Purpose
The dosimetric accuracies of volumetric modulated arc therapy (VMAT) plans were predicted using plan complexity parameters via machine learning.
Methods
The dataset consisted of 600 cases of clinical VMAT plans from a single institution. The predictor variables (n = 28) for each plan included complexity parameters, machine type, and photon beam energy. Dosimetric measurements were performed using a helical diode array (ArcCHECK), and the dosimetric accuracy of the passing rates for a 5% dose difference (DD5%) and gamma index of 3%/3 mm (γ3%/3 mm) were predicted using three machine learning models: regression tree analysis (RTA), multiple regression analysis (MRA), and neural networks (NNs). First, the prediction models were applied to 500 cases of the VMAT plans. Then, the dosimetric accuracy was predicted using each model for the remaining 100 cases (evaluation dataset). The error between the predicted and measured passing rates was evaluated.
Results
For the 600 cases, the mean ± standard deviation of the measured passing rates was 92.3% ± 9.1% and 96.8% ± 3.1% for DD5% and γ3%/3 mm, respectively. For the evaluation dataset, the mean ± standard deviation of the prediction errors for DD5% and γ3%/3 mm was 0.5% ± 3.0% and 0.6% ± 2.4% for RTA, 0.0% ± 2.9% and 0.5% ± 2.4% for MRA, and –0.2% ± 2.7% and –0.2% ± 2.1% for NN, respectively.
Conclusions
NNs performed slightly better than RTA and MRA in terms of prediction error. These findings may contribute to increasing the efficiency of patient‐specific quality‐assurance procedures.
We previously isolated three monomeric dihydrodiol dehydrogenases, DD1, DD2 and DD4, from human liver, and cloned a cDNA (C9) thought to encode DD2, which is identical with those for human bile-acid-binding protein and an oxidoreductase of human colon carcinoma HT29 cells. In the present study we have provided evidence that the C9 cDNA clone encodes DD1, not DD2. A recombinant enzyme expressed from the cDNA in a bacterial system was purified, and its catalytic properties, bile-acid-binding ability and primary sequence were compared with those of the hepatic dihydrodiol dehydrogenases. The results show that DD1 encoded by C9 possesses prostaglandin F synthase activity but low affinity for lithocholic acid, whereas DD2, showing differences of six amino acid residues from the DD1 sequence, exhibited high-affinity binding for the bile acid. Refined relationship between dihydrodiol dehydrogenases and their related proteins of human tissues is proposed.
The aim of this paper is to assess the overall geometric accuracy of the Novalis system using the Robotic Tilt Module in terms of the uncertainty in frameless stereotactic radiotherapy. We analyzed the following three metrics: (1) the correction accuracy of the robotic couch, (2) the uncertainty of the isocenter position with gantry and couch rotation, and (3) the shift in position between the isocenter and central point detected with the ExacTrac x-ray system. Based on the concept of uncertainty, the overall accuracy was calculated from these values. The accuracy in positional correction with the robotic couch was 0.07 +/- 0.22 mm, the positional shift of the isocenter associated with gantry rotation was 0.35 mm, the positional shift of the isocenter associated with couch rotation was 0.38 mm and the difference in position between the isocenter and the ExacTrac x-ray system was 0.30 mm. The accuracy of intracranial stereotactic radiosurgery with the Novalis system in our clinic was 0.31 +/- 0.77 mm. The overall geometric accuracy based on the concept of uncertainty was 0.31 +/- 0.77 mm, which is within the tolerance given in the American Association of Physicists in Medicine report no. 54.
Human liver contains two dihydrodiol dehydrogenases, DD2 and DD4, associated with 3 alpha-hydroxysteroid dehydrogenase activity. We have raised polyclonal antibodies that cross-reacted with the two enzymes and isolated two 1.2 kb cDNA clones (C9 and C11) for the two enzymes from a human liver cDNA library using the antibodies. The clones of C9 and C11 contained coding sequences corresponding to 306 and 321 amino acid residues respectively, but lacked 5'-coding regions around the initiation codon. Sequence analyses of several peptides obtained by enzymic and chemical cleavages of the two purified enzymes verified that the C9 and C11 clones encoded DD2 and DD4 respectively, and further indicated that the sequence of DD2 had at least additional 16 residues upward from the N-terminal sequence deduced from the cDNA. There was 82% amino acid sequence identity between the two enzymes, indicating that the enzymes are genetic isoenzymes. A computer-based comparison of the cDNAs of the isoenzymes with the DNA sequence database revealed that the nucleotide and amino acid sequences of DD2 and DD4 are virtually identical with those of human bile-acid binder and human chlordecone reductase cDNAs respectively.
Positional tracking errors correlated strongly with 4D-modeling errors in IR tracking. Thus, the accuracy of the 4D model must be verified before treatment, and margins are required to compensate for the 4D-modeling error.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.