Computer-aided diagnosis using embedded ensemble deep learning for multiclass drug-resistant tuberculosis classification.

Introduction: This study aims to develop a web application, TB-DRD-CXR, for the categorization of tuberculosis (TB) patients into subgroups based on their level of drug resistance. The application utilizes an ensemble deep learning model that classifies TB strains into five subtypes: drug sensitive tuberculosis (DS-TB), drug resistant TB (DR-TB), multidrug-resistant TB (MDR-TB), pre-extensively drug-resistant TB (pre-XDR-TB), and extensively drug-resistant TB (XDR-TB). Methods: The ensemble deep learning model employed in the TB-DRD-CXR web application incorporates novel fusion techniques, image segmentation, data augmentation, and various learning rate strategies. The performance of the proposed model is compared with state-of-the-art techniques and standard homogeneous CNN architectures documented in the literature. Results: Computational results indicate that the suggested method outperforms existing methods reported in the literature, providing a 4.0%-33.9% increase in accuracy. Moreover, the proposed model demonstrates superior performance compared to standard CNN models, including DenseNet201, NASNetMobile, EfficientNetB7, EfficientNetV2B3, EfficientNetV2M, and ConvNeXtSmall, with accuracy improvements of 28.8%, 93.4%, 2.99%, 48.0%, 4.4%, and 7.6% respectively. Conclusion: The TB-DRD-CXR web application was developed and tested with 33 medical staff. The computational results showed a high accuracy rate of 96.7%, time-based efficiency (ET) of 4.16 goals/minutes, and an overall relative efficiency (ORE) of 100%. The system usability scale (SUS) score of the proposed application is 96.7%, indicating user satisfaction and a likelihood of recommending the TB-DRD-CXR application to others based on previous literature.

Embedded AMIS-Deep Learning with Dialog-Based Object Query System for Multi-Class Tuberculosis Drug Response Classification.

A person infected with drug-resistant tuberculosis (DR-TB) is the one who does not respond to typical TB treatment. DR-TB necessitates a longer treatment period and a more difficult treatment protocol. In addition, it can spread and infect individuals in the same manner as regular TB, despite the fact that early detection of DR-TB could reduce the cost and length of TB treatment. This study provided a fast and effective classification scheme for the four subtypes of TB: Drug-sensitive tuberculosis (DS-TB), drug-resistant tuberculosis (DR-TB), multidrug-resistant tuberculosis (MDR-TB), and extensively drug-resistant tuberculosis (XDR-TB). The drug response classification system (DRCS) has been developed as a classification tool for DR-TB subtypes. As a classification method, ensemble deep learning (EDL) with two types of image preprocessing methods, four convolutional neural network (CNN) architectures, and three decision fusion methods have been created. Later, the model developed by EDL will be included in the dialog-based object query system (DBOQS), in order to enable the use of DRCS as the classification tool for DR-TB in assisting medical professionals with diagnosing DR-TB. EDL yields an improvement of 1.17-43.43% over the existing methods for classifying DR-TB, while compared with classic deep learning, it generates 31.25% more accuracy. DRCS was able to increase accuracy to 95.8% and user trust to 95.1%, and after the trial period, 99.70% of users were interested in continuing the utilization of the system as a supportive diagnostic tool.

Drug-Resistant Tuberculosis Treatment Recommendation, and Multi-Class Tuberculosis Detection and Classification Using Ensemble Deep Learning-Based System.

