Computer-aided diagnosis of keratoconus through VAE-augmented images using deep learning.

Detecting clinical keratoconus (KCN) poses a challenging and time-consuming task. During the diagnostic process, ophthalmologists are required to review demographic and clinical ophthalmic examinations in order to make an accurate diagnosis. This study aims to develop and evaluate the accuracy of deep convolutional neural network (CNN) models for the detection of keratoconus (KCN) using corneal topographic maps. We retrospectively collected 1758 corneal images (978 normal and 780 keratoconus) from 1010 subjects of the KCN group with clinically evident keratoconus and the normal group with regular astigmatism. To expand the dataset, we developed a model using Variational Auto Encoder (VAE) to generate and augment images, resulting in a dataset of 4000 samples. Four deep learning models were used to extract and identify deep corneal features of original and synthesized images. We demonstrated that the utilization of synthesized images during training process increased classification performance. The overall average accuracy of the deep learning models ranged from 99% for VGG16 to 95% for EfficientNet-B0. All CNN models exhibited sensitivity and specificity above 0.94, with the VGG16 model achieving an AUC of 0.99. The customized CNN model achieved satisfactory results with an accuracy and AUC of 0.97 at a much faster processing speed compared to other models. In conclusion, the DL models showed high accuracy in screening for keratoconus based on corneal topography images. This is a development toward the potential clinical implementation of a more enhanced computer-aided diagnosis (CAD) system for KCN detection, which would aid ophthalmologists in validating the clinical decision and carrying out prompt and precise KCN treatment.

Application of ensemble machine learning approach to assess the factors affecting size and polydispersity index of liposomal nanoparticles.

Liposome nanoparticles have emerged as promising drug delivery systems due to their unique properties. Assessing particle size and polydispersity index (PDI) is critical for evaluating the quality of these liposomal nanoparticles. However, optimizing these parameters in a laboratory setting is both costly and time-consuming. This study aimed to apply a machine learning technique to assess the impact of specific factors, including sonication time, extrusion temperature, and compositions, on the size and PDI of liposomal nanoparticles. Liposomal solutions were prepared and subjected to sonication with varying values for these parameters. Two compositions: (A) HSPC:DPPG:Chol:DSPE-mPEG2000 at 55:5:35:5 molar ratio and (B) HSPC:Chol:DSPE-mPEG2000 at 55:40:5 molar ratio, were made using remote loading method. Ensemble learning (EL), a machine learning technique, was employed using the Least-squares boosting (LSBoost) algorithm to accurately model the data. The dataset was randomly split into training and testing sets, with 70% allocated for training. The LSBoost algorithm achieved mean absolute errors of 1.652 and 0.0105 for modeling the size and PDI, respectively. Under conditions where the temperature was set at approximately 60 °C, our EL model predicted a minimum particle size of 116.53 nm for composition (A) with a sonication time of approximately 30 min. Similarly, for composition (B), the model predicted a minimum particle size of 129.97 nm with sonication times of approximately 30 or 55 min. In most instances, a PDI of less than 0.2 was achieved. These results highlight the significant impact of optimizing independent factors on the characteristics of liposomal nanoparticles and demonstrate the potential of EL as a decision support system for identifying the best liposomal formulation. We recommend further studies to explore the effects of other independent factors, such as lipid composition and surfactants, on liposomal nanoparticle characteristics.

Optimizing nanoliposomal formulations: Assessing factors affecting entrapment efficiency of curcumin-loaded liposomes using machine learning.

