Harnessing Machine Learning for Uterine Fibroid Detection Using Advanced Classification Techniques

In the 40 to 50 age range, uterine fibroids are prevalent in over 70% of women. The proposed system employs a dataset comprising 1985 ultrasound images of uterus for fibroid detection. Utilizing advanced classification techniques like ResNet-152 and MobileNet V2, and an ensemble of Logistic Regression and MobileNet V2, the proposed system attained remarkable precision at 87.23% and an impressive accuracy of 91.67%. This outperformed the previous HIFU-Net Segmentation method’s precision of 84.48% on MRI scans, highlighting the potential of advanced techniques in medical diagnostics.