Diagnosis and Nursing Intervention of Gynecological Ovarian Endometriosis with Magnetic Resonance Imaging under Artificial Intelligence Algorithm

Abstract

This research was aimed to study the application value of the magnetic resonance imaging (MRI) diagnosis under artificial intelligence algorithms and the effect of nursing intervention on patients with gynecological ovarian endometriosis. 116 patients with ovarian endometriosis were randomly divided into a control group (routine nursing) and an experimental group (comprehensive nursing), with 58 cases in each group. The artificial intelligence fuzzy C-means (FCM) clustering algorithm was proposed and used in the MRI diagnosis of ovarian endometriosis. The application value of the FCM algorithm was evaluated through the accuracy, Dice, sensitivity, and specificity of the imaging diagnosis, and the nursing satisfaction and the incidence of adverse reactions were used to evaluate the effect of nursing intervention. The results showed that, compared with the traditional hard C-means (HCM) algorithm, the artificial intelligence FCM algorithm gave a significantly higher partition coefficient, and its partition entropy and running time were significantly reduced, with significant differences ( <a:math xmlns:a="http://www.w3.org/1998/Math/MathML" id="M1"> <a:mi>P</a:mi> <a:mo>&lt;</a:mo> <a:mn>0.05</a:mn> </a:math> ). The average values of Dice, sensitivity, and specificity of patients’ MRI images were 0.77, 0.73, and 0.72, respectively, which were processed by the traditional HCM algorithm, while those values obtained by the improved artificial intelligence FCM algorithm were 0.92, 0.90, and 0.93, respectively; all the values were significantly improved ( <c:math xmlns:c="http://www.w3.org/1998/Math/MathML" id="M2"> <c:mi>P</c:mi> <c:mo>&lt;</c:mo> <c:mn>0.05</c:mn> </c:math> ). In addition, the accuracy of MRI diagnosis based on the artificial intelligence FCM algorithm was 94.32 ± 3.05%, which was significantly higher than the 81.39 ± 3.11% under the HCM algorithm ( <e:math xmlns:e="http://www.w3.org/1998/Math/MathML" id="M3"> <e:mi>P</e:mi> <e:mo>&lt;</e:mo> <e:mn>0.05</e:mn> </e:math> ). The overall nursing satisfaction of the experimental group was 96.5%, which was significantly better than the 87.9% of the control group ( <g:math xmlns:g="http://www.w3.org/1998/Math/MathML" id="M4"> <g:mi>P</g:mi> <g:mo>&lt;</g:mo> <g:mn>0.05</g:mn> </g:math> ). The incidence of postoperative adverse reactions in the experimental group (7.9%) was markedly lower than that in the control group (24.1%), with a significant difference ( <i:math xmlns:i="http://www.w3.org/1998/Math/MathML" id="M5"> <i:mi>P</i:mi> <i:mo>&lt;</i:mo> <i:mn>0.05</i:mn> </i:math> ). In short, MRI images under the artificial intelligence FCM algorithm could greatly improve the clinical diagnosis of ovarian endometriosis, and the comprehensive nursing intervention would also improve the prognosis and recovery of patients.