.png){kind=link}
Health is a crucial facet of everybody's life. We all have to take medicine prescribed by doctors from time to time to sustain the soundness of our bodies. It is coherently laborious to maintain track of all the medications throughout life. Dissimilarity exists among individuals and medical institutions for treating patients with diverse medicines. Specific drugs might treat a patient positively while they might not turn out to be that effective on others. Hence, understanding each patient's body and the side effects of the drugs on them is crucial for doctors. The work of the doctors can be eased if they can refer to the patient's medical records. On the contrary, it is a tedious and meticulous job for the patients to maintain their own drug history. The proposed model Medibox will become a bridge for the doctors and the patient's smooth communication. The users can easily upload their prescriptions that will be converted into a digital repository along with more features. Overall, Medibox is a one-spot solution for patients to start pursuing a wholesome life and to keep a record of their medical history effortlessly.
Medibox also provides convenience by offering features like finding nearby hospitals or doctors. Moreover, the application focuses on the well-being of the user by estimating their body mass index concerning their present-day height and weight. According to their BMI, Medibox will also suggest relative videos to the user for a better understanding of their situation and to develop a path to healthy living. For smooth navigation through the prescriptions and medications, Medibox also assists in presenting a chatbot for the users. A chatbot is a program that is designed to imitate intelligent dialogue over text or speech. These systems can self-learn and restore their knowledge with the help of humans or web resources. Because knowledge is saved in advance, this application is extremely important. To respond to user inquiries, the system application engages the question and answer protocol in the form of a chatbot. Apart from this, user can even schedule their medication timings flexibly at any date and time. Medibox also provides a calendar for patients to remind about their medications periodically. Hence, Medibox is a collection of modules that will induce the users to a better, fitter, and easily calculated healthy lifestyle.
The dataset utilized is available at drugs.com, drug names are used as annotations and then passed to reviews as our training data. It consists of about 3700 drugs names. Other than drug names, it also has reviews that are used to train the Spacy model.
Aim of this project is to provide a platform to the elderly people regarding their health concerns. It has group of modules that would ease the friction between their consultation journey. The platform has 5 modules all together. The main focus is on information retrieval from the prescriptions with the help of OCR. OCR is commonly used in python for image-to-text use cases as it can read and recognize text from the images. The information extracted from the prescription will be stored in the database. This data will not only help user to store their medical records at one place but also help user maintain their medication. The application also has a health chat-bot which will assist users regarding general health issues. Apart from this it will also suggest exercise videos for fitness of the users.
.png)
With the proliferation of technology in the medical field, the importance of keeping electronic medical records is upsurging. Although as easy as it seems, it is very difficult and challenging to store, maintain, and update the records efficiently. Therefore, this project focuses on implementing a practical solution to detect multiple details from the printed prescription such as the doctor's name, number, hospital address, drugs, and prescription date. All these predictions are performed by using the concepts of Machine learning, Deep Learning, Optical Character Recognition, and Natural Language Processing accordingly. In addition to keeping customers' medical information, Medibox will also provide them advice on how to live a healthy life based on those records.
This line of work can be further improved by making it more scalable for multifarious prescription formats. For now, Medibox uses a predefined format of prescription for prediction. For a more efficient prognosis, we can also increase the training dataset and the number of preprocessing steps. Chatbot responses can also be more accurate and detailed by adding more layers to Neural Network and also providing more question-answer patterns in its JSON file. Hence, Medibox is a one-stop application that tracks the patient's present and past medical records encouraging its users to lead healthier and more physically fit lives.