|
|
|
| A study of learning curve for high resolution microendoscopy in diagnosis of colon polyps |
| TAN Tao,QU Yawei,SHU Juan,ZHANG Ling,LIU Minli, and LIU Haifeng. |
| Department of Gastroenterology,General Hospital of Chinese People’s Armed Police Force,Beijing 100039,China |
|
|
|
|
Abstract Objective To study the learning curve for high resolution microendoscopy in diagnosis of colon polyps. Methods We selected 140 colon polyp biopsy specimens for the study, based on pre-established HRME diagnostic criteria for colon polyp, and retrospectively analyzed HRME photo features. An experienced HRME physician screened one white light endoscopic picture and four HRME pictures and three observers participated in learning, respectively. Before starting learning, observers needed to learn HRME basics and diagnostic criteria of colonic polyps for 20 cases of colon polyps. Then 120 specimens of HRME images were made to four stages of learning and observers made HRME diagnosis after each phase completed. HRME diagnostic results were compared with pathological findings,observer diagnostic efficiency of colon polyps was calculated to evaluate the consistency between HRME results and pathological findings. Results Four stages observers’ diagnostic sensitivity was 72.2%, 82.4%, 94.4% and 87.5%; specificity was 75.0%, 76.9%, 91.7% and 78.6%,and accuracy was 73.3%, 80.0%, 93.3% and 83.3%. Compared with the first three stages,the differences of sensitivity, specificity and accuracy were statistically significant (P<0.05). Compared with the third stage, the difference was not statistically significant. Four stages had a higher diagnostic consistency, K values were 0.63,0.75,0.89 and 0.81. Conclusions HRME has a high diagnostic value for diagnosis of colon polyps as a new real-time pathology techinique. For beginners, HRME is a simple and easy to master diagnostic technique and one can get a higher diagnostic yield after a short time of learning.
|
|
Received: 13 September 2015
|
|
|
|
|
|
| [1] |
. [J]. Med. J. Chin. Peop. Armed Poli. Forc., 2019, 30(9): 737-740. |
|
|
|
|
2016, Vol. 27 
