RLAS-BIABC: A Reinforcement Learning-Based Answer Selection Using the BERT Model Boosted by an Improved ABC Algorithm.

Answer selection (AS) is a critical subtask of the open-domain question answering (QA) problem. The present paper proposes a method called RLAS-BIABC for AS, which is established on attention mechanism-based long short-term memory (LSTM) and the bidirectional encoder representations from transformers (BERT) word embedding, enriched by an improved artificial bee colony (ABC) algorithm for pretraining and a reinforcement learning-based algorithm for training backpropagation (BP) algorithm. BERT can be comprised in downstream work and fine-tuned as a united task-specific architecture, and the pretrained BERT model can grab different linguistic effects. Existing algorithms typically train the AS model with positive-negative pairs for a two-class classifier. A positive pair contains a question and a genuine answer, while a negative one includes a question and a fake answer. The output should be one for positive and zero for negative pairs. Typically, negative pairs are more than positive, leading to an imbalanced classification that drastically reduces system performance. To deal with it, we define classification as a sequential decision-making process in which the agent takes a sample at each step and classifies it. For each classification operation, the agent receives a reward, in which the prize of the majority class is less than the reward of the minority class. Ultimately, the agent finds the optimal value for the policy weights. We initialize the policy weights with the improved ABC algorithm. The initial value technique can prevent problems such as getting stuck in the local optimum. Although ABC serves well in most tasks, there is still a weakness in the ABC algorithm that disregards the fitness of related pairs of individuals in discovering a neighboring food source position. Therefore, this paper also proposes a mutual learning technique that modifies the produced candidate food source with the higher fitness between two individuals selected by a mutual learning factor. We tested our model on three datasets, LegalQA, TrecQA, and WikiQA, and the results show that RLAS-BIABC can be recognized as a state-of-the-art method.

Spinal injury resulting from car accident: Focus to prevention.

AIM: To determine and compare the patterns of spinal injury in car occupants. MATERIALS AND METHODS: Retrospective cross-sectional study enrolling all patients with spinal fracture after car accidents, who were admitted to hospital more than 24 h during 2004-2009. RESULTS: The lumbosacral spine was the most commonly involved region (64.8). Six patients had spinal cord injury (6.6%). The majority of the victims were drivers of the vehicle (86.8%) and remaining were passengers. There was a significant difference in lumbar anatomic region (P = 0.05) and place of accident (P = 0.05) in car occupants' position (P = 0.05). Car rollover was the most common mechanism of spinal fractures. There was a significant difference in lumbar anatomic region (P = 0.05), and two or more associated organ injuries (P ≤ 0.05) in car accident mechanism (P = 0.05). CONCLUSION: The chance of sustaining serious spine and associated multiple injuries in car accidents is quite high in our today's society. This may be due to the low level of standards for car manufacturing, absence or inadequacy of appropriate safety measures in cars, and poorly designed roads and problems in quality of driving to mention some reasons. Therefore, these victims are prone to significant morbidity and even mortality and need more specific prehospital supportive interventions.

Repetitive Traumatic Brain Injury in Patients From Kashan, Iran.

BACKGROUND: Traumatic brain injury (TBI) is a worldwide problem, especially in countries with high incidence of road traffic accidents such as Iran. Patients with a single occurrence of TBI have been shown to be at increased risk to sustain future TBI. OBJECTIVES: The aim of this study was to present the incidence and characteristics of repeated TBI (RTBI) in Iranian patients. PATIENTS AND METHODS: During one year, all admitted TBI patients with prior TBI history were enrolled into the study. In each patient, data such as age, gender, past medical history, injury cause, anatomic site of injury, TBI severity, clinical findings and CT scan findings were collected. RESULTS: RTBI comprised 2.5% of TBI cases (41 of 1629). The incidence of RTBI per 100,000 individuals per years was 9.7. The main cause of RTBI was road traffic accident (68.3%); 9.7 % of cases had preexisting seizure/epilepsy disorder; 36.6% of patients with RTBI had pervious ICU admission due to severe TBI. Ten patients had Glasgow coma scale (GCS) ≤ 13 (24.4%). Seizure was seen in seven patients (17.1%). Thirty-nine percent of patients with RTBI had associated injuries. Eleven patients had abnormal CT scan findings (26.9%). CONCLUSIONS: Considering the high incidence of trauma in developing countries, RTBI may also be more common compared with that of developed countries. This mandates a newer approach to preventive strategies, particularly in those with a previous experience of head injury.

Seat Belt Usage in Injured Car Occupants: Injury Patterns, Severity and Outcome After Two Main Car Accident Mechanisms in Kashan, Iran, 2012.

