Renal Cyst Complex
Conditions
Keywords
Radiomics, Bosniak classification, Renal cell carcinoma, Medical imaging
Brief summary
More than 200,000 new cases of renal cancer are diagnosed in the world each year, with more than 63,000 new cases in Europe alone. Of those, renal cell carcinoma (RCC) is the most common type in adults, making up more than 90% of the cases. Deciding on the benign or malignant nature of some RCC on the basis of medical images (CT, MRI, US) is an issue, which often leads to unnecessary surgery, morbidity and costs. A categorization for renal cysts was introduced in the late 1980s known as the Bosniak classification. The Bosniak classification system classifies them into groups that are benign (I and II) and those that need surgical resection (III and IV), based on specific imaging features. However, defining the malignancy of category III lesions still remains a challenge. Though Bosniak classification for renal cysts is used worldwide and underwent a number of modifications, Bosniak III cysts still have almost a 1:1 chance of being malignant. So the problem is that approximately half of the Bosniak category III cystic lesions prove to be benign after surgery. The proposed project aims to develop a quantitative image analysis (QIA) based multifactorial decision support system (mDSS) capable of classifying renal cysts with high accuracy into benign or malignant status, thus reducing the amount of unnecessary surgeries performed. Using standard-of-care CT images and clinical parameters, the customized DSS will then guide experts in planning a safe and effective diagnostic and treatment strategy for all RCC patients.
Interventions
The high-throughput extraction of large amounts of quantitative image features from radiographic medical images
Sponsors
Study design
Eligibility
Inclusion criteria
* Patients who underwent contrast-enhanced CT-Scan, with radiological findings suggestive of Bosniak 3 renal cyst and have available results for pathology analysis of the cyst.
Exclusion criteria
* CT-Scans that are reformatted or secondary.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| malignancy classifier | 1 year | Machine learning algorithm that can differentiate between malignant and beingn bosniak 3 renal cysts. |
Countries
Netherlands