BACKGROUND: Curcumin faces challenges in clinical applications due to its low bioavailability and poor water solubility. Liposomes have emerged as a promising delivery system for curcumin. This study aims to apply ensemble learning, a machine learning technique, to determine the most effective experimental conditions for formulating stable curcumin-loaded liposomes with a high entrapment efficiency (EE). METHODS: Two liposomal formulations composed of HSPC:DPPG:Chol:DSPE-mPEG2000 and HSPC:Chol:DSPE-mPEG2000 at 55:5:35:5 and 55:40:5 M ratios, respectively, were prepared using the remote loading method, and their particle size and polydispersity index (PDI) were determined using Dynamic Light Scattering. To model the impact of five factors (molar ratios, particle size, sonication time, pH, and PDI) on EE%, the Least-squares boosting (LSBoost) ensemble learning algorithm was employed due to its capability to effectively handle nonlinear and non-stationary problems. The implementation and optimization of LSBoost were performed using MATLAB R2020a. The dataset was randomly split into training and testing sets, with 70% allocated for training. The mean absolute error (MAE) was used as the cost function to evaluate model performance. Additionally, a novel approach was employed to visualize the results using 3D plots, facilitating practical interpretation. RESULTS: The optimal model exhibited an MAE of 3.61, indicating its robust predictive capability. The study identified several optimal conditions for achieving the highest EE value of 100%. However, to ensure both the highest EE value and a suitable particle size, it is recommended to set the following conditions: a molar ratio of 55:5:35:5, a PDI within the range of 0.09-0.13, a particle size of approximately 130 nm, a sonication time of 30 min, and a pH within the range of 7.2-8. It is worth mentioning that adjusting the molar ratio to 55:40:5 resulted in a maximum EE of 88.38%. CONCLUSION: These findings underscore the high performance of ensemble learning in accurately predicting and optimizing the EE of the curcumin-loaded liposomes. The application of this technique provides valuable insights and holds promise for the development of efficient drug delivery systems.

PTEN hamartoma tumour syndrome: case report based on data from the Iranian hereditary colorectal cancer registry and literature review.

BACKGROUND: PTEN hamartoma tumour syndrome (PHTS) is a rare hereditary disorder caused by germline pathogenic mutations in the PTEN gene. This study presents a case of PHTS referred for genetic evaluation due to multiple polyps in the rectosigmoid area, and provides a literature review of PHTS case reports published between March 2010 and March 2022. CASE PRESENTATION: A 39-year-old Iranian female with a family history of gastric cancer in a first-degree relative presented with minimal bright red blood per rectum and resistant dyspepsia. Colonoscopy revealed the presence of over 20 polyps in the rectosigmoid area, while the rest of the colon appeared normal. Further upper endoscopy showed multiple small polyps in the stomach and duodenum, leading to a referral for genetic evaluation of hereditary colorectal polyposis. Whole-exome sequencing led to a PHTS diagnosis, even though the patient displayed no clinical or skin symptoms of the condition. Further screenings identified early-stage breast cancer and benign thyroid nodules through mammography and thyroid ultrasound. METHOD AND RESULTS OF LITERATURE REVIEW: A search of PubMed using the search terms "Hamartoma syndrome, Multiple" [Mesh] AND "case report" OR "case series" yielded 43 case reports, predominantly in women with a median age of 39 years. The literature suggests that patients with PHTS often have a family history of breast, thyroid and endometrial neoplasms along with pathogenic variants in the PTEN/MMAC1 gene. Gastrointestinal polyps are one of the most common signs reported in the literature, and the presence of acral keratosis, trichilemmomas and mucocutaneous papillomas are pathognomonic characteristics of PHTS. CONCLUSION: When a patient presents with more than 20 rectosigmoid polyps, PHTS should be considered. In such cases, it is recommended to conduct further investigations to identify other potential manifestations and the phenotype of PHTS. Women with PHTS should undergo annual mammography and magnetic resonance testing for breast cancer screening from the age of 30, in addition to annual transvaginal ultrasounds and blind suction endometrial biopsies.

Corrigendum to "A Dynamic Model for Imputing Missing Medical Data: A Multiobjective Particle Swarm Optimization Algorithm".

[This corrects the article DOI: 10.1155/2021/1203726.].

Colorectal Cancer in North-Eastern Iran: a retrospective, comparative study of early-onset and late-onset cases based on data from the Iranian hereditary colorectal cancer registry.

