A framework for prediction of personalized pediatric nuclear medical dosimetry based on machine learning and Monte Carlo techniques

Abstract

Goal: A methodology is introduced for the development of an internal dosimetry prediction toolkit for nuclear medical pediatric applications. The proposed study exploits Artificial Intelligence techniques using Monte Carlo simulations as ground truth for accurate prediction of absorbed doses per organ considering personalized anatomical characteristics of any new pediatric patient.
Method: GATE Monte Carlo simulations were performed using a population of computational pediatric models to calculate the specific absorbed dose rates (SADRs) in several organs. A simulated dosimetry database was developed for 28 pediatric phantoms (age range 2-17 years old, both genders) and 5 different radiopharmaceuticals. Machine Learning regression models were trained on the produced simulated dataset, with Leave One Out Cross Validation for the prediction model evaluation. Hyperparameter optimization and ensemble learning techniques for a variation of input features were applied for achieving the best predictive power, leading to the development of a SADR prediction toolkit for any new pediatric patient for the studied organs and radiopharmaceuticals.
Main results: SADR values for 30 organs of interest were calculated via Monte Carlo simulations for 28 pediatric phantoms for the cases of five radiopharmaceuticals. The relative percentage uncertainty in the extracted dose values per organ was lower than 2.7%. An internal dosimetry prediction toolkit which can accurately predict SADRs in 30 organs for five different radiopharmaceuticals, with mean absolute percentage error on the level of 8% was developed, with specific focus on pediatric patients, by using Machine Learning regression algorithms, Single or Multiple organ training and Artificial Intelligence ensemble techniques.
Conclusion: A large dosimetry simulated database was developed and utilized for the training of Machine Learning models. The developed predictive models provide very fast results (< 2 sec) with an accuracy >90% with respect to the ground truth of Monte Carlo, considering personalized anatomical characteristics and the biodistribution of each radiopharmaceutical. The proposed method is applicable to other medical dosimetry applications in different patients’ populations.

A radiomic‐ and dosiomic‐based machine learning regression model for pretreatment planning in 177Lu‐DOTATATE therapy

“…Both ML and DL based treatment plans have shown remarkable predictive ability in nuclear medicine. 36,69,70 We aimed to develop a model that can predict the absorbed dose from pretreatment imaging sessions but additionally identify patient-and dose-specific features for the purpose of discovering possible correlations between these features. These features can only be identified using ML-based models.…”

Theranostics and artificial intelligence: new frontiers in personalized medicine

“…However, traditional MIRD methods relied on a simplified, non-specific model of a 70 kg adult male or female as a phantom for these dosimetry calculations, resulting in a lack of patient-specific details 54 . Another prime example of dosimetry techniques is Monte Carlo (MC) simulations 55 . Monte Carlo simulations provide in-depth predictions of radiation dose distributions within the human body, taking into account complex bodily structures and radiation-tissue dynamics 56 .…”

Dosimetric Evaluation of 177Lu Peptide Receptor Radionuclide Therapy Using GATE and Planet Dose

“…One of the most popular methods is the use of Monte Carlo (MC) simulations for dosimetry in both diagnostic and therapeutic applications [8,9]. Numerous studies have utilized MC simulations to calculate the absorbed doses for patients undergoing therapy with 177 Lu [10][11][12][13][14][15][16]. MC simulations offer high-accuracy patient-specific dose calculations and are considered as the gold standard [17][18][19].…”