This research develops the TB/non-TB detection and drug-resistant categorization diagnosis decision support system (TB-DRC-DSS). The model is capable of detecting both TB-negative and TB-positive samples, as well as classifying drug-resistant strains and also providing treatment recommendations. The model is developed using a deep learning ensemble model with the various CNN architectures. These architectures include EfficientNetB7, mobileNetV2, and Dense-Net121. The models are heterogeneously assembled to create an effective model for TB-DRC-DSS, utilizing effective image segmentation, augmentation, and decision fusion techniques to improve the classification efficacy of the current model. The web program serves as the platform for determining if a patient is positive or negative for tuberculosis and classifying several types of drug resistance. The constructed model is evaluated and compared to current methods described in the literature. The proposed model was assessed using two datasets of chest X-ray (CXR) images collected from the references. This collection of datasets includes the Portal dataset, the Montgomery County dataset, the Shenzhen dataset, and the Kaggle dataset. Seven thousand and eight images exist across all datasets. The dataset was divided into two subsets: the training dataset (80%) and the test dataset (20%). The computational result revealed that the classification accuracy of DS-TB against DR-TB has improved by an average of 43.3% compared to other methods. The categorization between DS-TB and MDR-TB, DS-TB and XDR-TB, and MDR-TB and XDR-TB was more accurate than with other methods by an average of 28.1%, 6.2%, and 9.4%, respectively. The accuracy of the embedded multiclass model in the web application is 92.6% when evaluated with the test dataset, but 92.8% when evaluated with a random subset selected from the aggregate dataset. In conclusion, 31 medical staff members have evaluated and utilized the online application, and the final user preference score for the web application is 9.52 out of a possible 10.

Combining metabolic engineering and evolutionary adaptation in Klebsiella oxytoca KMS004 to significantly improve optically pure D-(-)-lactic acid yield and specific productivity in low nutrient medium.

In this study, K. oxytoca KMS004 (ΔadhE Δpta-ackA) was further reengineered by the deletion of frdABCD and pflB genes to divert carbon flux through D-(-)-lactate production. During fermentation of high glucose concentration, the resulted strain named K. oxytoca KIS004 showed poor in growth and glucose consumption due to its insufficient capacity to generate acetyl-CoA for biosynthesis. Evolutionary adaptation was thus employed with the strain to overcome impaired growth and acetate auxotroph. The evolved K. oxytoca KIS004-91T strain exhibited significantly higher glucose-utilizing rate and D-(-)-lactate production as a primary route to regenerate NAD+. D-(-)-lactate at concentration of 133 g/L (1.48 M), with yield and productivity of 0.98 g/g and 2.22 g/L/h, respectively, was obtained by the strain. To the best of our knowledge, this strain provided a relatively high specific productivity of 1.91 g/gCDW/h among those of other previous works. Cassava starch was also used to demonstrate a potential low-cost renewable substrate for D-(-)-lactate production. Production cost of D-(-)-lactate was estimated at $3.72/kg. Therefore, it is possible for the KIS004-91T strain to be an alternative biocatalyst offering a more economically competitive D-(-)-lactate production on an industrial scale. KEY POINTS: • KIS004-91T produced optically pure D-(-)-lactate up to 1.48 M in a low salts medium. • It possessed the highest specific D-(-)-lactate productivity than other reported strains. • Cassava starch as a cheap and renewable substrate was used for D-(-)-lactate production. • Costs related to media, fermentation, purification, and waste disposal were reduced.

Evaluation of inhibitory effect and feasible utilization of dilute acid-pretreated rice straws on succinate production by metabolically engineered Escherichia coli AS1600a.

This work demonstrated a pioneer work in the pre-treatment of rice straw by phosphoric acid (H3PO4) for succinate production. The optimized pre-treatment condition of rice straw was at 121 °C for 30 min with 2 N H3PO4. With this condition, total sugar concentration of 31.2 g/L with the highest hemicellulose saccharification yield of 94% was obtained. The physicochemical analysis of the pre-treated rice straw showed significant changes in its structure thus enhancing enzymatic saccharification. Succinate concentrations of 78.5 and 63.8 g/L were produced from hydrolysate liquor (L) and solid fraction (S) of the pre-treated rice straw respectively, with a comparable yield of 86% by E. coli AS1600a. Use of a combined L + S fraction in simultaneous saccharification and fermentation (LS + SSF) further improved succinate production at a concentration and yield of 85.6 g/L and 90% respectively. The results suggested that H3PO4 pre-treated rice straw may be utilized for economical succinate production by E. coli AS1600a.

