Ended on 28th Jun'21 03:30 PM (Coordinated Universal Time)
Data Sprint #38: The Kvasir-Capsule Dataset
61
Medium
Challenge Starts
18 Jun 03:30 pm
Registration Ends
28 Jun 03:30 pm
Challenge Ends
28 Jun 03:30 pm
Content
The small bowel constitutes the gastrointestinal (GI) tract’s mid-part, situated between the stomach and the large bowel. It is three to four meters long and has a surface of about 30 m^2, including the surface of the villi, and plays a crucial role in absorbing nutrients. Therefore, disorders in the small bowel may cause severe growth retardation in children and nutrient deficiencies in children and adults. This organ may be affected by chronic diseases, like Crohn’s disease, coeliac disease, and angiectasis,or malignant diseases like lymphoma and adenocarcinoma. These diseases may represent a substantial health challenge for both patients and society, and a thorough examination of the lumen is frequently necessary to diagnose and treat them. However, the small bowel, due to its anatomical location, is less accessible for inspection by flexible endoscopes commonly used for the upper GI tract and the large bowel. Since early 2000, video capsule endoscopy (VCE) has been used, usually as a complementary test for patients with GI bleeding. A VCE consists of a small capsule containing a wide-angle camera, lightsources, batteries, and other electronics. The patient swallows the capsule, which then captures a video as it moves passively through the GI tract. A recorder, carried by the patient or included in the capsule, stores the video before a medical expert assesses it after the procedure. [Source of information: Reasearch paper: Kvasir-Capsule, a video capsule endoscopy dataset]
Problem Statement
Artificial intelligence (AI) is predicted to have profound effects on the future of video capsule endoscopy (VCE) technology.The potential lies in improving anomaly detection while reducing manual labour. However, medical data is often sparse and unavailable to the research community, and qualified medical personnel rarely have time for the tedious labelling work.
Objective
You are required to build a machine learning model to recognize the disease label of the respective images.
What will you learn?
- Practical applications of Deep Learning Algorithms, optimizing neural networks, CNN, etc. (Learn ithere)