Deep learning describes a class of machine learning algorithms that are capable of combining raw inputs into layers of intermediate features. These algorithms have recently shown impressive results across a variety of domains. Biology and medicine are data-rich disciplines, but the data are complex and often ill-understood. Hence, deep learning techniques may be particularly well-suited to solve problems of these fields. We examine applications of deep learning to a variety of biomedical problems—patient classification, fundamental biological processes, and treatment of patients. The product helps in diagnosis of infectious diseases, it asks a series of questions designed to emulate the thinking of an expert in the field of infectious diseases, and from the responses to these questions give a list of possible diagnoses, with probability, as well as recommend treatment. The product also intends to adjust the patient’s dose and push reminders on when to take the pills and monitor it. It also uses the location data to detect common trends in an area and hence suggest the necessary precautions to be taken care of. Its portability makes it a highly useful and practical invention.
Keywords: Deep Learning, Neural Nets, Artificial Intelligence, Data Analysis and Visualizaiton.
Biology and medicine are rapidly becoming data-intensive. The term deep learning has come to refer to a collection of new techniques that, together, have demonstrated breakthrough gains over existing best-in-class machine learning algorithms across several fields. There is a great potential to create deep learning systems that augment biologists and clinicians by prioritizing experiments or streamlining tasks that do not require expert judgment. We divide the tasks into three broad classes: Disease and Patient Categorization, Fundamental Biological Study, and Treatment of Patients. A key challenge in biomedicine is the accurate classification of diseases and disease subtypes. In healthcare, individuals are diagnosed with a disease or condition based on symptoms, the results of certain diagnostic tests, or other factors. Once diagnosed with a disease, an individual might be assigned a stage based on another set of human-defined rules. Imagine a deep neural network is provided with clinical test results gleaned from electronic health records. Because physicians may order certain tests based on their suspected diagnosis, a deep neural network may learn to “diagnose” patients simply based on the tests that are ordered. The product asks a series of questions from the patients just like a physician and on the basis of the answers provided recommends certain tests that need to be undertaken. The patient submits the report of the test and on the basis of the report, outlines the diseases and then recommends essential treatment. The product also recommends necessary precautions to be undertaken on the basis of your location and push gentle reminders when to take pills and monitor it. We are optimistic about the future of deep learning in biology and medicine, its full potential has not been explore and there is vast scope of features that can be integrated with this product making it more worthwhile in the future.
2. PROBLEM DOMAIN
Problems faced by the patient’s are listed as follows:
The complete diagnosis is a cumbersome process where one needs to wait in long queue for getting the treatment and it more or less results in severe cases.
Any sort of medication is not provided for free, and usually even a small treatment is expensive.
Doctor to patient ratio. The sheer number of patients a doctor has to attend is overwhelming and leads to lack of personal attention to each patient. A visit to a govt hospital reveals how a doctor has to take important calls on life within a span of few minutes while attending large number of patients.
Doctors Negligence/error of Judgement. The error of judgement of doctors or their negligence often goes unnoticed and the patients pay for the mistakes of the doctor. Sometimes the mistakes can prove fatal to the patient.
Money poses big problem for many individuals, due to which they tend to ignore their diseases avoid the doctor thus posing them severe health problems.
Ignorance in taking pills has been the major issue cutting individual’s healthy life.
3. SOLUTION DOMAIN
The proposed solution will aim to improve the diagnosis system and provide the patients and affordable and homely accurate doctor.
The important features of the system are-
It is User-friendly and free of cost.
The patients often need to wait in the long queues for very long time in the hospital. The
proposed system is portable and can diagnose them anywhere in the world.
The need for rushing to a doctor for medication or small treatment is thus saved, indeed even saving the money.
The high accuracy of the system is equivalent to getting the treatment from the best doctor in the world.
The individual’s who avoid consulting doctor because of lack of money can get the treatment, hence preventing them from severe health problems
The system also pushes reminder for pills and suggests precautions to be undertaken on the basis of your location.
The basic idea for this project is to guide all the patients to a proper affordable diagnosis procedure, giving right treatment and suggestions for the precautions of certain diseases. The goal is not to replace a doctor , while machine learning might help with “suggestions” in a diagnostic situation, a doctor’s judgment would still be needed in order to factor for the specific context of the patient.
4. SYSTEM DOMAIN
The entire backend is built in Django,Python supported by Flask. The front end of the website is designed using client-side languages like HTML, CSS and Java Script.
The System uses a Convolutions neural network and deep learning, machine learning algorithms
so as to classify the information and train the model to get the expected outcome.
Hardware Requirements: -
• Processor: Intel core i3 and above
• System type: 64 bit Operating System
• RAM: atleast 2GB
• Hard Disk: 5GB (atleast excluding Data size)
Software Requirements: -
• Operating System: Linux, Windows10 or above.
• Database Server: MySql Server.
• Designing Tool: HTML, CSS, JS
• Development Tool: Jupyter, Spyder, Google Static API’s, Tensorflow
5. APPLICATION DOMAIN
The application will prove beneficial for every individual. It not only saves the time and money but also gives you the best possible diagnosis.
The application also uses your location feature and detects common trends or diseases in an area and thereby suggests necessary precautions.
The application allows is basically a personalized doctor who gives you complete time and suggests you the best treatment possible.
Seeing huge potential in biology and medicine, lot of giants have taken step towards exploring its domain, and with Amazon Alexa and Google Home Building up there’s huge scope of its future integration thus making it really valuable. The proliferation of consumer wearables and other medical devices combined with AI is also being applied to oversee early-stage heart disease, enabling doctors and other caregivers to better monitor and detect potentially life-threatening episodes at earlier, more treatable stages.
Money and Time are most impactful terms in a society and an application which helps save both of them has really a huge scope thus making it useful and practical invention.
6. EXPECTED OUTCOME
The system shall take the information from patients like Name, Age, Gender, Previous Medical Issues and current symptoms.
On the basis of above series of questions it uses a trained model to generate tests that patients should undergo.
Once done with the tests the patient needs to submit the report and on the basis of report it describes the disease and suggest necessary treatment.
For eg. If user enters symptoms like Polydipsia, Polyuria, Polyhagia,it suggests and AIC test(1.1) and diagnoses it as a diabetic patient with medicines TZP, Colesvlern.