Machine learning for dose-volume histogram based clinical decision-making support system in radiation therapy plans for brain tumors

Abstract Background Referral and clinical decision-making support are important for reducing delays in reaching and receiving appropriate and quality care. This paper presents analysis of the use of a pilot referral and decision making support call center for mothers and newborns in the Greater Accra region of Ghana, and challenges encountered in implementing such an intervention. Methods We analyzed longitudinal time series data from routine records of the call center over the first 33 months of its operation in Excel. Results During the first seventeen months of operation, the Information Communication Technology (ICT) platform was provided by the private telecommunication network MTN. The focus of the referral system was on maternal and newborn care. In this first phase, a total of 372 calls were handled by the center. 93% of the calls were requests for referral assistance (87% obstetric and 6% neonatal). The most frequent clinical reasons for maternal referral were prolonged labor (25%), hypertensive diseases in pregnancy (17%) and post-partum hemorrhage (7%). Birth asphyxia (58%) was the most common reason for neonatal referral. Inadequate bed space in referral facilities resulted in only 81% of referrals securing beds. The national ambulance service was able to handle only 61% of the requests for assistance with transportation because of its resource challenges. Resources could only be mobilized for the recurrent cost of running the center for 12 h (8.00 pm – 8.00 am) daily. During the second phase of the intervention we switched the use of the ICT platform to a free government platform operated by the National Security. In the next sixteen-month period when the focus was expanded to include all clinical cases, 390 calls were received with 51% being for medical emergency referrals and 30% for obstetrics and gynaecology emergencies. Request for bed space was honoured in 69% of cases. Conclusions The call center is a potentially useful and viable M-Health intervention to support referral and clinical decision making in the LMIC context of this study. However, health systems challenges such inadequacy of human resources, unavailability of referral beds, poor health infrastructure, lack of recurrent financing and emergency transportation need to be addressed for optimal functioning.

Download Full-text

Real-Time AI-Based Informational Decision-Making Support System Utilizing Dynamic Text Sources

Applied Sciences ◽

10.3390/app11136237 ◽

2021 ◽

Vol 11 (13) ◽

pp. 6237

Author(s):

Azharul Islam ◽

KyungHi Chang

Keyword(s):

Machine Learning ◽

Decision Making ◽

Random Forest ◽

Support System ◽

Classification Accuracy ◽

Short Term Memory ◽

Learning Algorithm ◽

Unstructured Data ◽

Stochastic Gradient Descent ◽

Decision Making Support

Unstructured data from the internet constitute large sources of information, which need to be formatted in a user-friendly way. This research develops a model that classifies unstructured data from data mining into labeled data, and builds an informational and decision-making support system (DMSS). We often have assortments of information collected by mining data from various sources, where the key challenge is to extract valuable information. We observe substantial classification accuracy enhancement for our datasets with both machine learning and deep learning algorithms. The highest classification accuracy (99% in training, 96% in testing) was achieved from a Covid corpus which is processed by using a long short-term memory (LSTM). Furthermore, we conducted tests on large datasets relevant to the Disaster corpus, with an LSTM classification accuracy of 98%. In addition, random forest (RF), a machine learning algorithm, provides a reasonable 84% accuracy. This research’s main objective is to increase the application’s robustness by integrating intelligence into the developed DMSS, which provides insight into the user’s intent, despite dealing with a noisy dataset. Our designed model selects the random forest and stochastic gradient descent (SGD) algorithms’ F1 score, where the RF method outperforms by improving accuracy by 2% (to 83% from 81%) compared with a conventional method.

Download Full-text

Neuroendoscopy Adapter Module Development for Better Brain Tumor Image Visualization

International Journal of Electrical and Computer Engineering (IJECE) ◽

10.11591/ijece.v7i6.pp3643-3654 ◽

2017 ◽

Vol 7 (6) ◽

pp. 3643

Author(s):

Sunil L. Bangare ◽

G. Pradeepini ◽

Shrishailappa Tatyasaheb Patil

Keyword(s):

