Use of a Deep Belief Network for Small High-Level Abstraction Data Sets Using Artificial Intelligence with Rule Extraction.

We describe a simple method to transfer from weights in deep neural networks (NNs) trained by a deep belief network (DBN) to weights in a backpropagation NN (BPNN) in the recursive-rule eXtraction (Re-RX) algorithm with J48graft (Re-RX with J48graft) and propose a new method to extract accurate and interpretable classification rules for rating category data sets. We apply this method to the Wisconsin Breast Cancer Data Set (WBCD), the Mammographic Mass Data Set, and the Dermatology Dataset, which are small, high-abstraction data sets with prior knowledge. After training these three data sets, our proposed rule extraction method was able to extract accurate and concise rules for deep NNs trained by a DBN. These results suggest that our proposed method could help fill the gap between the very high learning capability of DBNs and the very high interpretability of rule extraction algorithms such as Re-RX with J48graft.

Use of an artificial intelligence-based rule extraction approach to predict an emergency cesarean section.

OBJECTIVE: One of the major problems with artificial intelligence (AI) is that it is generally known as a "black box". Therefore, the present study aimed to construct an emergency cesarean section (CS) prediction system using an AI-based rule extraction approach as a "white box" to detect the cause for the emergency CS. METHODS: Data were collected from all perinatal records of all delivery outcomes at Osaka Medical College between December 2014 and July 2019. We identified the delivery method for all deliveries after 36 gestational weeks as either (1) vaginal delivery or scheduled CS, or (2) emergency CS. From among these, we selected 52 risk factors to feed into an AI-based rule extraction algorithm to extract rules to predict an emergency CS. RESULTS: We identified 1513 singleton deliveries (1285 [84.9%] vaginal deliveries, 228 emergency CS [15.1%]) and extracted 15 rules. We achieved an average accuracy of 81.90% using five-fold cross-validation and an area under the receiving operating characteristic curve of 71.46%. CONCLUSION: To our knowledge, this is the first study to use interpretable AI-based rule extraction technology to predict an emergency CS. This system appears to be useful for identifying hidden factors for emergency CS.

Detection of Lower Albuminuria Levels and Early Development of Diabetic Kidney Disease Using an Artificial Intelligence-Based Rule Extraction Approach.

The aim of the present study was to determine the lowest cut-off value for albuminuria levels, which can be used to detect diabetic kidney disease (DKD) using the urinary albumin-to-creatinine ratio (UACR). National Health and Nutrition Examination Survey (NHANES) data for US adults were used, and participants were classified as having diabetes or prediabetes based on a self-report and physiological measures. The study dataset comprised 942 diabetes and 524 prediabetes samples. This study clarified the significance of the lower albuminuria (UACR) levels, which can detect DKD, using an artificial intelligence-based rule extraction approach. The diagnostic rules (15 concrete rules) for both samples were extracted using a recursive-rule eXtraction (Re-RX) algorithm with continuous attributes (continuous Re-RX) to discriminate between prediabetes and diabetes datasets. Continuous Re-RX showed high test accuracy (77.56%) and a large area under the receiver operating characteristics curve (75%), which derived the two cut-off values (6.1 mg/g Cr and 71.00 mg/g Cr) for the lower albuminuria level in the UACR to detect early development of DKD. The early cut-off values for normoalbuminuria (NA) and microalbuminuria (MA) will be determined to help detect CKD and DKD, and to detect diabetes before MA develop and to prevent diabetic complications.

The Right Direction Needed to Develop White-Box Deep Learning in Radiology, Pathology, and Ophthalmology: A Short Review.

The popularity of deep learning (DL) in the machine learning community has been dramatically increasing since 2012. The theoretical foundations of DL are well-rooted in the classical neural network (NN). Rule extraction is not a new concept, but was originally devised for a shallow NN. For about the past 30 years, extensive efforts have been made by many researchers to resolve the "black box" problem of trained shallow NNs using rule extraction technology. A rule extraction technology that is well-balanced between accuracy and interpretability has recently been proposed for shallow NNs as a promising means to address this black box problem. Recently, we have been confronting a "new black box" problem caused by highly complex deep NNs (DNNs) generated by DL. In this paper, we first review four rule extraction approaches to resolve the black box problem of DNNs trained by DL in computer vision. Next, we discuss the fundamental limitations and criticisms of current DL approaches in radiology, pathology, and ophthalmology from the black box point of view. We also review the conversion methods from DNNs to decision trees and point out their limitations. Furthermore, we describe a transparent approach for resolving the black box problem of DNNs trained by a deep belief network. Finally, we provide a brief description to realize the transparency of DNNs generated by a convolutional NN and discuss a practical way to realize the transparency of DL in radiology, pathology, and ophthalmology.

