Liu, Hongcheng

Abstract Purpose This study investigates a novel optimization framework that integrates the newly developed MCF MKM relative biological effectiveness (RBE) model into intensity modulated particle therapy (IMPT) planning for carbon ion therapy. Unlike conventional RBE models such as MKM or LEM, the MCF MKM provides detailed microdosimetric information, which improves predictive accuracy of cell survival but elevates both memory and time requirement in the optimization process. Furthermore, to the implicit nature of the MCF MKM, the gradient information, which is essential in inverse planning, is hardly accessible. In response, a randomized gradient‐free optimization method (RGFM) was developed to mitigate the computational challenge and achieve improved RBE dose distributions. Materials and Methods An RBE dose optimization problem was formulated based on the discretized MCF MKM, known as the Abridged Microdosimetry Distribution Model (AMDM). The RGFM was derived to find feasible and high‐quality settings of monitor unit for IMPT plans without the need for explicit gradient calculations. To verify the clinical plausibility of the proposed method, it was tested on two clinical scenarios—prostate and brain cases in comparison with benchmark plans derived solely from physical dose optimization. Computational times and dose‐volume histograms (DVHs) for both physical and RBE doses were evaluated. Results The RGFM algorithm successfully optimized the MCF MKM‐based IMPT plans. Compared to the benchmark, the RBE dose distributions showed improved coverage of the target and reduced hot spots. Such improvements were observed to coincide with trade‐offs in physical dose distributions. The complexity of the MCF MKM extended the computational time, yet high‐quality solutions were attainable in a clinically reasonable timeframe. Multi‐field optimization (MFO) further enhanced target coverage and critical organ sparing over single‐field optimization (SFO). Conclusions This proof‐of‐principle study confirmed the feasibility of using a randomized gradient‐free method to optimize RBE dose based on the MCF MKM in IMPT planning for carbon ion therapy. The approach provides a pathway for clinical adoption of advanced RBE models, potentially leading to more biologically informed and effective carbon ion therapy treatments. Future work involves refining algorithms, incorporating parallel computing, and performing additional case studies to improve computational efficiency and generalizability.

Abstract Background Dose‐driven continuous scanning (DDCS) enhances the efficiency and precision of proton pencil beam delivery by reducing beam pauses inherent in discrete spot scanning (DSS). However, current DDCS optimization studies using traveling salesman problem (TSP) formulations often rely on fixed beam intensity and computationally expensive interpolation for move spot generation, limiting efficiency and methodological robustness. Purpose This study introduces a Break Spot‐Guided (BSG) method, combined with two acceleration strategies—dose rate skipping and bounding—to optimize beam intensity while minimizing beam delivery time (BDT). In addition, a ‐method is proposed for efficient move spot generation, aiming to balance computational efficiency and dose calculation accuracy. Methods The BSG framework simplifies the original mixed‐integer nonlinear programming (MINLP) problem by transforming it into a series of TSP evaluations using break‐spot dose rates. The two acceleration strategies are integrated to reduce computational complexity. The ‐method is introduced to efficiently generate move spots without compromising dose fidelity. The performance of the BSG method, with and without the acceleration strategies, and the ‐method, was evaluated across multiple clinical cases. Results By jointly optimizing the scan path and dose rate, the proposed BSG‐S‐B method offered the most efficient delivery, achieving a 64% reduction in BDT compared to DSS and a 57% reduction compared to Liu's method in the prostate case. These gains were consistently observed across all cases. Utilizing acceleration strategies, BSG‐S‐B significantly reduced computation time from hours to minutes (e.g., from 10 525.7 to 108.5 s in the prostate case, and from 6202.4 to 129.8 s in the head‐and‐neck case). When combined with the ‐method, it reduced the number of move spots by more than 50%, while still ensuring dosimetric quality. Conclusions The proposed framework significantly improves computational efficiency, making scan path optimization with dose rate selection feasible for clinical DDCS delivery. The ‐method effectively reduces computational complexity while maintaining dose calculation accuracy, demonstrating strong potential for clinical adoption in proton therapy treatment planning.

AbstractBackgroundIntensity Modulated Proton Therapy (IMPT) is a sophisticated radiation treatment allowing for precise dose distributions. However, conventional spot selection strategies in IMPT face challenges, particularly with minimum monitor unit (MU) constraints, affecting planning quality and efficiency.PurposeThis study introduces an innovative Two‐Stage Mixed Integer Linear Programming (MILP) method to optimize spot intensity in IMPT with Lower Bound (LB) constraints. This method seeks to improve treatment planning efficiency and precision, overcoming limitations of existing strategies.MethodsOur approach evaluates prevalent IMPT spot selection strategies, identifying their limitations, especially concerning MU constraints. We integrated LB constraints into a MILP framework, using a novel three‐phase strategy for spot pool selection, to enhance performance over traditional heuristic methods and L1 + L∞ strategies. The method's efficacy was tested in eight study cases, using Dose‐Volume Histograms (DVHs), spot selection efficiency, and computation time analysis for benchmarking against established methods.ResultsThe proposed method showed superior performance in DVH quality, adhering to LB constraints while maintaining high‐quality treatment plans. It outperformed existing techniques in spot selection, reducing unnecessary spots and balancing precision with efficiency. Cases studies confirmed the method's effectiveness in producing clinically feasible plans with enhanced dose distributions and reduced hotspots, especially in cases with elevated LB constraints.ConclusionsOur Two‐Stage MILP strategy signifies a significant advancement in IMPT treatment planning. By incorporating LB constraints directly into the optimization process, it achieves superior plan quality and deliverability compared to current methods. This approach is particularly advantageous in clinical settings requiring minimum spot number and high MU LB constraints, offering the potential for improved patient outcomes through more precise and efficient radiation therapy plans.

Abstract Objective . UNet-based deep-learning (DL) architectures are promising dose engines for traditional linear accelerator (Linac) models. Current UNet-based engines, however, were designed differently with various strategies, making it challenging to fairly compare the results from different studies. The objective of this study is to thoroughly evaluate the performance of UNet-based models on magnetic-resonance (MR)-Linac-based intensity-modulated radiation therapy (IMRT) dose calculations. Approach . The UNet-based models, including the standard-UNet, cascaded-UNet, dense-dilated-UNet, residual-UNet, HD-UNet, and attention-aware-UNet, were implemented. The model input is patient CT and IMRT field dose in water, and the output is patient dose calculated by DL model. The reference dose was calculated by the Monaco Monte Carlo module. Twenty training and ten test cases of prostate patients were included. The accuracy of the DL-calculated doses was measured using gamma analysis, and the calculation efficiency was evaluated by inference time. Results . All the studied models effectively corrected low-accuracy doses in water to high-accuracy patient doses in a magnetic field. The gamma passing rates between reference and DL-calculated doses were over 86% (1%/1 mm), 98% (2%/2 mm), and 99% (3%/3 mm) for all the models. The inference times ranged from 0.03 (graphics processing unit) to 7.5 (central processing unit) seconds. Each model demonstrated different strengths in calculation accuracy and efficiency; Res-UNet achieved the highest accuracy, HD-UNet offered high accuracy with the fewest parameters but the longest inference, dense-dilated-UNet was consistently accurate regardless of model levels, standard-UNet had the shortest inference but relatively lower accuracy, and the others showed average performance. Therefore, the best-performing model would depend on the specific clinical needs and available computational resources. Significance . The feasibility of using common UNet-based models for MR-Linac-based dose calculations has been explored in this study. By using the same model input type, patient training data, and computing environment, a fair assessment of the models’ performance was present.