Optimization of sodium hydroxide pretreatment and enzyme loading for efficient hydrolysis of rice straw to improve succinate production by metabolically engineered Escherichia coli KJ122 under simultaneous saccharification and fermentation.

Rice straw was pretreated with sodium hydroxide (NaOH) before subsequent use for succinate production by Escherichia coli KJ122 under simultaneous saccharification and fermentation (SSF). The NaOH pretreated rice straw was significantly enhanced lignin removal up to 95%. With the optimized enzyme loading of 4% cellulase complex + 0.5% xylanase (endo-glucanase 67 CMC-U/g, ß-glucosidase 26 pNG-U/g and xylanase 18 CMC-U/g dry biomass), total sugar conversion reached 91.7 ±â€¯0.8% (w/w). The physicochemical analysis of NaOH pretreated rice straw indicated dramatical changes in its structure, thereby favoring enzymatic saccharification. In batch SSF, succinate production of 69.8 ±â€¯0.3 g/L with yield and productivity of 0.84 g/g pretreated rice straw and 0.76 ±â€¯0.02 g/L/h, respectively, was obtained. Fed-batch SSF significantly improved succinate concentration and productivity to 103.1 ±â€¯0.4 g/L and 1.37 ±â€¯0.07 g/L/h with a comparable yield. The results demonstrated a feasibility of sequential saccharification and fermentation of rice straw as a promising process for succinate production in industrial scale.

Re-engineering Escherichia coli KJ122 to enhance the utilization of xylose and xylose/glucose mixture for efficient succinate production in mineral salt medium.

Escherichia coli KJ122 was previously engineered to produce high concentration and yield of succinate in mineral salt medium containing glucose and sucrose under anaerobic conditions. However, this strain does not efficiently utilize xylose. To improve the xylose uptake and utilization in the strain KJ122, xylFGH and xylE genes were individually and simultaneously deleted. E. coli KJ12201 (KJ122::ΔxylFGH) exhibited superior abilities in growth, xylose consumption, and succinate production compared to those of the parental strain KJ122. However, E. coli KJ12202 (KJ122::ΔxylE) lessened xylose consumption due to an ATP deficit for metabolizing xylose thus making succinate production from xylose not preferable. Moreover, E. coli KJ12203 (KJ122::ΔxylFGHΔxylE) exhibited an impaired growth on xylose due to lacking of xylose transporters. After performing metabolic evolution, the evolved KJ12201-14T strain exhibited a great improvement in succinate production from pure xylose with higher concentration and productivity about 18 and 21%, respectively, compared to KJ12201 strain. During fed-batch fermentation, KJ12201-14T also produced succinate from xylose at a concentration, yield, and overall productivity of 84.6 ± 0.7 g/L, 0.86 ± 0.01 g/g and 1.01 ± 0.01 g/L/h, respectively. KJ12201 and KJ12201-14T strains co-utilized glucose/xylose mixture without catabolite repression. Both strains produced succinate from glucose/xylose mixture at concentration, yield, and overall and specific productivities of about 85 g/L, 0.85 g/g, 0.70 g/L/h, and 0.44 g/gCDW/h, respectively. Based on our results, KJ12201 and KJ12201-14T strains exhibited a greater performance in succinate production from xylose containing medium than those of other published works. They would be potential strains for the economic bio-based succinate production from xylose.

Genome analysis of food-processing stressful-resistant probiotic Bifidobacterium animalis subsp. lactis BF052, and its potential application in fermented soymilk.

In this study, Bifidobacterium animalis subsp. lactis BF052 was demonstrated the growth capability in soymilk and could be thus supplemented as a probiotic starter that employed soymilk as one of its food vehicles. The complete genome sequence of BF052 was therefore determined to understand the genetic basis of BF052 as a technological and functional probiotic starter. The whole genome sequence of BF052 consists of a circular genome of 1938 624 bp with a G+C content of 60.50%. This research highlights relevant genes involving in its adaptive responses to industrial and/or environmental stresses and utilization of α-galacto-oligosaccharides in BF052 strain compared with other representative bifidobacterial genomes.