Greedy rule generation from discrete data and its use in neural network rule extraction.

This paper proposes a GRG (Greedy Rule Generation) algorithm, a new method for generating classification rules from a data set with discrete attributes. The algorithm is "greedy" in the sense that at every iteration, it searches for the best rule to generate. The criteria for the best rule include the number of samples and the size of subspaces that it covers, as well as the number of attributes in the rule. This method is employed for extracting rules from neural networks that have been trained and pruned for solving classification problems. The classification rules are extracted from the neural networks using the standard decompositional approach. Neural networks with one hidden layer are trained and the proposed GRG algorithm is applied to their discretized hidden unit activation values. Our experimental results show that neural network rule extraction with the GRG method produces rule sets that are accurate and concise. Application of GRG directly on three medical data sets with discrete attributes also demonstrates its effectiveness for rule generation.

Non-invasive prediction of non-alcoholic steatohepatitis in Japanese patients with morbid obesity by artificial intelligence using rule extraction technology.

AIM: To construct a non-invasive prediction algorithm for predicting non-alcoholic steatohepatitis (NASH), we investigated Japanese morbidly obese patients using artificial intelligence with rule extraction technology. METHODS: Consecutive patients who required bariatric surgery underwent a liver biopsy during the operation. Standard clinical, anthropometric, biochemical measurements were used as parameters to predict NASH and were analyzed using rule extraction technology. One hundred and two patients, including 79 NASH and 23 non-NASH patients were analyzed in order to create the prediction model, another cohort with 77 patients including 65 NASH and 12 non-NASH patients were analyzed to validate the algorithm. RESULTS: Alanine aminotransferase, C-reactive protein, homeostasis model assessment insulin resistance, albumin were extracted as predictors of NASH using a recursive-rule extraction algorithm. When we adopted the extracted rules for the validation cohort using a highly accurate rule extraction algorithm, the predictive accuracy was 79.2%. The positive predictive value, negative predictive value, sensitivity and specificity were 88.9%, 35.7%, 86.2% and 41.7%, respectively. CONCLUSION: We successfully generated a useful model for predicting NASH in Japanese morbidly obese patients based on their biochemical profile using a rule extraction algorithm.

Volumetric brain tumour detection from MRI using visual saliency.

Medical image processing has become a major player in the world of automatic tumour region detection and is tantamount to the incipient stages of computer aided design. Saliency detection is a crucial application of medical image processing, and serves in its potential aid to medical practitioners by making the affected area stand out in the foreground from the rest of the background image. The algorithm developed here is a new approach to the detection of saliency in a three dimensional multi channel MR image sequence for the glioblastoma multiforme (a form of malignant brain tumour). First we enhance the three channels, FLAIR (Fluid Attenuated Inversion Recovery), T2 and T1C (contrast enhanced with gadolinium) to generate a pseudo coloured RGB image. This is then converted to the CIE L*a*b* color space. Processing on cubes of sizes k = 4, 8, 16, the L*a*b* 3D image is then compressed into volumetric units; each representing the neighbourhood information of the surrounding 64 voxels for k = 4, 512 voxels for k = 8 and 4096 voxels for k = 16, respectively. The spatial distance of these voxels are then compared along the three major axes to generate the novel 3D saliency map of a 3D image, which unambiguously highlights the tumour region. The algorithm operates along the three major axes to maximise the computation efficiency while minimising loss of valuable 3D information. Thus the 3D multichannel MR image saliency detection algorithm is useful in generating a uniform and logistically correct 3D saliency map with pragmatic applicability in Computer Aided Detection (CADe). Assignment of uniform importance to all three axes proves to be an important factor in volumetric processing, which helps in noise reduction and reduces the possibility of compromising essential information. The effectiveness of the algorithm was evaluated over the BRATS MICCAI 2015 dataset having 274 glioma cases, consisting both of high grade and low grade GBM. The results were compared with that of the 2D saliency detection algorithm taken over the entire sequence of brain data. For all comparisons, the Area Under the receiver operator characteristic (ROC) Curve (AUC) has been found to be more than 0.99 ± 0.01 over various tumour types, structures and locations.

