Bone Cancer Detection using Neural Network

Project Title

Bone Cancer Detection using Neural Network

Project Area of Specialization Artificial IntelligenceProject Summary

Bone cancer detection is a desktop application which is use for detecting cancer tumor from MRI scan report of cancer patient. User can upload or enter MRI scan image to our system and our system analyze that image and tell the user that whether the tumor exist or not. Our system provide  functionality to user for entering MRI scan image in our system, user can enter image from local disk storage of PC or from external drives like USB. After receiving input from user our system perform different image processing techniques on input image for tumor detection. When tumor is diagnosed, we provide a functionality to user for getting brief description about his disease, our system tell the user about the size of the tumor and also tell the user about the stage of the disease. Our system also highlighted the tumor area for giving best response to user about the presence of tumor, by highlighting the tumor area a layman who don’t have enough knowledge about the cancer disease also come to know about the present situation of disease. Our system save the User’s time, because they come to know about the condition of his disease by using this app instead of taking appointment from doctor. If the tumor is identified in MRI report after using this app, then patient could go to doctor for proper treatment.User enter his MRI scan image to our system and our system will predict that whether the tumor is exist or not. 

Project Objectives

The main objective of our system is to save user time.Cancer diagnosis is a very time taken process, doctors advise patient to perform different lab tests, scans and Biopsy etc. After performing all these tests then patient go to Cancer specilist and then this cancer specilist check and analyze this report and then tell the patient about his disease. Our system provide functionality to cancer patient to know about the status of his disease by simply uploading his MRI scan image to our system.

Our system also facilate to Cancer specilist to save his time also, doctor just enetr each MRI scan image of cnacer patient and then tell about their disease rather the to analayze each report individually.

Project Implementation Method

The diagnosing of cancer is mostly done through MRI scan. MRI uses a strong magnetic field and radio waves to create detailed images of the organs and tissues within the body. It provides us a detailed image but it cannot find the tumor where it locates exactly, if a normal person or Radiologist reads the MRI report then they cannot come to know about the tumor. Our system is capable to provide information about the tumor to normal person or a Radiologist. Our system takes the MRI images as input and tells the user where the tumor is located. 

Benefits of the Project

As we mentioned above that the diagnosing process of cancer is a time taken process in which firstly patient go for his MRI scan of his body part where the possibility of tumor is exist. After MRI scan the patient take an appointment from cancer specialist doctor (Orthopedic Oncologist) and the MRI report is examine by doctor. After examining the report the doctor tells to patient about the disease.

So in order to address this type of problem and to save patient time we provide a system in which there is no need to take an appointment from doctor and our system directly tells to patient about the description of disease. Our system saves patient’s time and patient money.

Our system also provide facility to cancer specialist (Orthopedic Oncologist), because they have numbers of reports to be examine. By entering MRI report in our system the tumor area will be  highlighted. So it also saves the doctor’s time.

Technical Details of Final Deliverable

Our system make prediction about the presence of cancer tumor using trained model. We use different Neural networks and classifier like Convolutional Neural Network CNN and Artificial neural Network and SVM in order to trained our system. We use a large number of dataset of cancer and non cancer images and extract some usefull feature from this dataset and provide these feature to neural networks during training process.

Final Deliverable of the Project Software SystemType of Industry Medical Technologies Artificial Intelligence(AI)Sustainable Development Goals Good Health and Well-Being for PeopleRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	60000
dataset of MRI scan images	Equipment	10000	6	60000