Process Optimization on Micro-Aeration Supply for High Production Yield of 2,3-Butanediol from Maltodextrin by Metabolically-Engineered Klebsiella oxytoca.

An optimization process with a cheap and abundant substrate is considered one of the factors affecting the price of the production of economical 2,3-Butanediol (2,3-BD). A combination of the conventional method and response surface methodology (RSM) was applied in this study. The optimized levels of pH, aeration rate, agitation speed, and substrate concentration (maltodextrin) were investigated to determine the cost-effectiveness of fermentative 2,3-BD production by metabolically-engineered Klebsiella oxytoca KMS005. Results revealed that pH, aeration rate, agitation speed, and maltodextrin concentration at levels of 6.0, 0.8 vvm, 400 rpm, and 150 g/L respectively were the optimal conditions. RSM also indicated that the agitation speed was the most influential parameter when either agitation and aeration interaction or agitation and substrate concentration interaction played important roles for 2,3-BD production by the strain from maltodextrin. Under interim fed-batch fermentation, 2,3-BD concentration, yield, and productivity were obtained at 88.1±0.2 g/L, 0.412±0.001 g/g, and 1.13±0.01 g/L/h respectively within 78 h.

Effect of physicochemical pre-treatment of rice straw on its digestibility by Clostridium cellulolyticum.

Rice straw (RS) may serve as a low-cost biomass for the production of biofuels and biochemicals, but its native structure is resistant to enzymatic and microbial deconstruction. Therefore, an efficient pre-treatment method is required to modify crystalline cellulose to a more reactive amorphous form. This work investigated pre-treatments of rice straw involving size reduction (S) followed by either sodium hydroxide (NaOH) or diluted sulfuric acid (H2SO4) and liquid hot water (LHW). The shrinkage of the vascular bundles in the rice straw structure pre-treated with NaOH-LHW-S was higher than that with LHW-S and H2SO4-LHW-S pre-treatments. The highest levels of total fermentative products and residual sugars were obtained at the concentrations of 7.8 ± 0.2 and 2.1 ± 0.3 g/L, respectively, after fermentation by Clostridium cellulolyticum for NaOH-LHW-S pre-treated rice straw at 121 °C for 120 min. Overall, the combined physicochemical pre-treatment of RS led to improved microbial hydrolysis during cellulose degradation at the percentage of 85.5 ± 0.5.

Effects of the Food Manufacturing Chain on the Viability and Functionality of Bifidobacterium animalis through Simulated Gastrointestinal Conditions.

The viability and functionality of probiotics may be influenced by industrial production processes resulting in a decrease in probiotic efficiency that benefit the health of humans. This study aimed to investigate the probiotic characteristics of Bifidobacterium strains isolated from fecal samples of healthy Thai infants. In the present work, three local strains (BF014, BF052, and BH053) belonging to Bifidobacterium animalis showed a great resistance against conditions simulating the gastrointestinal tract. Among these, B. animalis BF052 possessed considerable probiotic properties, including high acid and bile tolerance, strong adhesion capability to Caco-2 cells, and inhibitory activity against pathogens including Salmonella typhimurium and Vibrio cholerae. This strain also exhibited a high survival rate compared to commercial strains during storage in a wide variety of products, including pasteurized milk, soy milk, drinking yogurt, and orange juice. The impact of food processing processes as well as the freeze-drying process, storage of freeze-dried powders, and incorporation of freeze-dried cells in food matrix on probiotic properties was also determined. The stability of the probiotic properties of the BF052 strain was not affected by food processing chain, especially its resistance in the simulated gastrointestinal conditions and its adherence ability to Caco-2 cells. It indicates that it satisfies the criteria as a potential probiotic and may be used as an effective probiotic starter in food applications.

