Radiomics from Various Tumour Volume Sizes for Prognosis Prediction of Head and Neck Squamous Cell Carcinoma: A Voted Ensemble Machine Learning Approach.

Background: Traditionally, cancer prognosis was determined by tumours size, lymph node spread and presence of metastasis (TNM staging). Radiomics of tumour volume has recently been used for prognosis prediction. In the present study, we evaluated the effect of various sizes of tumour volume. A voted ensemble approach with a combination of multiple machine learning algorithms is proposed for prognosis prediction for head and neck squamous cell carcinoma (HNSCC). Methods: A total of 215 HNSCC CT image sets with radiotherapy structure sets were acquired from The Cancer Imaging Archive (TCIA). Six tumour volumes, including gross tumour volume (GTV), diminished GTV, extended GTV, planning target volume (PTV), diminished PTV and extended PTV were delineated. The extracted radiomics features were analysed by decision tree, random forest, extreme boost, support vector machine and generalized linear algorithms. A voted ensemble machine learning (VEML) model that optimizes the above algorithms was used. The receiver operating characteristic area under the curve (ROC-AUC) were used to compare the performance of machine learning methods, including accuracy, sensitivity and specificity. Results: The VEML model demonstrated good prognosis prediction ability for all sizes of tumour volumes with reference to GTV and PTV with high accuracy of up to 88.3%, sensitivity of up to 79.9% and specificity of up to 96.6%. There was no significant difference between the various target volumes for the prognostic prediction of HNSCC patients (chi-square test, p > 0.05). Conclusions: Our study demonstrates that the proposed VEML model can accurately predict the prognosis of HNSCC patients using radiomics features from various tumour volumes.

Non-coplanar VMAT plans for postoperative primary brain tumour to reduce dose to hippocampus, temporal lobe and cochlea: a planning study.

OBJECTIVES: This study aimed to compare radiotherapy plan quality of coplanar volumetric modulated arc therapy (CO-VMAT) and non-coplanar VMAT (NC-VMAT) for post-operative primary brain tumour. METHODS: A total of 16 patients who were treated for primary brain tumours were retrospectively selected for this study. For each patient, identical CT sets with structures were used for both CO-VMAT and NC-VMAT planning. For CO-VMAT, one full arc and two coplanar half arcs were used. For NC-VMAT, one full coplanar and two non-coplanar half arcs with couch rotation of 315° or 45° were used. Dose constraints were adhered to the RTOG 0614, RTOG 0933 and TMH protocol. Dose volumetric parameters were collected for statistical analysis. RESULTS: .NC-VMAT achieved significant dose reduction in contralateral hippocampus, both temporal lobes and cochleae, and other OARs while the plan qualities remained the same. In particular, NC-VMAT decreased contralateral hippocampus mean dose by 1.67Gy. Similarly, the NC-VMAT decreased temporal lobe mean dose by 6.29Gy and 2.8Gy for ipsilateral and contralateral side respectively. Furthermore, it decreased cochlea mean dose by 5.34Gy and 0.97Gy for ipsilateral and contralateral side respectively. Overall, there was a reduction of 5.4% of normal brain tissue volume receiving low dose irradiation. CONCLUSION: The proposed NC-VMAT showed more favourable plan quality than the CO-VMAT for primary brain tumours, in particular to hippocampus, temporal lobes, cochleae and OARs located to the contralateral side of tumours. ADVANCES IN KNOWLEDGE: For primary brain tumours radiotherapy, NC-VMAT can reduce doses to the hippocampus, both temporal lobes, and cochleae, as well as OARs located to the contralateral side of tumours.

The Role of Virtual Environment for Radiotherapy Training (VERT) in Medical Dosimetry Education.

Medical dosimetry is an important component in training of radiation therapist, yet it is not easy for student to understand the principle of treatment planning and to be familiar with the relationship of the clinical target volume (CTV), planned target volume (PTV), and the nearby organs at risk (OARs) by just imagination. This study is conducted to evaluate whether using VERT in teaching medical dosimetry can help student to improve their learning experience. Students of cohort 2015 and 2016 were taught under TPS mode and TPS + VERT mode respectively. Direct comparison was conducted through self-evaluation survey, between two groups of students, in terms of their understanding of the concept of medical dosimetry and their level of confidence in completing different types of plans after the course. Both groups of students were able to understand the concept of medical dosimetry and able to complete 3D conformal plans after taking the course. Though, the students received TPS mode reported that they had lower level of confidence in completing the planning and required longer time for self-study and practice compared to the students who received the TPS + VERT mode. This study demonstrated that including VERT into medical dosimetry education can improve students' learning experience, by improving their self-confidence, as well as reducing time required for their self-study and practice.