BACKGROUND: The incidence rate of colorectal cancer (CRC) is increasing among patients below 50 years of age. The reason for this is unclear, but could have to do with the fact that indicative variables, such as tumour location, gender preference and genetic preponderance have not been followed up in a consistent mann er. The current study was primarily conducted to improve the hereditary CRC screening programme by assessing the demographic and clinicopathological characteristics of early-onset CRC compared to late-onset CRC in northeast Iran. METHODS: This retrospective study, carried out over a three-year follow-up period (2014-2017), included 562 consecutive CRCs diagnosed in three Mashhad city hospital laboratories in north-eastern Iran. We applied comparative analysis of pathological and hereditary features together with information on the presence of mismatch repair (MMR) gene deficiency with respect to recovery versus mortality. Patients with mutations resulting in absence of the MMR gene MLH1 protein product and normal BRAF status were considered to be at high risk of Lynch syndrome (LS). Analyses using R studio software were performed on early-onset CRC (n = 222) and late-onset CRC (n = 340), corresponding to patients ≤50 years of age and patients > 50 years. RESULTS: From an age-of-onset point of view, the distribution between the genders differed with females showing a higher proportion of early-onset CRC than men (56% vs. 44%), while the late-onset CRC disparity was less pronounced (48% vs. 52%). The mean age of all participants was 55.6 ± 14.8 years, with 40.3 ± 7.3 years for early-onset CRC and 65.1 ± 9.3 years for late-onset CRC. With respect to anatomical tumour location (distal, rectal and proximal), the frequencies were 61, 28 and 11%, respectively, but the variation did not reach statistical significance. However, there was a dramatic difference with regard to the history of CRC in second-degree relatives between two age categories, with much higher numbers of family-related CRCs in the early-onset group. Expression of the MLH1 and PMS2 genes were significantly different between recovered and deceased, while this finding was not observed with regard to the MSH6 and the MSH2 genes. Mortality was significantly higher in those at high risk of LS. CONCLUSION: The variation of demographic, pathological and genetic characteristics between early-onset and late-onset CRC emphasizes the need for a well-defined algorithm to identify high-risk patients.

Predicting the Relation between Biopsychosocial Factors and Type of Childbirth using the Decision Tree Method: A Cohort Study.

Background: With the growing rate of cesarean sections, rising morbidity and mortality thereafter is an important health issue. Predictive models can identify individuals with a higher probability of cesarean section, and help them make better decisions. This study aimed to investigate the biopsychosocial factors associated with the method of childbirth and designed a predictive model using the decision tree C4.5 algorithm. Methods: In this cohort study, the sample included 170 pregnant women in the third trimester of pregnancy referring to Shahroud Health Care Centers (Semnan, Iran), from 2018 to 2019. Blood samples were taken from mothers to measure the estrogen hormone at baseline. Birth information was recorded at the follow-up time per 30-42 days postpartum. Chi square, independent samples t test, and Mann-Whitney were used for comparisons between the two groups. Modeling was performed with the help of MATLAB software and C4.5 decision tree algorithm using input variables and target variable (childbirth method). The data were divided into training and testing datasets using the 70-30% method. In both stages, sensitivity, specificity, and accuracy were evaluated by the decision tree algorithm. Results: Previous method of childbirth, maternal body mass index at childbirth, maternal age, and estrogen were the most significant factors predicting the childbirth method. The decision tree model's sensitivity, specificity, and accuracy were 85.48%, 94.34%, and 89.57% in the training stage, and 82.35%, 83.87%, and 83.33% in the testing stage, respectively. Conclusion: The decision tree model was designed with high accuracy successfully predicted the method of childbirth. By recognizing the contributing factors, policymakers can take preventive action.It should be noted that this article was published in preprint form on the website of research square (https://www.researchsquare.com/article/rs-34770/v1).

A Dynamic Model for Imputing Missing Medical Data: A Multiobjective Particle Swarm Optimization Algorithm.

Missing data occurs in all research, especially in medical studies. Missing data is the situation in which a part of research data has not been reported. This will result in the incompatibility of the sample and the population and misguided conclusions. Missing data is usual in research, and the extent of it will determine how misinterpreted the conclusions will be. All methods of parameter estimation and prediction models are based on the assumption that the data are complete. Extensive missing data will result in false predictions and increased bias. In the present study, a novel method has been proposed for the imputation of medical missing data. The method determines what algorithm is suitable for the imputation of missing data. To do so, a multiobjective particle swarm optimization algorithm was used. The algorithm imputes the missing data in a way that if a prediction model is applied to the data, both specificity and sensitivity will be optimized. Our proposed model was evaluated using real data of gastric cancer and acute T-cell leukemia (ATLL). First, the model was then used to impute the missing data. Then, the missing data were imputed using deletion, average, expectation maximization, MICE, and missForest methods. Finally, the prediction model was applied for both imputed datasets. The accuracy of the prediction model for the first and the second imputation methods was 0.5 and 16.5, respectively. The novel imputation method was more accurate than similar algorithms like expectation maximization and MICE.

Predicting the incidence of COVID-19 using data mining.

