Evaluation of real-time tumor contour prediction using LSTM networks for MR-guided radiotherapy. (May 2023)

• Record Type: Journal Article
• Authors: Lombardo, Elia; Rabe, Moritz; Xiong, Yuqing; Nierer, Lukas; Cusumano, Davide; Placidi, Lorenzo; Boldrini, Luca; Corradini, Stefanie; Niyazi, Maximilian; Reiner, Michael; Belka, Claus; Kurz, Christopher; Riboldi, Marco; Landry, Guillaume
• Is Part Of: Radiotherapy and oncology. Volume 182(2023)
• Publication Date: 2023-05
• Subjects: Artificial intelligence -- Long short-term memory networks -- Time series prediction -- Respiratory motion -- MR-linac -- MLC-tracking; Oncology -- Periodicals; Radiotherapy -- Periodicals; Tumors -- Periodicals; Medical Oncology -- Periodicals; Neoplasms -- radiotherapy -- Periodicals; Radiotherapy -- Periodicals; Radiothérapie -- Périodiques; Cancérologie -- Périodiques; Tumeurs -- Périodiques; Electronic journals; 616.9940642
• DOI: 10.1016/j.radonc.2023.109555
• Languages: English
• ISSNs: 0167-8140

Abstract

Highlights: Comparison of three AI algorithms for real-time prediction of future tumor contours. Usage of clinical cine MRI data from low-field MR-linacs from two institutions. Prediction times compatible with reported MLC-tracking latencies on MR-linacs. Abstract: Background and purpose: Magnetic resonance imaging guided radiotherapy (MRgRT) with deformable multileaf collimator (MLC) tracking would allow to tackle both rigid displacement and tumor deformation without prolonging treatment. However, the system latency must be accounted for by predicting future tumor contours in real-time. We compared the performance of three artificial intelligence (AI) algorithms based on long short-term memory (LSTM) modules for the prediction of 2D-contours 500 ms into the future. Materials and methods: Models were trained (52 patients, 3.1 h of motion), validated (18 patients, 0.6 h) and tested (18 patients, 1.1 h) with cine MRs from patients treated at one institution. Additionally, we used three patients (2.9 h) treated at another institution as second testing set. We implemented 1) a classical LSTM network (LSTM-shift) predicting tumor centroid positions in superior-inferior and anterior-posterior direction which are used to shift the last observed tumor contour. The LSTM-shift model was optimized both in an offline and online fashion. We also implemented 2) a convolutional LSTM model (ConvLSTM) to directly predict future tumor contours and 3) a convolutional LSTM combined with spatial transformer layers (ConvLSTM-STL) to predict displacement fields used to warp the last tumor contour. Results: The online LSTM-shift model was found to perform slightly better than the offline LSTM-shift and significantly better than the ConvLSTM and ConvLSTM-STL. It achieved a 50% Hausdorff distance of 1.2 mm and 1.0 mm for the two testing sets, respectively. Larger motion ranges were found to lead to more substantial performance differences across the models. Conclusion: LSTM networks predicting future centroids and shifting the last tumor contour are the most suitable for tumor contour prediction. The obtained accuracy would allow to reduce residual tracking errors during MRgRT with deformable MLC-tracking.