Developing a hybrid machine learning model to predict treatment time duration as a workflow regulation tool in public...

This study aimed to design a desktop application that implements machine learning algorithms to predict dental treatment time durations, assess the accuracy of the model, and assess its clinical efficiency. The Python programming language was used to develop software that uses Machine Learning and Google SerpApi service for the prediction process. The sample consisted of 2750 records, 2500 records for training, and 250 records for testing the model. Spearman correlation test result was (r (250) = 0.96, p = < 0.001), the R2 value was (0.97), which means that the actual durations can predict 97.32% of the change in predicted durations, and the Mean Absolute Error metric, yielding a result of 2.6432 min. Age and sex of participants showed no statistically significant effect. The application of Machine Learning is promising in dentistry and the medical field to help regulate the workflow. The integration of the Google SerpApi service was successful and can be helpful in cases without training data. Also, the availability of electronic patient records is necessary in all medical facilities. Finally, Python is a powerful tool in designing software that implements Machine Learning algorithms.