Subspecialty Second-Opinion in Multiple Myeloma CT: Emphasis on Clinically Significant Lytic Lesions.
Alberto Stefano TagliaficoLiliana BelgioiaAlessandro BonsignoreFederica RossiGiulia SuccioBianca BignottiAlida DominiettoPublished in: Medicina (Kaunas, Lithuania) (2020)
Background and objectives: In order to increase the accuracy of lytic lesion detection in multiple myeloma, a dedicated second-opinion interpretation of medical images performed by subspecialty musculoskeletal radiologists could increase accuracy. Therefore, the purpose of this study is to evaluate the added value (increased accuracy) of subspecialty second-opinion (SSO) consultations for Computed Tomography (CT) examinations in Multiple Myeloma (MM) patients undergoing stem cell transplantation on standard computed tomography with a focus on focal lesion detection. Materials and Methods: Approval from the institutional review board was obtained. This retrospective study included 70 MM consecutive patients (mean age, 62 years ± 11.3 (standard deviation); range, 35-88 years) admitted in the last six years. Pre-transplant total-body CT (reported by general radiologists) was the only inclusion criteria. Each of these CT examinations had a second-opinion interpretation by two experienced subspecialty musculoskeletal (MSK) radiologists (13 years of experience and 6 years of experience, mean: 9.5 years), experts in musculoskeletal radiology and bone image interpretation with a focus on lytic lesions. Results: Per lesion intra- and inter-observer agreement between the two radiologists was calculated with K statistics and the results were good (K = 0.67: Confidence Inteval (CI) 95%: 0.61-0.78). When the initial CT reports were compared with the re-interpretation reports, 46 (65%) of the 70 cases (95% CI: 37-75%) had no discrepancy. There was a discrepancy in detecting a clinically unimportant abnormality in 10/70 (14%) patients (95% CI: 7-25%) unlikely to alter patient care or irrelevant to further clinical management. A discrepancy in interpreting a clinically important abnormality was registered in 14/70 (21%) patients for focal lesions. The mean diameter of focal lesions was: 23 mm (95% CI: 5-57 mm). The mean number of focal lesions per patient was 3.4 (95% CI). Conclusions: subspecialty second-opinion consultations in multiple myeloma CT is more accurate to identify lesions, especially lytic lesions, amenable to influence patients' care.
Keyphrases
- computed tomography
- end stage renal disease
- multiple myeloma
- image quality
- ejection fraction
- newly diagnosed
- dual energy
- stem cell transplantation
- positron emission tomography
- chronic kidney disease
- patients undergoing
- artificial intelligence
- healthcare
- prognostic factors
- emergency department
- peritoneal dialysis
- magnetic resonance imaging
- high dose
- deep learning
- primary care
- case report
- patient reported outcomes
- soft tissue
- postmenopausal women
- palliative care
- machine learning
- electronic health record