Rectal cancer
Conditions
Interventions
Group 1: This is a non-interventional observational study.
Retrospective and prospective data from routine clinical practice will be analyzed for patients who have undergone robot-assisted rectal rese
Sponsors
TU Dresden
Eligibility
Sex/Gender
All
Age
18 Years to No maximum
Inclusion criteria
Inclusion criteria: Retrospective: Robot-assisted rectal resections performed at a participating center Prospective: Indication for robot-assisted rectal resection using a surgical robot (e.g., the DaVinci system) Patient understands German
Exclusion criteria
Exclusion criteria: Conversion to open surgery Technical difficulties with data collection or data loss Missing or incomplete surgical videos
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| Intersection-over-Union (IoU) as a measure of the overlap between the actual location of a structure in the image (“ground truth”) and the location predicted by the AI model. Scale from 0 (no overlap) to 1 (perfect detection). | — |
Secondary
| Measure | Time frame |
|---|---|
| Additional metrics of predictive performance: F1 score, precision, recall, specificity, AUROC Qualitative endpoints: Assessment of AI-assisted detection of nerves and dissection planes by two surgeons, with evaluation of interrater agreement Technical performance parameters of the AI models: computation latencies, reliability, image distortions Functional outcomes (questionnaires): quality of life (QLQ-C30), fecal incontinence (LARS), urinary function (IPSS), female sexual function (FSFI), male erectile function (IIEF-5) Oncological endpoints: Local recurrence rate, overall survival, progression-free survival Surgical outcome parameters: Pathological quality of the specimen (CRM status, resection status), surgical complications, surgeons’ satisfaction with TME | — |
Countries
Germany, Switzerland
Outcome results
None listed