No abstract

Purpose To develop a deep learning method for prediction of three‐dimensional (3D) voxel‐by‐voxel dose distributions of helical tomotherapy (HT). Methods Using previously treated HT plans as training data, a deep learning model named U‐ResNet‐D was trained to predict a 3D dose distribution. First, the contoured structures and dose volumes were converted from plan database to 3D matrix with a program based on a developed visualization toolkit (VTK), then transferred to U‐ResNet‐D for correlating anatomical features and dose distributions at voxel‐level. One hundred and ninety nasopharyngeal cancer (NPC) patients treated by HT with multiple planning target volumes (PTVs) in different prescription patterns were studied. The model was typically trained from scratch with weights randomly initialized rather than using transfer‐learning method, and used to predict new patient's 3D dose distributions. The predictive accuracy was evaluated with three methods: (a) The dose difference at the position r, δ(r, r) = Dc(r) − Dp(r), was calculated for each voxel. The mean (μδ(r,r)) and standard deviation (σδ(r,r)) of δ(r, r) were calculated to assess the prediction bias and precision; (b) The mean absolute differences of dosimetric indexes (DIs) including maximum and mean dose, homogeneity index, conformity index, and dose spillage for PTVs and organ at risks (OARs) were calculated and statistically analyzed with the paired‐samples t test; (c) Dice similarity coefficients (DSC) between predicted and clinical isodose volumes were calculated. Results The U‐ResNet‐D model predicted 3D dose distribution accurately. For twenty tested patients, the prediction bias ranged from −2.0% to 2.3% and prediction error varied from 1.5% to 4.5% (relative to prescription) for 3D dose differences. The mean absolute dose differences for PTVs and OARs are within 2.0% and 4.2%, and nearly all the DIs for PTVs and OARs had no significant differences. The averaged DSC ranged from 0.95 to 1 for different isodose volumes. Conclusions The study developed a new deep learning method for 3D voxel‐by‐voxel dose prediction, and shown to be able to produce accurately dose predictions for nasopharyngeal patients treated by HT. The predicted 3D dose map can be useful for improving radiotherapy planning design, ensuring plan quality and consistency, making clinical technique comparison, and guiding automatic treatment planning.

Background Limonene is an important biologically active natural product widely used in the food, cosmetic, nutraceutical and pharmaceutical industries. However, the low abundance of limonene in plants renders their isolation from plant sources non-economically viable. Therefore, engineering microbes into microbial factories for producing limonene is fast becoming an attractive alternative approach that can overcome the aforementioned bottleneck to meet the needs of industries and make limonene production more sustainable and environmentally friendly. Results In this proof-of-principle study, the oleaginous yeast Yarrowia lipolytica was successfully engineered to produce both d-limonene and l-limonene by introducing the heterologous d-limonene synthase from Citrus limon and l-limonene synthase from Mentha spicata, respectively. However, only 0.124 mg/L d-limonene and 0.126 mg/L l-limonene were produced. To improve the limonene production by the engineered yeast Y. lipolytica strain, ten genes involved in the mevalonate-dependent isoprenoid pathway were overexpressed individually to investigate their effects on limonene titer. Hydroxymethylglutaryl-CoA reductase (HMGR) was found to be the key rate-limiting enzyme in the mevalonate (MVA) pathway for the improving limonene synthesis in Y. lipolytica. Through the overexpression of HMGR gene, the titers of d-limonene and l-limonene were increased to 0.256 mg/L and 0.316 mg/L, respectively. Subsequently, the fermentation conditions were optimized to maximize limonene production by the engineered Y. lipolytica strains from glucose, and the final titers of d-limonene and l-limonene were improved to 2.369 mg/L and 2.471 mg/L, respectively. Furthermore, fed-batch fermentation of the engineered strains Po1g KdHR and Po1g KlHR was used to enhance limonene production in shake flasks and the titers achieved for d-limonene and l-limonene were 11.705 mg/L (0.443 mg/g) and 11.088 mg/L (0.385 mg/g), respectively. Finally, the potential of using waste cooking oil as a carbon source for limonene biosynthesis from the engineered Y. lipolytica strains was investigated. We showed that d-limonene and l-limonene were successfully produced at the respective titers of 2.514 mg/L and 2.723 mg/L under the optimal cultivation condition, where 70% of waste cooking oil was added as the carbon source, representing a 20-fold increase in limonene titer compared to that before strain and fermentation optimization. Conclusions This study represents the first report on the development of a new and efficient process to convert waste cooking oil into d-limonene and l-limonene by exploiting metabolically engineered Y. lipolytica strains for fermentation. The results obtained in this study lay the foundation for more future applications of Y. lipolytica in converting waste cooking oil into various industrially valuable products.