The number of microvascular complications is associated with an increased risk for severity of periodontitis in type 2 diabetes patients: Results of a multicenter hospital-based cross-sectional study.

AIMS/INTRODUCTION: To explore the relationships between periodontitis and microvascular complications as well as glycemic control in type 2 diabetes patients. MATERIALS AND METHODS: This multicenter, hospital-based, cross-sectional study included 620 patients with type 2 diabetes. We compared the prevalence and severity of periodontitis between patients with ≥1 microvascular complication and those without microvascular complications. We also compared the prevalence and severity of periodontitis among patients with different degrees of glycemic control. RESULTS: After adjusting for confounding factors, multiple logistic regression analysis showed that the severity of periodontitis was significantly associated with the number of microvascular complications (odds ratio 1.3, 95% confidence interval 1.1-1.6), glycated hemoglobin ≥8.0% (64 mmol/mol; odds ratio 1.6; 95% confidence interval 1.1-2.3), and older age (≥50 years; odds ratio 1.7; 95% confidence interval 1.1-2.6). However, the prevalence of periodontitis was not significantly associated with the number of microvascular complications, but was associated with male sex, high glycated hemoglobin (≥8.0% [64 mmol/mol]), older age (≥40 years), longer duration of diabetes (≥15 years) and fewer teeth (≤25). Furthermore, propensity score matching for age, sex, diabetes duration and glycated hemoglobin showed that the incidence of severe periodontitis was significantly higher among patients with microvascular complications than among those without microvascular complications (P < 0.05). CONCLUSIONS: The number of microvascular complications is a risk factor for more severe periodontitis in patients with type 2 diabetes, whereas poor glycemic control is a risk factor for increased prevalence and severity of periodontitis.

A Novel GBM Saliency Detection Model Using Multi-Channel MRI.

The automatic computerized detection of regions of interest (ROI) is an important step in the process of medical image processing and analysis. The reasons are many, and include an increasing amount of available medical imaging data, existence of inter-observer and inter-scanner variability, and to improve the accuracy in automatic detection in order to assist doctors in diagnosing faster and on time. A novel algorithm, based on visual saliency, is developed here for the identification of tumor regions from MR images of the brain. The GBM saliency detection model is designed by taking cue from the concept of visual saliency in natural scenes. A visually salient region is typically rare in an image, and contains highly discriminating information, with attention getting immediately focused upon it. Although color is typically considered as the most important feature in a bottom-up saliency detection model, we circumvent this issue in the inherently gray scale MR framework. We develop a novel pseudo-coloring scheme, based on the three MRI sequences, viz. FLAIR, T2 and T1C (contrast enhanced with Gadolinium). A bottom-up strategy, based on a new pseudo-color distance and spatial distance between image patches, is defined for highlighting the salient regions in the image. This multi-channel representation of the image and saliency detection model help in automatically and quickly isolating the tumor region, for subsequent delineation, as is necessary in medical diagnosis. The effectiveness of the proposed model is evaluated on MRI of 80 subjects from the BRATS database in terms of the saliency map values. Using ground truth of the tumor regions for both high- and low- grade gliomas, the results are compared with four highly referred saliency detection models from literature. In all cases the AUC scores from the ROC analysis are found to be more than 0.999 ± 0.001 over different tumor grades, sizes and positions.

A single bout of exercise at higher intensity enhances glucose effectiveness in sedentary men.