Efficient reduction of the formation of by-products and improvement of production yield of 2,3-butanediol by a combined deletion of alcohol dehydrogenase, acetate kinase-phosphotransacetylase, and lactate dehydrogenase genes in metabolically engineered Klebsiella oxytoca in mineral salts medium.

Klebsiella oxytoca KMS005 (∆adhE∆ackA-pta∆ldhA) was metabolically engineered to improve 2,3-butanediol (BDO) yield. Elimination of alcohol dehydrogenase E (adhE), acetate kinase A-phosphotransacetylase (ackA-pta), and lactate dehydrogenase A (ldhA) enzymes allowed BDO production as a primary pathway for NADH re-oxidation, and significantly reduced by-products. KMS005 was screened for the efficient glucose utilization by metabolic evolution. KMS005-73T improved BDO production at a concentration of 23.5±0.5 g/L with yield of 0.46±0.02 g/g in mineral salts medium containing 50 g/L glucose in a shake flask. KMS005-73T also exhibited BDO yields of about 0.40-0.42 g/g from sugarcane molasses, cassava starch, and maltodextrin. During fed-batch fermentation, KMS005-73T produced BDO at a concentration, yield, and overall and specific productivities of 117.4±4.5 g/L, 0.49±0.02 g/g, 1.20±0.05 g/Lh, and 27.2±1.1 g/gCDW, respectively. No acetoin, lactate, and formate were detected, and only trace amounts of acetate and ethanol were formed. The strain also produced the least by-products and the highest BDO yield among other Klebsiella strains previously developed.

Efficient utilization of cassava pulp for succinate production by metabolically engineered Escherichia coli KJ122.

A metabolically engineered Escherichia coli KJ122 was efficiently utilized for succinate production from cassava pulp during batch separate hydrolysis and fermentation (SHF) under simple anaerobic conditions. Succinate concentration of 41.46 ± 0.05 g/L with yield and productivity of 82.33 ± 0.14 g/100 g dry pulp and 0.84 ± 0.02 g/L/h was obtained. In batch simultaneous saccharification and fermentation (SSF), hydrolysis of 12 % (w/v) cassava pulp with an enzyme loading of 2 % AMG + 3 % Cel (v/w) at pH 6.5 was optimized at 39 °C. Succinate concentration of 80.86 ± 0.49 g/L with a yield of 70.34 ± 0.37 g/100 g dry pulp and a productivity of 0.84 ± 0.01 g/L/h was attained using E. coli KJ122. Fed-batch SSF significantly enhanced succinate concentration to 98.63 ± 0.12 g/L at yield and productivity of 71.64 ± 0.97 g/100 g dry pulp and 1.03 ± 0.01 g/L/h. This result indicated an efficient and economical succinate production from cassava pulp using SHF and SSF by the use of E. coli KJ122.

Metabolic engineering of Klebsiella oxytoca M5a1 to produce optically pure D-lactate in mineral salts medium.

Klebsiella oxytoca strains were constructed to produce optical pure d-lactate by pH-controlled batch fermentation in mineral salts medium. The alcohol dehydrogenase gene, adhE, and the phospho-transacetylase/acetate kinase A genes, pta-ackA, were deleted from the wild type. KMS002 (ΔadhE) and KMS004 (ΔadhE Δpta-ackA) exhibited d-lactate production as a primary pathway for the regeneration of NAD(+). Both strains produced 11-13 g/L of d-lactate in medium containing 2% (w/v) glucose with yields of 0.64-0.71 g/g glucose used. In sugarcane molasses, KMS002 and KMS004 produced 22-24 g/L of d-lactate with yields of 0.80-0.87 g/g total sugars utilized. Both strains also utilized maltodextrin derived from cassava starch and produced d-lactate at a concentration of 33-34 g/L with yields of 0.91-0.92 g/g maltodextrin utilized. These d-lactate yields are higher than those reported for engineered E. coli strains.