Send to

Choose Destination
Clin Physiol Funct Imaging. 2017 Jan;37(1):30-36. doi: 10.1111/cpf.12264. Epub 2015 Jul 3.

Three-phase bone scintigraphy for diagnosis of Charcot neuropathic osteoarthropathy in the diabetic foot - does quantitative data improve diagnostic value?

Author information

Department of Clinical Physiology and Nuclear Medicine, Herlev Hospital, Herlev, Denmark.
Department of Clinical Physiology Nuclear Medicine & PET, Rigshospitalet, Copenhagen, Denmark.
Department of Clinical Physiology and Nuclear Medicine, Hvidovre Hospital, Hvidovre, Denmark.
Steno Diabetes Center, Gentofte, Denmark.
Health Aarhus University, Aarhus, Denmark.
NNF CBMR University of Copenhagen, Copenhagen, Denmark.



To investigate whether inclusion of quantitative data on blood flow distribution compared with visual qualitative evaluation improve the reliability and diagnostic performance of 99 m Tc-hydroxymethylene diphosphate three-phase bone scintigraphy (TPBS) in patients suspected for charcot neuropathic osteoarthropathy (CNO) of the foot.


A retrospective cohort study of TPBS performed on 148 patients with suspected acute CNO referred from a single specialized diabetes care centre. The quantitative blood flow distribution was calculated based on the method described by Deutsch et al. All scintigraphies were re-evaluated by independent, blinded observers twice with and without quantitative data on blood flow distribution at ankle and focus level, respectively. The diagnostic validity of TPBS was determined by subsequent review of clinical data and radiological examinations.


A total of 90 patients (61%) had confirmed diagnosis of CNO. The sensitivity, specificity and accuracy of three-phase bone scintigraphy without/with quantitative data were 89%/88%, 58%/62% and 77%/78%, respectively. The intra-observer agreement improved significantly by adding quantitative data in the evaluation (Kappa value 0·79/0·94). The interobserver agreement was not significantly improved.


Adding quantitative data on blood flow distribution in the interpretation of TBPS improves intra-observer variation, whereas no difference in interobserver variation was observed. The sensitivity of TPBS in the diagnosis of CNO is high, but holds limited specificity. Diagnostic performance does not improve using quantitative data in the evaluation. This may be due to the reference intervals applied in the study or the absence of a proper gold standard diagnostic procedure for comparison.


blood flow distribution; charcot foot; diabetic neuropathy; dynamic bone imaging; quantitative 99Tc-diphosphonate scintigraphy

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Wiley
Loading ...
Support Center