Decoding Medical Diagnosis with Machine Learning Classifiers

International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 2019

Nowadays, every field is digitizing their data for easy access at anytime and anywhere or even for enclosed cabinet servers, especially the health care sector. But, that is not the only reason health care sector is computerizing its data. These huge chucks of records are used for research purposes. Many hospitals are working with education institutes with research departments (Damian Borbolla et.al 2010).CDSS performs Knowledge-based analyses on these EHRs and running disease prediction models on these data is done. There may be many complications. We have reviewed the problems faced by such system from previous researches and implemented systems.

Studies in health technology and informatics, 2008

Well-designed medical decision support system (DSS) have been shown to improve health care quality. However, before they can be used in real clinical situations, these systems must be extensively tested, to ensure that they conform to the clinical guidelines (CG) on which they are based. Existing methods cannot be used for the systematic testing of all possible test cases. We describe here a new exhaustive dynamic verification method. In this method, the DSS is considered to be a black box, and the Quinlan C4.5 algorithm is used to build a decision tree from an exhaustive set of DSS input vectors and outputs. This method was successfully used for the testing of a medical DSS relating to chronic diseases: the ASTI critiquing module for type 2 diabetes.

Health and Technology, 2013

The rise in living standards that has occurred with the advancement of new technologies has increased the demand for sophisticated standards-based health-care applications that provide services anytime, anywhere, and with low cost. To achieve this objective, we have designed and developed the Smart Clinical Decision Support System (Smart CDSS) that takes input from diverse modalities, such as sensors, user profile information, social media, clinical knowledge bases, and medical experts to generate standards-based personalized recommendations. Smartphone-based, accelerometer-based, environment-based activity-recognition algorithms are developed with this system that recognizes users' daily life activities. For example, social media data are captured for a diabetic patient from his/her social interactions on Twitter, e-mail, and Trajectory and then combined with clinical observations from real encounters in health-care facilities. The input is converted into standard interface following HL7 vMR standards and submitted to the Smart CDSS for it to generate recommendations. We tested the system for 100 patients from Saint Mary's Hospital: 20 with type-1 diabetes, 40 with type-2 diabetes mellitus, and 40 with suspicions for diabetes but no diagnosis during clinical observations. The system knowledge base was initialized with standard guidelines from online resources for diabetes, represented in HL7 Arden syntax. The system generates recommendations based on physicians' guidelines provided at the hospital during patient follow-ups. With support from the Azure cloud infrastructure, the system executed the set of guidelines represented in Arden syntax in a reasonable amount of time. Scheduling and executing the 3-5 guidelines called medical logic modules (MLMs) required less than a second.

Neural Computing and Applications, 2012

Decision support systems help physicians and also play an important role in medical decision-making. They are based on different models, and the best of them are providing an explanation together with an accurate, reliable and quick response. This paper presents a decision support tool for the detection of breast cancer based on three types of decision tree classifiers. They are single decision tree (SDT), boosted decision tree (BDT) and decision tree forest (DTF). Decision tree classification provides a rapid and effective method of categorizing data sets. Decision-making is performed in two stages: training the classifiers with features from Wisconsin breast cancer data set, and then testing. The performance of the proposed structure is evaluated in terms of accuracy, sensitivity, specificity, confusion matrix and receiver operating characteristic (ROC) curves. The results showed that the overall accuracies of SDT and BDT in the training phase achieved 97.07 % with 429 correct classifications and 98.83 % with 437 correct classifications, respectively. BDT performed better than SDT for all performance indices than SDT. Value of ROC and Matthews correlation coefficient (MCC) for BDT in the training phase achieved 0.99971 and 0.9746, respectively, which was superior to SDT classifier. During validation phase, DTF achieved 97.51 %, which was superior to SDT (95.75 %) and BDT (97.07 %) classifiers. Value of ROC and MCC for DTF achieved 0.99382 and 0.9462, respectively. BDT showed the best performance in terms of sensitivity, and SDT was the best only considering speed. Keywords Computer-aided diagnosis (CAD) Á Decision support systems (DSS) Á Decision tree classification Á Single decision tree Á Boosted decision tree Á Decision tree forest Á k-fold cross-validation