BACKGROUND: The high prevalence of COVID-19 has made it a new pandemic. Predicting both its prevalence and incidence throughout the world is crucial to help health professionals make key decisions. In this study, we aim to predict the incidence of COVID-19 within a two-week period to better manage the disease. METHODS: The COVID-19 datasets provided by Johns Hopkins University, contain information on COVID-19 cases in different geographic regions since January 22, 2020 and are updated daily. Data from 252 such regions were analyzed as of March 29, 2020, with 17,136 records and 4 variables, namely latitude, longitude, date, and records. In order to design the incidence pattern for each geographic region, the information was utilized on the region and its neighboring areas gathered 2 weeks prior to the designing. Then, a model was developed to predict the incidence rate for the coming 2 weeks via a Least-Square Boosting Classification algorithm. RESULTS: The model was presented for three groups based on the incidence rate: less than 200, between 200 and 1000, and above 1000. The mean absolute error of model evaluation were 4.71, 8.54, and 6.13%, respectively. Also, comparing the forecast results with the actual values in the period in question showed that the proposed model predicted the number of globally confirmed cases of COVID-19 with a very high accuracy of 98.45%. CONCLUSION: Using data from different geographical regions within a country and discovering the pattern of prevalence in a region and its neighboring areas, our boosting-based model was able to accurately predict the incidence of COVID-19 within a two-week period.

Spatial-time analysis of cardiovascular emergency medical requests: enlightening policy and practice.

BACKGROUND: Response time to cardiovascular emergency medical requests is an important indicator in reducing cardiovascular disease (CVD) -related mortality. This study aimed to visualize the spatial-time distribution of response time, scene time, and call-to-hospital time of these emergency requests. We also identified patterns of clusters of CVD-related calls. METHODS: This cross-sectional study was conducted in Mashhad, north-eastern Iran, between August 2017 and December 2019. The response time to every CVD-related emergency medical request call was computed using spatial and classical statistical analyses. The Anselin Local Moran's I was performed to identify potential clusters in the patterns of CVD-related calls, response time, call-to-hospital arrival time, and scene-to-hospital arrival time at small area level (neighborhood level) in Mashhad, Iran. RESULTS: There were 84,239 CVD-related emergency request calls, 61.64% of which resulted in the transport of patients to clinical centers by EMS, while 2.62% of callers (a total of 2218 persons) died before EMS arrival. The number of CVD-related emergency calls increased by almost 7% between 2017 and 2018, and by 19% between 2017 and 2019. The peak time for calls was between 9 p.m. and 1 a.m., and the lowest number of calls were recorded between 3 a.m. and 9 a.m. Saturday was the busiest day of the week in terms of call volume. There were statistically significant clusters in the pattern of CVD-related calls in the south-eastern region of Mashhad. Further, we found a large spatial variation in scene-to-hospital arrival time and call-to-hospital arrival time in the area under study. CONCLUSION: The use of geographical information systems and spatial analyses in modelling and quantifying EMS response time provides a new vein of knowledge for decision makers in emergency services management. Spatial as well as temporal clustering of EMS calls were present in the study area. The reasons for clustering of unfavorable time indices for EMS response requires further exploration. This approach enables policymakers to design tailored interventions to improve response time and reduce CVD-related mortality.

Altered Expression of Cell Cycle Regulators in Adult T-Cell Leukemia/ Lymphoma Patients.

BACKGROUND: Adult T-cell leukemia/lymphoma (ATLL) is caused by human T-cell lymphotropic virus type-1 (HTLV-1). HTLV-1 oncogenes can induce malignancy through controlled gene expression of cell cycle checkpoints in the host cell. HTLV-I genes play a pivotal role in overriding cell cycle checkpoints and deregulate cellular division. In this study, we aimed to determine and compare the HTLV-1 proviral load and the gene expression levels of cyclin-dependent kinase-2 (CDK2), CDK4, p53, and retinoblastoma (Rb) in ATLL and carrier groups. METHODS: A total of twenty-five ATLL patients (12 females and 13 males) and 21 asymptomatic carriers (10 females and 11 males) were included in this study. TaqMan real-time polymerase chain reaction assay was used for evaluation of proviral load and gene expression levels of CDK2, CDK4, p53, and Rb. Statistical analysis was used to compare proviral load and gene expression levels between two groups, using SPSS version 18. RESULTS: The mean scores of the HTLV-1 proviral load in the ATLL patients and healthy carriers were 13067.20±6400.41 and 345.79±78.80 copies/104 cells, respectively (P=0.000). There was a significant correlation between the gene expression levels of CDK2 and CDK4 (P=0.01) in the ATLL group. CONCLUSION: Our findings demonstrated a significant difference between the ATLL patients and healthy carriers regarding the rate of proviral load and the gene expression levels of p53 and CDK4; accordingly, proviral load and expression levels of these genes may be useful in the assessment of disease progression and prediction of HTLV-1 infection outcomes.

