 96 soft tissue masses with our prospective material of 50, the statistical model predicted the first together with the second diagnosis alternative correctly in 68-90% and for the whole material in 79%.

Conclusion: Ultrasonography can with good accuracy correctly diagnose dislocated ulnar collateral ligaments of the thumb, as well as ganglions and 5 groups of benign soft tissue masses.