Article Preview
TopIntroduction
Artificial Intelligence (AI) plays a crucial role in the domain of medical science such as diagnosis of symptoms and processing of radiological images. AI applies the process of learning that is machine learning, iteratively for improving in response of training procedure, additionally refining outcomes for decision making and predictions. Learning process includes various types (Han, Kamber, & Pei, 2011) (Farooqui & Mehra, 2019). Supervised learning is one of the types where predefined data is trained for applying further predictions on new data (Velchamy, Subramanian, & Vasudevan, 2012). Afterwards, classification technique is executed on afresh entered data for isolating them into groups with the help of trained model and is used to generate decisions (Veni & Rani, 2015).
On the other hand, the techniques of unsupervised learning can be applied to generate decisions without having any trained models (Mishra & Soni, 2014) . The process of learning includes various techniques which are used for analysing, computation and implementation of data (Koteeswaran, Visu, & Janet, 2012) (Li, Roy, Khan, Wang, & Bai, 2012). Usually, learning is considered as the procedure for training the data set. Like learning, Deep learning can be considered as a combination of techniques based on unsupervised, supervised learning and additionally having networks capability such as artificial neural network. It is an automated computer science technology applied to identify patterns and trends in dissimilar groups, for example, identifying fractures in images of medical field. Deep learning tools plays a crucial role in the domain of image processing, computer visualization and medical domain (Dimililer, 2017).
Concept of artificial intelligence i.e., machine intelligence is reflected in deep learning which is comprised of many layers of networks to extract features to improve the level of perception and precision (CHUNG, et al., Automated detection and classið cation of the proximal humerus fracture by using deep learning algorithm, 2018). Artificial intelligence requires deep learning for computations and processing. Deep learning is composed of neural networks, which is further comprised of input layers and hidden layers required for processing and calculations (Johari & Singh, 2018). Therefore, deep neural network has now become an empowered technology for medical image diagnosis for enhanced performance and accuracy in diagnosis.
A fracture is a broken bone, which may be a thin crack or major bone battered badly, and still, it further may be a closed or open type (Al-Ayyoub, Hmeidi, & Rababah, 2013). The severity of fracture may vary which depends on external force (Maheshwari & Mhaskar, 2011). Severity also generally depends on the age of person and on the part of Bone. A fracture is a broken bone which is dealt in Orthopaedics branch of medical science. Fracture can be vertical, horizontal, oblique or may be in smaller pieces in bone after contact with some external force or stimulus. Clinical diagnosis of fracture in lack of specialized expertise may steer to consequences, i.e., Failure to diagnose fracture may otherwise lead to grave consequences and poor outcome.
Usually, all type of fractures can be visible and diagnosed in X-Ray. Later, CT and MRI can be done for confirmation of fracture. X-rays is an imaging technique based on electromagnetic radiation to create picture of internal body parts and bones underneath skin. For acquiring the image of body part in X-ray, the patient is positioned accordingly. Computer Tomography is generally known as CT scan, utilizes computer and X-rays machines for creating cross-sectional images and three-dimensional reconstruction images for more accuracy and understanding. MRI also called as magnetic resonance imaging utilizes the paired and unpaired magnetic fields of proton nucleus of hydrogen atoms of water (which is present everywhere in the body) to create images. MRI is especially useful in detecting fractures which have not yet become precipitated or become imminent e.g., Stress Fractures, or where X-rays and CT scan are contraindicated e.g., Pregnancy.