Providing an imputation algorithm for missing values of longitudinal data using Cuckoo search algorithm: A case study on cervical dystonia.

BACKGROUND: Missing values in data are found in a large number of studies in the field of medical sciences, especially longitudinal ones, in which repeated measurements are taken from each person during the study. In this regard, several statistical endeavors have been performed on the concepts, issues, and theoretical methods during the past few decades. METHODS: Herein, we focused on the missing data related to patients excluded from longitudinal studies. To this end, two statistical parameters of similarity and correlation coefficient were employed. In addition, metaheuristic algorithms were applied to achieve an optimal solution. The selected metaheuristic algorithm, which has a great search functionality, was the Cuckoo search algorithm. RESULTS: Profiles of subjects with cervical dystonia (CD) were used to evaluate the proposed model after applying missingness. It was concluded that the algorithm used in this study had a higher accuracy (98.48%), compared with similar approaches. CONCLUSION: Concomitant use of similar parameters and correlation coefficients led to a significant increase in accuracy of missing data imputation.

Evaluation of the role of TAX, HBZ, and HTLV-1 proviral load on the survival of ATLL patients.

BACKGROUND: Adult T-cell leukemia/lymphoma (ATLL) is an aggressive malignancy with very poor prognosis and short survival, caused by the human T-lymphotropic virus type-1 (HTLV-1). The HTLV-1 biomarkers trans-activator x (TAX) and HTLV-1 basic leucine zipper factor (HBZ) are main oncogenes and life-threatening elements. This study aimed to assess the role of the TAX and HBZ genes and HTLV-1 proviral load (PVL) in the survival of patients with ATLL. METHODS: Forty-three HTLV-1-infected individuals, including 18 asymptomatic carriers (AC) and 25 ATLL patients (ATLL), were evaluated between 2011 and 2015. The mRNA expression of TAX and HBZ and the HTLV-1 PVL were measured by quantitative PCR. RESULTS: Significant differences in the mean expression levels of TAX and HBZ were observed between the two study groups (ATLL and AC, P=0.014 and P=0.000, respectively). In addition, the ATLL group showed a significantly higher PVL than AC (P=0.000). There was a significant negative relationship between PVL and survival among all study groups (P=0.047). CONCLUSION: The HTLV-1 PVL and expression of TAX and HBZ were higher in the ATLL group than in the AC group. Moreover, a higher PVL was associated with shorter survival time among all ATLL subjects. Therefore, measurement of PVL, TAX, and HBZ may be beneficial for monitoring and predicting HTLV-1-infection outcomes, and PVL may be useful for prognosis assessment of ATLL patients. This research demonstrates the possible correlation between these virological markers and survival in ATLL patients.

Predictive model for survival in patients with gastric cancer.

BACKGROUND AND AIM: Gastric cancer is one of the most prevalent cancers in the world. Characterized by poor prognosis, it is a frequent cause of cancer in Iran. The aim of the study was to design a predictive model of survival time for patients suffering from gastric cancer. METHODS: This was a historical cohort conducted between 2011 and 2016. Study population were 277 patients suffering from gastric cancer. Data were gathered from the Iranian Cancer Registry and the laboratory of Emam Reza Hospital in Mashhad, Iran. Patients or their relatives underwent interviews where it was needed. Missing values were imputed by data mining techniques. Fifteen factors were analyzed. Survival was addressed as a dependent variable. Then, the predictive model was designed by combining both genetic algorithm and logistic regression. Matlab 2014 software was used to combine them. RESULTS: Of the 277 patients, only survival of 80 patients was available whose data were used for designing the predictive model. Mean ?SD of missing values for each patient was 4.43?.41 combined predictive model achieved 72.57% accuracy. Sex, birth year, age at diagnosis time, age at diagnosis time of patients' family, family history of gastric cancer, and family history of other gastrointestinal cancers were six parameters associated with patient survival. CONCLUSION: The study revealed that imputing missing values by data mining techniques have a good accuracy. And it also revealed six parameters extracted by genetic algorithm effect on the survival of patients with gastric cancer. Our combined predictive model, with a good accuracy, is appropriate to forecast the survival of patients suffering from Gastric cancer. So, we suggest policy makers and specialists to apply it for prediction of patients' survival.