Improved Hidden Markov Algorithm Based on Bayes in Low Dose CT Images

The study investigates the application of a Bayesian-based improved hidden Markov algorithm for enhancing low-dose CT images, which are used to reduce patient radiation exposure but suffer from reduced image quality. The improved algorithm was compared to other methods like naive Bayes and traditional hidden Markov models, showing superior performance in terms of image clarity and resolution. Experimental results demonstrated that this algorithm achieved higher Jaccard, Dice, and CCR values across different noise levels, improved the peak signal-to-noise ratio, retained detailed image features better, and enhanced the visual effect by 46.78%. Additionally, it achieved a correct segmentation rate of 95.75%, indicating its effectiveness in improving low-dose CT image quality.