necessity of data mining in clinical emergency medicine; a narrative review of the current literatrue

Clinical databases can be categorized as big data, include large quantities of information about patients and their medical conditions. Analyzing the quantitative and qualitative clinical data in addition with discovering relationships among huge number of samples using data mining techniques could unveil hidden medical knowledge in terms of correlation and association of apparently independent variables. The aim of this research is using predictive algorithm for prediction of trauma patients on admission to hospital to be able to predict the necessary treatment for patients and provided the necessary measures for the trauma patients who are before entering the critical situation. This study provides a review on data mining in clinical medicine. The relevant, recently-published studies of data mining on medical data with a focus on emergency medicine were investigated to tackle pros and cons of such approaches. The results of this study can be used in prediction of trauma patient's status at six hours after admission to hospital

Predicting implantation outcome of in vitro fertilization and intracytoplasmic sperm injection using data mining techniques

Background: In vitro fertilization [IVF] and intracytoplasmic sperm injection [ICSI] are two important subsets of the assisted reproductive techniques, used for the treatment of infertility. Predicting implantation outcome of IVF/ICSI or the chance of pregnancy is essential for infertile couples, since these treatments are complex and expensive with a low probability of conception

Materials and Methods: In this cross-sectional study, the data of 486 patients were collected using census method. The IVF/ICSI dataset contains 29 variables along with an identifier for each patient that is either negative or positive. Mean accuracy and mean area under the receiver operating characteristic [ROC] curve are calculated for the classifiers. Sensitivity, specificity, positive and negative predictive values, and likelihood ratios of classifiers are employed as indicators of performance. The state-of-art classifiers which are candidates for this study include support vector machines, recursive partitioning [RPART], random forest [RF], adaptive boosting, and one-nearest neighbor

Results: RF and RPART outperform the other comparable methods. The results revealed the areas under the ROC curve [AUC] as 84.23 and 82.05%, respectively

The importance of IVF/ICSI features was extracted from the output of RPART. Our findings demonstrate that the probability of pregnancy is low for women aged above 38

Conclusion: Classifiers RF and RPART are better at predicting IVF/ICSI cases compared to other decision makers that were tested in our study. Elicited decision rules of RPART determine useful predictive features of IVF/ICSI. Out of 20 factors, the age of woman, number of developed embryos, and serum estradiol level on the day of human chorionic gonadotropin administration are the three best features for such prediction

Sleep apnea headaches

Obstructive sleep apnea syndrome [OSAS] is a common disorder characterized by recurrent apnea during sleep. Nocturnal laboratory-based polysomnography [PSG] is the gold standard test for diagnosis of OSA. The sufferers may complain from daytime sleepiness, snoring or occasional headaches. Serious consequences such as cardiovascular complications, stroke or symptoms of depression may complicate the syndrome. Headache prevalence due to sleep apnea is estimated 1%-2% in general population and affects 2%-8% of middle age population. Morning headache is more common in the OSAS patients. OSAS patients present with various characteristics of morning headache. Treatment with continuous positive airway pressure usually reduces headache. The pathophysiologic background for a relation between obstructive sleep apnea and morning headache is multifactorial. Some theories have been proposed for OSAS-related headaches such as changing oxygen saturation during sleep, cerebral vasodilation and increased intracranial pressure due to cerebral vasodilation, sleep disruption and depression but the definite cause of headaches in OSAS patients is not yet clear

Magnetic resonance imaging for Human T-cell lymphotropic virus type 1 [HTLV1-] associated myelopathy/tropical spastic paraparesis patients: a systematic review

Introduction: Human T-cell lymphotropic virus type 1 [HTLV-1]-associated myelopathy/tropical spastic paraparesis is a chronic progressive neurologic disease, which might be associated with brain and spinal cord atrophy and lesions. Here, we systematically reviewed the brain and spinal cord abnormalities reported by magnetic resonance imaging [MRI] modality on HTLV-1-associated myelopathy/tropical spastic paraparesis patients

Methods: PubMed was searched for all the relevant articles, which used MRI in patients with HTLV-1-associated myelopathy/tropical spastic paraparesis. Included criteria were all the cohort and case series with at least 10 patients. We had no time limitation for searched articles, but only English language articles were included in our systematic review. Exclusion criteria were none-English articles, case reports, articles with less than 10 patients, spastic paraparesis patients with unknown etiology and patients with HTLV-II

