Posts

Showing posts from May, 2022

Report on various kind of bio-inks development process for bioprinting Applications

Image
Abstract Bio-ink is a relatively recent field of study. As the area of 3D printing matures, it throws up plenty of new opportunities for artificial tissue development. However, because we're working with a substance whose end component will be implanted within the human body, finding a decent ink is usually tough.  In this review, we provide an in-depth discussion on various types of bio-ink available, compare their properties, check their use cases, and their development procedure for bioprinting applications. Key Words:-  bio-ink, bioprinting, tissue-development, polymer, 3D printer 1. Introduction The increase in demand for organ and tissue regeneration has been always challenging. It is always a matter of concern whether the donated organ or tissue will be biocompatible with the patient or not. 3D bioprinting has emerged to be a good solution to tackle this issue. Like a normal 3D printer it also requires ink, the composition of the ink should be suitable for the environment on

High benchmark for deep-learning researcher, is it hurting the research community? Are the Deep-Learning research metric tweaked?

Image
This all started with the minor project that we chose for my pre-final semester. We were working on a Pneumonia binary image classification problem. Despite the fact that this was a simple project. As suggested by my mentor, we looked for previous work in this domain. We found a good enough research article to begin working on the review section.  Image by Gerd Altmann from Pixabay It is normally a very excellent habit to conduct some research on the domain on which you are working; it prevents plagiarism and proposes new ideas from the future scope and drawbacks discussed in any article. So, while reviewing the papers, we discovered that many of them claimed to be more accurate than 98 percent, and some were even more accurate than 99.5 percent. One issue that lingered in my mind was, if there are numerous AI solutions with accuracy below 95 percent, why isn't anyone utilising such models with accuracy above 99.5 percent? Is this model theoretically sound, or is it actually feasi