OBJECTIVE: Previous studies have shown that glucose effectiveness and insulin sensitivity are acutely enhanced by exercise at various intensities. The aim of this study was to determine the effects of a single bout of exercise at intensities recommended by the American Diabetes Association (ADA) and the American College of Sports Medicine (ACSM) on glucose uptake-specific glucose effectiveness (S(G)(2)*) and insulin sensitivity (S(I)(2)*). S(G)(2)* and S(I)(2)* were estimated by a two-compartment minimal model. DESIGN: Six healthy men (age, 28.5 +/- 2.0 yr) performed a stable-labeled frequently sampled iv glucose tolerance test (FSIGT) under three separate conditions: without any prior exercise, and immediately after single 20-min bouts of cycle ergometer exercise at an intensity of 50% and 70% of maximal oxygen uptake (Vo(2max)). The exercise intensities were close to the lower and upper boundaries recommended by the ADA and ACSM. RESULTS: Glucose disappearance constant (K(G)), S(G)(2)*, and S(I)(2)* increased after exercise in an intensity-dependent manner. Increases in S(G)(2)* (+237.1 +/- 50.5%), S(I)(2)* (+225.6 +/- 51.9%), and K(G) (+151.7 +/- 16.5%) following exercise at 70% Vo(2max) were statistically significant (P < 0.05), whereas those at 50% Vo(2max) were not. CONCLUSIONS: In conclusion, a single bout of exercise acutely improves S(I)(2)* and S(G)(2)* in individuals with normal glucose tolerance in an intensity-dependent manner.

[Clinical significant of semiquantificating DNA topoisomerase- I mRNA in colorectal cancer].

PURPOSE: To examine the clinical significance of determining the expression levels of DNA topoisomerase- I (topo-I) mRNA of colorectal cancer. METHODS: The relative expression levels of topo-I mRNA in primary colorectal cancer and adjacent normal mucosa were semiquantificated by the RT-PCR method. The relative expression of thymidylate synthase (TS) mRNA of the primary lesions was also examined. RESULTS: The topo- I mRNA expression was higher in the tumorous tissue than in the normal mucosa (n=22, p<0.01). The topo- I mRNA expression did not significantly correlate with 9 clinicopathological variables examined (n=22). In patients who received irinotecan hydrochloride (CPT-11) following the failure of 5-fluorouracil-based treatment, the topoI mRNA expression did not differ nor correlate with the response to CPT-11 (PR, n=14; SD, n=11; PR; n=24) (p=0.91). In these patients, there was no relationship between the topo I mRNA expression and the TS mRNA expression (p=0.22, r=0.18). In addition, the efficacy of CPT-11 did not correlate with combinations subdivided according to the expression levels of topo- I mRNA and TS mRNA. CONCLUSIONS: Determination of topo- I mRNA levels of primary colorectal cancer may not be useful for predicting the efficacy of CPT-11 treatment alone or in combination with TS mRNA levels.

[Expression of mRNA levels of thymidylate synthase, dihydropyrimidine dehydrogenase, and orotate phosphoribosyltransferase of colorectal cancer--relationships among mRNA levels in association with response to 5-FU based treatment].

We semiquantificated the mRNA levels of thymidylate synthase (TS), dihydropyrimidine dehydrogenase (DPD), and orotate phosphoribosyltransferase(OPRT) from the frozen tissue of colorectal cancer, in order to examine the relationships among these enzymes and the efficacy for 5-fluorouracil (FU)-based treatment in terms of the combinations of the expression levels of these enzymes. There were weak and positive relationships among TS mRNA expression and OPRT mRNA expression or DPD mRNA expression, while there was no significant relationship between DPD mRNA expression and OPRT mRNA expression (n = 112). In 39 patients who received 5-FU-based chemotherapy with evaluable lesions, patients with low TS expression (n = 23) showed a higher response rate (52%) as compared to those with high TS expression (13%, p = 0.02). Patients with low TS, low DPD, or high DPD expression tended to show a higher response rate (50%) than those with the other combinations (26%, p = 0.09). In addition, all of the responders with only one favorable factor (n = 4) were associated with low TS or low DPD expression. In conclusion, low TS expression followed by low DPD expression is important to predict the efficacy of 5-FU-based treatment for colorectal cancer. However, it may be of little significance to semiquantify the expression of OPRT.

Knowledge acquisition in the fuzzy knowledge representation framework of a medical consultation system.