Decision Making ◽

Brain Tumors ◽

Clinical Decision Making ◽

Clinical Decision ◽

Image Resolution ◽

Minimally Invasive Surgical Procedure ◽

Narrow Band Image ◽

Minimally Invasive Surgical ◽

Auto Fluorescence ◽

Adapter Module

The issue of brain magnetic resonance image exploration together with classification receives a significant awareness in recent years. Indeed, various computer-aided-diagnosis solutions were suggested to support radiologist in decision-making. In this circumstance, adequate image classification is extremely required as it is the most common critical brain tumors which often develop from subdural hematoma cells, which might be common type in adults. In healthcare milieu, brain MRIs are intended for identification of tumor. In this regard, various computerized diagnosis systems were suggested to help medical professionals in clinical decision-making. As per recent problems, Neuroendoscopy is the gold standard intended for discovering brain tumors; nevertheless, typical Neuroendoscopy can certainly overlook ripped growths. Neuroendoscopy is a minimally-invasive surgical procedure in which the neurosurgeon removes the tumor through small holes in the skull or through the mouth or nose. Neuroendoscopy enables neurosurgeons to access areas of the brain that cannot be reached with traditional surgery to remove the tumor without cutting or harming other parts of the skull. We focused on finding out whether or not visual images of tumor ripped lesions ended up being much better by auto fluorescence image resolution as well as narrow-band image resolution graphic evaluation jointly with the latest neuroendoscopy technique. Also, within the last several years, pathology labs began to proceed in the direction of an entirely digital workflow, using the electronic slides currently being the key element of this technique. Besides lots of benefits regarding storage as well as exploring capabilities with the image information, among the benefits of electronic slides is that they can help the application of image analysis approaches which seek to develop quantitative attributes to assist pathologists in their work. However, systems also have some difficulties in execution and handling. Hence, such conventional method needs automation. We developed and employed to look for the targeted importance along with uncovering the best-focused graphic position by way of aliasing search method incorporated with new Neuroendoscopy Adapter Module (NAM) technique.

Download Full-text

Machine learning to assist clinical decision-making during the COVID-19 pandemic

Bioelectronic Medicine ◽

10.1186/s42234-020-00050-8 ◽

2020 ◽

Vol 6 (1) ◽

Cited By ~ 5

Author(s):

Shubham Debnath ◽

◽

Douglas P. Barnaby ◽

Kevin Coppa ◽

Alexander Makhnevich ◽

...

Keyword(s):

Machine Learning ◽

Decision Making ◽

Clinical Decision Making ◽

Clinical Decision

Download Full-text

An Empirical Analysis of Machine Learning Classifiers for Clinical Decision Making in Asthma

Communications in Computer and Information Science - Cognitive Computing and Information Processing ◽

10.1007/978-981-10-9059-2_11 ◽

2018 ◽

pp. 105-117

Author(s):

M. R. Pooja ◽

M. P. Pushpalatha

Keyword(s):

Machine Learning ◽

Decision Making ◽

Empirical Analysis ◽

Clinical Decision Making ◽

Clinical Decision ◽

Machine Learning Classifiers ◽

Learning Classifiers

Download Full-text

579. Machine-Learning Based Models for Prediction of Recurrence-free Catheter Retention After ALT Treatment of CLABSI in a Pediatric Population

Open Forum Infectious Diseases ◽

10.1093/ofid/ofz360.648 ◽

2019 ◽

Vol 6 (Supplement_2) ◽

pp. S273-S274

Author(s):

Lorne W Walker ◽

Andrew J Nowalk ◽

Shyam Visweswaran

Keyword(s):

Machine Learning ◽

Decision Making ◽

Blood Culture ◽

Clinical Decision Making ◽

Pediatric Population ◽

Recurrent Infection ◽

Clinical Decision ◽

Definitive Treatment ◽

Support Vector ◽

The Future

Abstract Background Deciding whether to attempt salvage of an infected central venous catheter (CVC) can be challenging. While line removal is the definitive treatment for central-line associated bloodstream infection (CLABSI), salvage may be attempted with systemic antibiotics and antibiotic lock therapy (ALT). Weighing risk and benefit of CVC salvage is limited by uncertainty in the future viability of salvaged CVCs. If a CVC is likely to require subsequent removal (e.g., due to recurrent infection) salvage may not be beneficial, whereas discarding a viable CVC is also not desirable. Here we describe a machine learning approach to predicting outcomes in CVC salvage. Methods Episodes of pediatric CLABSI cleared with ALT were identified by retrospective record review between January 1, 2008 and December 31, 2018 and were defined by a single positive central blood culture of a known pathogen or two matching cultures of a possible contaminant. Clearance was defined as 48-hours of negative cultures and relapse was defined as a matching positive blood culture after clearance. Predictive models [logistic regression (LR), random forest (RF), support vector machine (SVM) and an ensemble combining the three] were used to predict recurrence-free CVC retention (RFCR) at various time points using a training and test set approach. Results Overall, 712 instances CLABSI cleared with ALT were identified. Demographic and microbiological data are summarized in Tables 1 and 2. Few (8%) instances recurred in the first 28 days. 58% recurred at any time within the study period. Rates of RFCR were 75%, 43%, 22% and 10% at 28, 91, 182 and 365 days. Machine learning (ML) models varied in their ability to predict RFCR (Table 3). RF models performed best overall, although no model performed well at 91 days. Conclusion ML models provide an opportunity to augment clinical decision making by learning patterns from data. In this case, estimating the likelihood of useful line retention in the future could help guide informed decisions on salvage vs. removal of infected CVCs. Limitations include the heterogeneity of clinical data and the use of an outcome capturing both clinical decision making (line removal) and infection recurrence. With further model development and prospective validation, practical machine learning models may prove useful to clinicians. Disclosures All authors: No reported disclosures.

Download Full-text