International Journal of Computing and Digital Systems, 2021

Millions of folks around the earth are affliction from late disease identification and diagnosis. An incredible amount of health information has been obtained by the latest technologies in digital medical services and information communication technologies. Disease diagnosis and artificially intelligent decision support systems have drawn tremendous attention from many scientists and research community worldwide. Various algorithms developed and applied with the aid of machine learning techniques that can substantially lead to the resolution of the system of health care and can help personnel involved in the early diagnosis of diseases. This research paper will propose an artificial intelligent algorithm which helps us to effectively, rapidly and accurately classify the information. The Proposed Disease Diagnosis Support Systems (DDSS) can assist clinicians to monitor the information, facilitate their evaluation by means of a preparatory treatment and decrease evaluation time per patient. The patients may inevitably be notified and recommended dietary suggestions also. This system will allow clinicians to focus on attending patients in accordance with their homeostasis. It decreases the volume of work of doctors and enables them to define patients who need to be examined more urgently or meticulously. Even with the widespread growth of such systems security of digital data and its privacy is still a major challenge yet to solve.

2020

Recent trends in the incorporation of computational science into medical and biological field of research have led to the accumulation of huge amounts of data regarding medical and experimental information. The application of data mining in healthcare sectors enables early prediction of patient conditions and their behaviors by performing data analysis and discovering relations from seemingly unrelated large volume of collected data. There is also increasing popularity of data mining in healthcare operations due to its ability to benefit all parties. For instance, data mining application in this sector aids in ensuring that patients receive more affordable and better healthcare services, physicians identify the best practices and effective treatments, healthcare firms make informed decisions about customer relationship management, and healthcare insurers detect abuse and fraud. Despite these promising trends, however, the resultant and huge data amounts that healthcare transactions ...

This chapter provides an overview of the Machine Learning (ML) concepts in the clinical field which data may be collected, either by Health Care Professionals (HCP) or patients. These data may include activities and medication reminders, objective measurement of physiological parameters, feedback based on observed patterns, questionnaires and scores that require computational processes that give rise to useful information capable of supporting clinical decision making. The chapter describes ML in terms of learning concepts emphasizing the following approaches: supervised, unsupervised, semi-supervised, and reinforcement learning. The principles of concept classification are explained and the mathematical concepts of several methodologies are presented, such as neural networks and support vector machine among other techniques. Finally, a case study based on a radial basis function neural network aiming at the estimation of ECG waveform is presented. The proposed method reveals its suitability to support HCP on clinical decisions and practices.

2004

The present study explored dichotomic classification methods for medical diagnosis data through three experiments. A first experiment run in Weka used four different classification schemes on two different sets of medical test data thus permitting comparison of each scheme’s performance. A second experiment tested the application of attribute selection, information gain, and boosting to Weka’s support vector classification scheme (SMO). Finally, in the third experiment when a cost matrix was applied to breast cancer diagnostic data, false negatives were effectively reduced to under one percent while overall accuracy was slightly improved. The first experiment suggests that SMO may classify better than J48, IBk and Naïve Bayes with respect to medical test data from the UCI repository. The data of the first experiment also suggests that support vector classification-based diagnosis outperforms manual diagnosis of fine needle aspirate results. While the second experiment showed no enha...

— Many healthcare providers and clinicians are using various types of systems known as CDSS which assist them to take decision about each case of patient. There is increasing recognition that if CDSS is well planned and put into service, has an immense potential to get better health care quality and possibly even increase effectiveness and decrease healthcare costs. Some existing CDSS are reviewed in this paper. The various types of CDSS available encompass a range of preferences, from general references, through precise detailed guidelines for a given condition. Technological intervention in developing CDSS has to be evaluated while designing, developing and implementing it. The basic CDSS is designed and proposed outline is given. The implementation issues for CDSS have to be considered and for successful execution of CDSS meticulous planning has to be done. Healthcare providers and users of CDSS are required to understand CDSS benefits and limitations, and the inimitable challenges of designing and implementing various types of CDSS to attain the optimal advantage of CDSS.