Innovative Long Short-Term Memory is a more accurate and efficient method of predicting the stage of ovarian cancer than Multi-Layer Perceptron

AIM: using a unique long short-term memory algorithm (LSTM) over Multi-Layer Perceptron (MLP), analyse various medical ovarian cancer colposcopy images for the purpose of predicting the presence and stage of ovarian cancer. Materials and Methods: The sample size for two groups is 30. In group 1, 15 samples were collected for the Multi-Layer Perceptron (MLP) algorithm, and 15 samples were collected for the Long Short-Term Memory (LSTM) method. To assess, contrast, and comprehend the correctness of suggested algorithms, roughly 15 medical samples were used in the SPSS study. The SPSS software use a G power of 80% to predict outcomes accurately. The metrics assessed comprise 95% confidence intervals and an alpha of 0.001 and a 0.001 significance value. Results: The analysis of the sample colposcopy photos for the presence of ovarian cancer and its stage using both algorithms is successful, and results are obtained based on the SPSS analysis with alpha = 0.001 and Gpower = 0.8, Conclusion: The MLP algorithm has obtained an accuracy rate as 90.17 % and the innovative LSTM has obtained the accuracy rate as 92.67 %.