– Can a deep learning tool improve diagnostics in patients with bladder cancer?
Short title: BC-DEEDS
Period: Spring 2022 – Spring 2025
Aim: The study is a collaboration between the Department and the Danish Medtech Company ‘ Cystotech ApS ‘.
‘Cystotech ApS’ is currently developing ‘ CystoAID® ‘ which is software to be applied in a clinical setting for the detection of malignant tissue in the bladder wall, as well as a decision helping tool for the surgeon during cystoscopies and TURBTs.
The aim of the study is to facilitate the collection, storage, and anonymization of images and vides obtained during elective Cystoscopies and TURBTs before it is transferred to Cystotech.
Number of patients: Currently 935 patients have consented to have their procedure recorded.
Method: Before routine cystoscopies patients are asked to consent to having videos and pictures taken during their routine procedure. If the patient is willing to participate, staff will record the procedure and take still pictures of suspicious areas in the bladder before storing the data. When available, the patients’ pathology results from the procedure is entered into a study database and linked to the images. After the data has been irrevocably anonymized it is transferred to Cystotech who uses it to educate the algorithm CystoAid is based on, making CystoAid and deep learning tool.
Post marketing, Urological surgeons will not have to rely solely on their experience when detecting malignant tissue and making decisions as to how much tissue should be surgically removed, but have the option to utilize software which is educated on thousands and thousands of images and pathology, for clinical decision making when performing cystoscopies and TURBT.
Sites: The collection of images and videos during Cystoscopies is currently taking place at the Urological Clinic at Viborg Hospital, at the Regionshospitalet in Horsens and Randers as well as the Urological Clinic at Aarhus University Hospital.
