放射治疗技术作为一种可靠而有效的治疗方式，在肿瘤治疗领域的应用不断提高，甚至可以取代某些肿瘤的手术治疗。随着精准放疗时代的到来，单次大剂量照射技术产生的生物学优势，以及质子布拉格峰产生的生物学优势在临床上受到越来越多的关注。研究单次大剂量光子和质子辐射后细胞存活模型对精准预测评估放射治疗剂量具有重要的临床意义。 论文基于 DNA 亚致死损伤机制原理，结合细胞敏感度分析，采用两种新型模型-LQR 模型与 LQI 模型进行光子单次大剂量照射剂量预测验证。对人肺腺癌细胞(A549) 与前列腺癌细胞 (DU145) 经 6 MV 光子辐射后产生的细胞存活曲线值进行研究，基于 LQR 模型、 LQI 模型、LQ 模型、 USC 模型、 LQL 模型以及 GLQ 模型等进行剂量预测。结果表明 LQR 模型与 LQI 模型的 A549 与 DU145 拟合曲线预测值与实验值最符合，拟合度分别为?𝑄?𝐴549： ?2/𝑑?𝑓=0.1308 、 ?𝑄?𝐴549： ?2/𝑑?𝑓=0.1035、?𝑄?𝐷?145： 𝑥2/?𝑜?=0.0030、?𝑄?𝐷?145： 𝑥2/?𝑜?=0.0035。 LQR 与 LQI 模型在光子单次大剂量照射技术下可准确预测剂量。 分析 A549、 DU145、中国仓鼠卵巢细胞 (CHO) 以及人胶质母细胞瘤 (T98) 经不同能量不同深度质子辐照后产生的细胞存活曲线值，基于 LQ 模型拟合发现， LQ模型在布拉格曲线上其他位置均能准确预测细胞存活值，但布拉格峰位置处产生的辐射剂量预测是不准确的， LQ 模型高估了质子对细胞杀伤能力，这将对临床工作带来不可靠因素。基于 LQR 模型和 LQI 模型拟合发现， LQR 模型更能准确预测质子剂量， LQI 模型在布拉格峰处的拟合能力与 LQ 模型表现相当。同时，布拉格曲线上不同位置的传能线密度 (LET) 不同，对细胞产生不同的杀伤。 D3(布拉格峰)、 D4(拖尾区) 位置处产生生物效应更高，且 D4 处辐射杀伤能力高于 D3 位置。质子束的生物效应在能量为 71 MeV 时显著高于其能量为 160 MeV 时的效应。 传统 LQ 的生物等效剂量（BED）经证明无法准确转换为不同分次剂量和治疗次数的等效剂量。 LQR 模型在光子与质子辐照细胞拟合能力表现最优，基于 LQR模型建立 BED，为临床评估放射治疗剂量提供了参考依据。

Radiation therapy continues to advance in the field of oncology as a reliable and effective treatment modality. sometimes even replacing surgical interventions for certain tumors. With the advent of precision radiotherapy, the biological advantages offered by single high-dose fractionation techniques and the biological benefits of the proton Bragg peak have garnered increasing attention clinically. Investigating the cell survival models after single high-dose photon and proton irradiation is of significant clinical importance for the precise prediction and assessment of radiation therapy doses. This paper employed the principles of DNA sublethal damage mechanisms and integrates cell sensitivity analysis to validate photon single high-dose irradiation dose predictions using two novel models: the LQR model and the LQI model. The study focused on the cell survival curve values of human lung adenocarcinoma (A549) and prostate cancer cells (DU145) following 6 MV photon irradiation, comparing dose predictions based onthe LQR, LQI, LQ, USC, LQL, and GLQ models. The findings indicated that the LQR and LQI models provided the best fit to the experimental data for both A549 and DU145,with goodness of fit indices of?𝑄?𝐴549:?2/𝑑?𝑓=0.1308 、 ?𝑄?𝐴549： ?2/𝑑?𝑓=0.1035、?𝑄?𝐷?145： 𝑥2/?𝑜?=0.0030、 ?𝑄?𝐷?145： 𝑥2/?𝑜?=0.0035.The LQR and LQI modelsaccurately predicted doses under single high-dose photon irradiation techniques. Furthermore, analyzing the cell survival curve values for A549, DU145, Chinese hamster ovary cells (CHO), and human glioblastoma (T98) post-irradiation with protons of varying energies and depths revealed that while the LQ model can accurately predict cell survival values at various positions on the Bragg curve, it inaccurately predicted the radiation dose at the Bragg peak, overestimating the proton’s cell-killing potential,which introduces unreliability in clinical practice. Fitting with the Linear-QuadraticLinear (LQL) and Linear-Quadratic-Interpolation (LQI) models showed that the LQLmodel could more accurately predict proton doses, while the LQI model’s fitting ability at the Bragg peak was comparable to that of the LQ model. Additionally, the Linear Energy Transfer (LET) varies at different positions on the Bragg curve, affecting the cellkilling differently. Biological effects were higher at positions D3 (Bragg peak) and D4 (tail region), with the radiation’s cell-killing ability being greater at D4 than at D3. The biological effects of the proton beam were significantly higher at an energy of 71 MeV compared to 160 MeV. The traditional biological equivalent dose (BED) based on the conventional LQ model had been demonstrated to be inadequate for converting into equivalent doses for different fractionation schedules and treatment regimens. The LQR model exhibits optimal fitting capability for cell survival curves under photon and proton irradiation, and establishing BED based on the LQR model provided a reference basis for clinical assessment of radiation therapy doses.