A Literature Survey on Data Mining Approach to Effectively Handle Cancer Treatment

Abstract

The effective treatment of cancer is not very easy since diagnosis of cancer involves many stages of treatment with gradually changing lifestyles. Physicians play vital role in identifying the correct cause and feel ambiguity for making perfect decisions about hundreds of data available from the internet resource. IDA (Intelligent Data Analysis) which is a part from Data Mining techniques is quiet useful to most of the physicians for decision making about types of cancers. IDA facilitates physicians to classify, detect and analyze the cancer outcome to patients. Healthcare Management System also aids the practitioners to practically search, analyze and compare the result analysis of the patient with existing data in the HMS and guide proper treatment to the cancer affected patient. Health care data analysis comprises enormous data with diversity of health information. One among the most important points that pull down the practitioner's confidence is that utility of latest software and most sophisticated computing machines. This put them in to the state of confusion for proper and elegant decision making for treating the cancer affected patients. Problems in user interaction, lack of awareness in data mining, improper knowledge in electronic guidelines makes physicians to work with old methods of treatment. Traditional medical practicing and modern methods of computing do not match either because of ignorance. IDA and HMS have significant impact for cancer treatment with speedy diagnosis and faster recovery. This also shows great impact on costs, clinical outcomes and proper guidelines for clinical approach. The prime motto of this survey article is to analyze the survey application, bring out the importance of comparison strategies of IDA to improve decision making for medical practitioner for effective cancer treatment.

Diagnosis and Grading of Diabetic Retinopathy using Deep Learning

“…Patients lose time between having their eyes scanned, having their images examined by medical professionals, and setting up a follow-up visit. [7][8][9][10][11][12][13][14][15] By using automated systems, the proposed method can solve the above mentioned problems. Previous attempts have used image classification, pattern recognition, and machine learning to make quick progress in developing a thorough and automated approach of DR screening.…”

Segmentation and Classification of Dermatoscopic Skin Lesion images using U-Net and MobileNet models

“…The most prevalent form of cancer in humans, skin cancer is typically identified visually after a first clinical screening and maybe after dermoscopic analysis, a biopsy, and a histological study [7]. The fine-grained variety in how skin lesions form makes it difficult to automatically classify skin lesions using photographs.…”

“…Three different models, MobileNet model, VGG-16 model, and a custom model is trained and tested on two different datasets. On HAM-10000 dataset achieved the accuracy of 82% using the MobileNet architecture.The proposed methodologies in above mentioned papers helped us to gain knowledge about the present-day technologies used in Skin LesionClassification techniques [7][8][9][10][12][13][14][15]…”

Crop and Fertilizer Recommendation System

“…Technological Changes in the agricultural sector tend to be very slow. It is often observed By reviewing the considerations and methodologies used in the research papers [10][11][12][13][14][15][16][17][18], they have tried to develop ML models to recommend crops and fertilizers to produce a good yield. Some considered building a web application to ease the process of accessing the recommendation system.…”