This paper describes the fuzzy knowledge representation framework of the medical computer consultation system MedFrame/CADIAG-IV as well as the specific knowledge acquisition techniques that have been developed to support the definition of knowledge concepts and inference rules. As in its predecessor system CADIAG-II, fuzzy medical knowledge bases are used to model the uncertainty and the vagueness of medical concepts and fuzzy logic reasoning mechanisms provide the basic inference processes. The elicitation and acquisition of medical knowledge from domain experts has often been described as the most difficult and time-consuming task in knowledge-based system development in medicine. It comes as no surprise that this is even more so when unfamiliar representations like fuzzy membership functions are to be acquired. From previous projects we have learned that a user-centered approach is mandatory in complex and ill-defined knowledge domains such as internal medicine. This paper describes the knowledge acquisition framework that has been developed in order to make easier and more accessible the three main tasks of: (a) defining medical concepts; (b) providing appropriate interpretations for patient data; and (c) constructing inferential knowledge in a fuzzy knowledge representation framework. Special emphasis is laid on the motivations for some system design and data modeling decisions. The theoretical framework has been implemented in a software package, the Knowledge Base Builder Toolkit. The conception and the design of this system reflect the need for a user-centered, intuitive, and easy-to-handle tool. First results gained from pilot studies have shown that our approach can be successfully implemented in the context of a complex fuzzy theoretical framework. As a result, this critical aspect of knowledge-based system development can be accomplished more easily.

Combining neural network predictions for medical diagnosis.

We present our results from combining the predictions of an ensemble of neural networks for the diagnosis of hepatobiliary disorders. To improve the accuracy of the diagnosis, we train the second level networks using the outputs of the first level networks as input data. The second level networks achieve an accuracy that is higher than that of the individual networks in the first level. Compared to the simple method which averages the outputs of the first level networks, the second level networks are also more accurate. We discuss how the overall predictive accuracy can be improved by introducing bias during the training of the level one networks.

[Efficacy of hepatic arterial infusion of adriamycin and mitomycin C mixed with degradable starch microspheres for liver metastasis of colorectal cancer--correlation with the mRNA expression of DNA topoisomerase-IIalpha and glutathione-S transferase-pi in primary lesions].

DNA topoisomerase-IIalpha (topo-IIalpha) is a target enzyme of adriamycin (ADM). Glutathione-S-transferase-pi is known to be correlated with the resistance of various anticancer drugs including mitomycin C (MMC) and ADM. Expression levels of topo-IIalpha and GST-pi mRNA of primary colorectal lesions were semi-quantitatively determined by the RT-PCR method in 22 patients with colorectal cancer, who underwent hepatic arterial infusion of ADM and MMC mixed with degradable starch microspheres for synchronous (n=17) or metachronous (n=5) liver metastasis. Expression of topo-IIalpha mRNA/beta-actin mRNA was 0.872+/-0.564 (mean+/-SD) in responders (PR, n=10) and 0.369+/-0.133 in non-responders (SD+PD, n=12) (p=0.047). The relative expression of GST-pi was 0.638+/-0.593 in responders and 1.014+/-0.682 in non-responders (p=0.22). These results suggest that determining the mRNA expression of topo-IIalpha is useful for predicting the efficacy for this regimen, whereas determining the mRNA expression of GST-pi is not.

Evaluation of effects of multiple candidate genes (LEP, LEPR, MC4R, PIK3C3, and VRTN) on production traits in Duroc pigs.

We evaluated multiple effects of genetic variations of five candidate loci (LEP, LEPR, MC4R, PIK3C3 and VRTN) on four production traits (average daily weight gain (ADG); backfat thickness (BFT); loin eye muscle area (EMA); and intramuscular fat content (IMF)) in a closed nucleus herd of pure Duroc pigs. Polymorphisms in LEPR, MC4R and PIK3C3 had significant single gene effects on ADG and BFT. The additive genetic variance in ADG and BFT (16.99% and 22.51%, respectively) was explained by genetic effects of these three loci. No correlations were observed between the LEP genotype and production traits in this study. Although we detected marginally epistatic interactions between LEPR and PIK3C3 on the eye muscle area, there were no significant epistatic effects on any traits among all loci pairs. These results suggest that LEPR, MC4R, PIK3C3 and VRTN may independently influence growth rate and fat deposition. Furthermore, the statistical models for predicting the breeding values of each trait had the lowest Akaike's information criterion values when considering the effect of the MC4R, LEPR, PIK3C3 and VRTN genotype simultaneously. These results suggest that LEPR, MC4R, PIK3C3 and VRTN are useful markers for accurately predicting breeding values in Duroc pigs.