Results: Total of 14 relevant articles were extracted after studying title, abstracts and full text of the irrelevant articles. Only 2/14 articles reported brain atrophy incidence. Five out of 14 articles studied the brain lesions prevalence. Spinal cord atrophy and lesions were studied in 6/14 articles

Discussion: According to the extracted data, brain atrophy does not seem to happen frequently in patients with HTLV-1 associated myelopathy/tropical spastic paraparesis. None-specific brain lesions identified in articles are indicative of low specificity of MRI technique despite its high sensitivity

Conclusion: Prevalence of spinal cord lesions and atrophy in these patients might be due to the degenerative processes associated with aging phenomenon. Further and larger studies in endemic areas could more accurately reveal the specificity of MRI in these patients

Vasculitic peripheral neuropathy

Primary systemic vasculitis in pre-capillary arteries is associated with peripheral neuropathy. In some types of systematic vasculitis about 60% of patients have peripheral nervous system [PNS] involvement. In vasculitic peripheral neuropathies [VPN] a necrotizing and inflammatory process leads to narrowing of vasa nervorum lumen and eventually the appearance of ischemic lesions in peripheral nerves. Some features might be suggestive of VPN, like: axonal nerve degeneration, wallerian-like degeneration, and diameter irregularity of nerve. Peripheral nervous system [PNS] destruction during systemic vasculitides should be considered, due to its frequency and early occurrence in vasculitis progression. The first line treatment of non systematic VPNs is corticosteroid agents, but these drugs might worsen the VPNs or systemic vasculitis

[Evaluation of the frequency of polyneuropathy based upon the clinical and paraclinical findings in diabetic patients]

Peripheral neuropathy has been known as the main cause of diabetic foot ulcer and limb amputation. Early diagnosis of this complication can prevent more severe morbidity as well as enormous economic costs. Based on the duration of the disease, 110 diabetic patients were divided in two groups of less than 10 and more than 10-year history. After taking complete history about the symptoms of neuropathy, the patients were examined neurologically. Then, electrodiagnostic studies were performed on the patients. 96 patients [87/2%] had diabetic neuropathy. The prevalence of neuropathy had a significant relationship with the duration of the disease [P value= 0.004]. 61 patients [55/4%] complained of neuropathic symptoms. The results of the neurological exams were abnormal in 65 patients [59.09%]. The most common symptom was paresthesia [50% of patients] and the most frequent sign was abnormal Achilles' reflex. 77 patients [70%] had abnormal indices in nerve conduction studies the most common of which was decrease amplitude of peroneal and sural nerves. 16 patients [14/5%] showed alterations in NCS in the absence of clinical signs or symptoms and 19 patients [17/2%] had normal NCS indices despite the presence of signs or symptoms of neuropathy. High sensitivity of taking history and a careful neurological examination in the diagnosis of diabetic neuropathy necessitates physicians to pay more attention to clinical examinations and patient complaints, and avoid costly electrodiagnostic investigations. Thus, early diagnosis of sub-clinical neuropathy in these patients and providing them with necessary recommendations, irreversible complications such as amputation can be prevented

Combination of beamforming and synchronization methods for epileptic source localization, using simulated EEG signals

Localization of sources in patients with focal seizure has recently attracted many attentions. In the severe cases of focal seizure, there is a possibility of doing neurosurgery operation to remove the defected tissue. The prosperity of this heavy operation completely depends on the accuracy of source localization. To increase this accuracy, this paper presents a new weighted beamforming method to precisely localize the focal seizure sources from the electroencephalogram [EEG] signals. First, synchronization value is determined just between each two adjacent channels, and the channel with maximum average in synchronization index is selected as the nearest channel to the dominant focal sources. Next, weight of each channel is determined based on its Euclidean's distance to the selected channel. The determined weights act as a prior knowledge, incorporating in multiple signal classification [MUSIC] and some other beamforming methods to localize the exact place of seizure source. Next, the effect of estimated source is removed from the signals repeatedly to find the second focal source. This process continues till all focal sources being determined. To verify and validate the proposed scheme, 65 channels EEG signals were simulated and a linear weighting was applied to the three groups based on some sources. The proposed scheme and some known beamforming methods such as conventional beamformer, MUSIC, Weighted MUSIC, Capon's, Eigenvector and also SLORETA were applied to the simulated epileptic signals to find the location of sources. Experimental results reveal the superiority of the proposed method to the rival schemes in terms of localization accuracy, both in clean and noisy environments