A Human Versus Machine Comparison Study
- Conditions
- Head and Neck Cancer
- Registration Number
- NCT06035653
- Lead Sponsor
- East and North Hertfordshire NHS Trust
- Brief Summary
The study aims to retrospectively, and anonymously analyse the data of patients treated over a period of 12 months. These are the patients whose radiotherapy plan were needed to be reviewed in the weekly on-treatment head and neck multi-professional radiotherapy meeting during the course of their radiation.
Collected clinical data will be included in the retrospective analysis, comparing the clinical decision (gold standard of care), against the predictive capability of intelligence software
- Detailed Description
This is a single centre, non-interventional study comparing "machine" against clinician's standard of care approach. The utilisation of intelligence software is for research purposes only and does not permit active changes to patients' clinical management. Any software related analyses will be done after patients have completed radiation treatment.
All consented patients who were discussed in the head and neck CBCT meeting and deemed to require new masks / recontouring / replanning - Cohort A will be analysed on completion of radiotherapy. These patients typically are ones with nodal disease of \>3cm at presentation, and whose weight change is notable 5 - 10% of baseline in the first 2 weeks of their treatment. There have been 28 patients in 12 months previously.
Other patients discussed in the head and neck CBCT meeting on completion of radiation - Cohort B, whose radiotherapy was not deemed to require total replan but were clinically deemed 'marginal' will be included. These patients tend to be ones with weight loss measuring close to 1.29cm lateral separation or between the skin surface and the head shell.
The radiotherapy replan volumes CTVs of Cohort A patients will retrospectively compared to the CTVs produced by the intelligence software. The time using the software to produce dynamic CTVs will be compared with the time needed for the clinicians to create new CTV for the radiotherapy replan, which will be recorded prospectively.
The CBCTs of both Cohorts A \& B will be analysed using the intelligence software to determine the adequacy of PTV coverage. This is to be compared with the departmental quality assurance/audit of PTV margins.
The software will also be used to calculate various aspects of the radiotherapy plans, including the cumulative dose distribution over the period of dynamic anatomical change Cohorts A \& B.
Recruitment & Eligibility
- Status
- RECRUITING
- Sex
- All
- Target Recruitment
- 40
- i. Head and neck cancer patients who require new mask for radiotherapy replanning will be included (Cohort A). These patients typically are ones with rapid shrinkage of the nodal disease, or whose weight change is notable (5 - 10% of baseline) in the first 2 weeks of radiotherapy ii. Patients on close observation of the head and neck multi-professional team but whose radiotherapy did no undergo total replan treatment will be included (Cohort B)
i. Patients whose primary malignancy is not of head and neck origin ii. Patients who specifically opt against their information be used anonymously iii. Head and neck cancer patients whose radiotherapy plans did not require review in the head and neck multi-professional radiotherapy meeting iv. Patients under the age of 18
Study & Design
- Study Type
- OBSERVATIONAL
- Study Design
- Not specified
- Primary Outcome Measures
Name Time Method Interobserver variability studies (IOV) between manually contoured CTVs and software-generated deformed CTVs 1 year Measured by volume differences
Interobserver variability studies (IOV) between manually contoured CTVs and software-generated deformed CTVs - Geographical miss index 1 year Measurement of under contouring
Interobserver variability studies (IOV) between manually contoured CTVs and software-generated deformed CTVs - Hausdorff distance 1 year Measurement of distance between 2 volumes
Interobserver variability studies (IOV) between manually contoured CTVs and software-generated deformed CTVs - Sorensen-DICE Coefficient 1 year Measurement of ratio of overlap
Interobserver variability studies (IOV) between manually contoured CTVs and software-generated deformed CTVs - Discordance Index 1 year Measurement of over contouring
PTV margins compared against the departmental standards 1 year Proportion of clinical tumour volume covered by different radiotherapy set up margins 3, 4 and 5mm
- Secondary Outcome Measures
Name Time Method Potential time saved recontouring with software and the associated cost of re-planning adaptive radiotherapy 1 year Difference between time spent on re- contouring and time spent on editing software deformed CTVs
Evaluation of clinical decision on total radiotherapy re-plan 1 year Measured by retrospective assessment of dose accumulation of the CTVs
Trial Locations
- Locations (1)
East & North Hertfordshire NHS Trust
🇬🇧Northwood, UK, United Kingdom