Association of swine vertnin (VRTN) gene with production traits in Duroc pigs improved using a closed nucleus breeding system.

Vertnin (VRTN) is involved in the variation of vertebral number in pigs and it is located on Sus scrofa chromosome 7. Vertebral number is related to body size in pigs, and many reports have suggested presence of an association between body length (BL) and meat production traits. Therefore, we analyzed the relationship between the VRTN genotype and the production and body composition traits in purebred Duroc pigs. Intramuscular fat content (IMF) in the Longissimus muscle was significantly associated with the VRTN genotype. The mean IMF of individuals with the wild-type genotype (Wt/Wt) (5.22%) was greater than that of individuals with the Wt/Q (4.99%) and Q/Q genotypes (4.79%). In addition, a best linear unbiased predictor of multiple traits animal model showed that the Wt allele had a positive effect on the IMF breeding value. No associations were observed between the VRTN genotype and other production traits. The VRTN genotype was related to BL. The Q/Q genotype individuals (100.0 cm) were longer than individuals with the Wt/Q (99.5 cm) and Wt/Wt genotypes (98.9 cm). These results suggest that in addition to the maintenance of an appropriate backfat thickness value, VRTN has the potential to act as a genetic marker of IMF.

Genetic networks and soft computing.

The analysis of gene regulatory networks provides enormous information on various fundamental cellular processes involving growth, development, hormone secretion, and cellular communication. Their extraction from available gene expression profiles is a challenging problem. Such reverse engineering of genetic networks offers insight into cellular activity toward prediction of adverse effects of new drugs or possible identification of new drug targets. Tasks such as classification, clustering, and feature selection enable efficient mining of knowledge about gene interactions in the form of networks. It is known that biological data is prone to different kinds of noise and ambiguity. Soft computing tools, such as fuzzy sets, evolutionary strategies, and neurocomputing, have been found to be helpful in providing low-cost, acceptable solutions in the presence of various types of uncertainties. In this paper, we survey the role of these soft methodologies and their hybridizations, for the purpose of generating genetic networks.

Influence of acute aerobic exercise on adiponectin oligomer concentrations in middle-aged abdominally obese men.

Exercise intensity may induce changes in total adiponectin and adiponectin oligomer levels. However, the effects of acute aerobic exercise on total adiponectin and adiponectin oligomers in middle-aged abdominally obese men remain unknown. The purpose of this study was to investigate the influence of aerobic exercise intensity on changes in the concentrations of total adiponectin and adiponectin oligomers (high-molecular weight [HMW] and middle- plus low-molecular weight [MLMW] adiponectin), and the endocrine mechanisms involved in exercise-induced changes in adiponectin oligomer profiles in middle-aged abdominally obese men. Using a crossover design, 9 middle-aged abdominally obese men (age, 54.1 ± 2.4 years; body mass index, 27.9 ± 0.6 kg/m²) underwent 2 trials that consisted of 60 minutes of stationary cycle exercise at either moderate-intensity (ME) or high-intensity (HE) aerobic exercise (50% or 70% of peak oxygen uptake, respectively). Blood samples were collected to measure the concentrations of adiponectin oligomers, hormones (catecholamines, insulin, and growth hormone), metabolites (free fatty acid, glycerol, triglyceride, and glucose), and cytokines (interleukin-6 and tumor necrosis factor-α). After exercise, plasma catecholamine concentrations were higher during HE than during ME (P < .05). Total adiponectin concentration decreased at the end of HE (P < .05), but remained unchanged after ME. The HMW adiponectin concentration did not change at either intensity, whereas the MLMW concentration decreased at the end of HE (P < .05). The ratio of HMW to total adiponectin concentration increased significantly (P < .05), whereas the ratio of MLMW to total adiponectin concentration decreased significantly (P < .05), at the end of HE. The percentage changes in epinephrine concentration from baseline to the end of exercise were correlated with the percentage changes in total adiponectin concentration (r = -0.67, P < .05) and MLMW adiponectin concentration (r = -0.82, P < .05) from baseline to the end of HE. Our results indicate that the change in total adiponectin was mainly due to a change in MLMW adiponectin concentration during high-intensity exercise in middle-aged abdominally obese men. Epinephrine may partially regulate the decrease in total and MLMW adiponectin concentrations during high-intensity exercise.