BACKGROUND: Road traffic accidents (RTAs) are the main public health problems in Iran. The seat belts, which are vehicle safety devices, are imperative to reduce the risk of severe injuries and mortality. OBJECTIVES: The aim of the study was to evaluate injury patterns, severity and outcome among belted and unbelted car occupants who were injured in car accidents. PATIENTS AND METHODS: This cross-sectional prospective study was performed on all car occupants injured in RTAs (n = 822) who were transported to hospital and hospitalized for more than 24 hours from March 2012 to March 2013. Demographic profile of the patients, including age, gender, position in the vehicle, the use of seat belts, type of car crashes, injured body regions, revised trauma score (RTS), Glasgow coma score (GCS), duration of hospital stay and mortality rate were analyzed by descriptive analysis, chi-square and independent t-test. P < 0.05 was considered significant. RESULTS: A total of 560 patients used seat belts (68.1%). The unbelted occupants were younger (28 years vs. 38 years) and had more frequently sustained head, abdomen and multiple injuries (P = 0.01, P = 0.01 and P = 0.009, respectively). Also, these patients had significantly lower GCS and elongated hospitalization and higher death rate (P = 0.001, P = 0.001 and P = 0.05, respectively). Tendency of severe head trauma and low RTS and death were increased in unbelted occupants in car rollover accident mechanisms (P = 0.001, P = 0.01 and P = 0.008, respectively). CONCLUSIONS: During car crashes, especially car rollover, unbelted occupants are more likely to sustain multiple severe injuries and death. Law enforcement of the seat belt usage for all occupants (front and rear seat) is obligatory to reduce severe injuries sustained as a result of car accidents, especially in vehicles with low safety.

A seven-year study on head injuries in infants, Iran---the changing pattern.

OBJECTIVE: Head injury (HI) is the leading cause of mortality and life-long disability in infants. Infants have different anatomical and pathophysiological brain structures from other age groups. The aim of this study was to survey infant HI patients admitted to Shahid Behest Hospital in Kashan, Iran from 2004 to 2010, and to identify the causes of HIs in this age group. METHODS: In this retrospective study, all HI patients under the age of two who were hospitalized for more than 24 hours between January 2004 and January 2010 were enrolled in the study. Demographic, etiologic, and injury data were collected and a descriptive analysis was performed. RESULTS: Infants comprised 20.8% of all children (under 15 years old) with HIs and 65.1% of the injuries occurred in the home. Falls were the most common cause of injury (63.4%). In hospital mortality was 6.6 per 100 000 infants. A decreasing trend was seen in home events, but HIs caused by traffic accidents were increasing during the study period. The amount of HI infants resulting from car accidents has tripled from the years 2004 to 2010. CONCLUSION: Although home events and falling are the main causes of infant HIs and need attention, our study showed an increase of HIs caused by road traffic accidents, especially by car accidents, thus legislation for the implementation of protective equipment such as child safety seats and programs is urgently needed.

Cisplatin Resistant Patterns in Ovarian Cell Line Using FTIR and Principle Component Analysis.

Cisplatin is a common chemotherapeutic agent that used for treatment of many solid cancers. Rapid identification of chemotherapy resistance is very important and may lead to effective treatment plan. Spectroscopy techniques, such as infrared spectroscopy, which are sensitive to biochemical composition of samples, have shown potentials to discriminate tissues. Developing in Fourier transform infrared (FTIR) as a diagnostic tool support conventional technique in investigating cell phenotype. By this goal three different cell lines, two cisplatin resistant OV2008-DDP (C13) and A2780-CP ovarian cell lines and one cisplatin sensitive A2780 cell line were investigated by FTIR spectroscopy. Data were subjected to principle component analysis (PCA) to obtain FTIR pattern for cisplatin resistance. Using FTIR spectroscopy on these cells in the range of 400-4000 cm(-1) was shown dramatic change in cells. Results shows that Cisplatin resistance pattern is characterized in spectrum with the alteration of conformation in secondary structure of proteins and a shift toward the high wave numbers of CH2 stretching vibration. The FTIR data set between 1000 and 3000 cm(-1) could be consumed as biochemical typicality spectra among resistant and sensitive cell lines while correctly classified by PCA model. Our work supports the promise of PCA analysis of FTIR data as a powerful combined approach for the development of automated methods to recognize resistant to cisplatin in experimental cell lines. One of the advantages of this tool is to investigate the resistant percent of cancer cells .Such technique may bring new tool in cancer diagnosis and stage definition in cancerous tissues.