Machine Learning in Differentiating Gliomas from Primary CNS Lymphomas: A Systematic Review, Reporting Quality, and Risk of Bias Assessment.

BACKGROUND: Differentiating gliomas and primary CNS lymphoma represents a diagnostic challenge with important therapeutic ramifications. Biopsy is the preferred method of diagnosis, while MR imaging in conjunction with machine learning has shown promising results in differentiating these tumors. PURPOSE: Our aim was to evaluate the quality of reporting and risk of bias, assess data bases with which the machine learning classification algorithms were developed, the algorithms themselves, and their performance. DATA SOURCES: Ovid EMBASE, Ovid MEDLINE, Cochrane Central Register of Controlled Trials, and the Web of Science Core Collection were searched according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. STUDY SELECTION: From 11,727 studies, 23 peer-reviewed studies used machine learning to differentiate primary CNS lymphoma from gliomas in 2276 patients. DATA ANALYSIS: Characteristics of data sets and machine learning algorithms were extracted. A meta-analysis on a subset of studies was performed. Reporting quality and risk of bias were assessed using the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) and Prediction Model Study Risk Of Bias Assessment Tool. DATA SYNTHESIS: The highest area under the receiver operating characteristic curve (0.961) and accuracy (91.2%) in external validation were achieved by logistic regression and support vector machines models using conventional radiomic features. Meta-analysis of machine learning classifiers using these features yielded a mean area under the receiver operating characteristic curve of 0.944 (95% CI, 0.898-0.99). The median TRIPOD score was 51.7%. The risk of bias was high for 16 studies. LIMITATIONS: Exclusion of abstracts decreased the sensitivity in evaluating all published studies. Meta-analysis had high heterogeneity. CONCLUSIONS: Machine learning-based methods of differentiating primary CNS lymphoma from gliomas have shown great potential, but most studies lack large, balanced data sets and external validation. Assessment of the studies identified multiple deficiencies in reporting quality and risk of bias. These factors reduce the generalizability and reproducibility of the findings.

Digestibility of amino acids in corn, corn coproducts, and bakery meal fed to growing pigs.

The objectives of this experiment were to determine the apparent ileal digestibility and the standardized ileal digestibility (SID) of CP and AA in bakery meal, corn gluten meal, corn gluten feed, corn germ meal, and hominy feed and to compare these values with the apparent ileal digestibility and SID of CP and AA in corn and distillers dried grains with solubles (DDGS). Eight growing barrows (initial BW: 82.5 ± 5.5 kg) were randomly allotted to an 8 × 8 Latin square design with 8 diets and 8 periods. Diets contained corn, DDGS, bakery meal, corn gluten meal, corn gluten feed, corn germ meal, or hominy feed as the sole source of protein and AA. An N-free diet was used to measure basal endogenous losses of AA and protein. Pigs were fed experimental diets for eight 7-d periods, with ileal digesta being collected on d 6 and 7 of each period. Results indicated that the SID of Lys in corn gluten meal (78.7%) was greater (P < 0.01) than in DDGS, bakery meal, corn germ meal, and hominy feed (46.0, 48.4, 68.4, and 58.8%, respectively). The SID of all indispensable AA except Arg, Leu, and Met in bakery meal were not different from those in DDGS. The SID of Arg, His, Leu, and Met in corn gluten feed were less (P < 0.01) than in corn, but the SID of all other indispensable AA in corn gluten feed were not different from those in corn. However, for most indispensable AA, the SID in corn gluten feed was not different from the SID in DDGS. The SID of all indispensable AA in corn germ meal, except Arg, His, Leu, and Met, were not different from the SID in corn. Likewise, the SID of all indispensable AA in corn germ meal, except Arg and Leu, were not different from those in DDGS. The SID of Ile, Met, Phe, and Val in hominy feed were less (P < 0.01) than in corn, but the SID of the remaining indispensable AA in hominy feed were not different from the SID of indispensable AA in corn. All indispensable AA in hominy feed also had SID values that were not different from the SID values of AA in DDGS, except for Arg and Lys, which had greater (P < 0.01) SID than in DDGS. In conclusion, bakery meal had SID values of most AA that were less than in corn, but corn gluten meal had SID values for most AA that were greater than the SID of AA in corn, bakery meal, and corn coproducts.

Relative bioavailability of phosphorus in inorganic phosphorus sources fed to growing pigs.

The relative bioavailability of P in 5 sources of inorganic P was determined using growing pigs. The 5 sources of inorganic P were dicalcium phosphate (DCP), monocalcium phosphate (MCP) containing 50% MCP (MCP50), MCP containing 70% MCP (MCP70), MCP containing 100% MCP (MCP100), and monosodium phosphate (MSP). A total of 11 diets were formulated. The basal diet was formulated to contain 0.10% P, and 10 additional diets were formulated by adding 0.07 or 0.14% P from each of the 5 P sources to the basal diet. Growing pigs (n = 44; initial BW: 16.8 ± 4.3 kg) were individually housed and randomly allotted to the 11 experimental diets. Feed was provided on an ad libitum basis throughout the 28-d experimental period. At the conclusion of the experiment, all pigs were killed, and 4 bones (i.e., the third and fourth metacarpals on both front feet) were harvested. Bone-breaking strength, bone ash, and Ca and P concentrations were determined. The concentration of bone ash increased (P < 0.05) as MCP50, MCP70, MCP100, or MSP were added to the basal diet, and the concentration of bone P also increased (P < 0.05) as MCP70, MCP100, or MSP were added to the basal diet. The relative bioavailability of P in each of the feed phosphates was determined using slope ratio methodologies based on breaking strength, and expressed relative to MSP. The slope of the regression line for diets containing MSP or MCP100 was steeper (P < 0.05) than the slope for pigs fed the diet containing DCP, but not different (P > 0.05) from that of pigs fed diets supplemented with MCP50 or MCP70. In conclusion, P in MSP and MCP100 is more bioavailable than P in DCP, but there were no differences within MCP sources.

Amino acid digestibility and energy concentration in a novel source of high-protein distillers dried grains and their effects on growth performance of pigs.

Three experiments were conducted to evaluate a novel source of high-protein distillers dried grains produced by Buhler Inc. (HP-DDG(Buhler)) and fed to growing pigs. The standardized ileal digestibility (SID) of AA in HP-DDG(Buhler) and in soybean meal (SBM) was determined in Exp. 1. Nine pigs (109.8 +/- 2.78 kg of BW) were fitted with a T-cannula in the distal ileum and allotted to a triplicated 3 x 3 Latin square design with 3 diets and 3 periods per square. Diets containing HP-DDG(Buhler) or SBM as the only source of AA and an N-free diet were formulated. The SID of indispensable AA was less (P < 0.01) in HP-DDG(Buhler) than in SBM (Arg, 87.5 vs. 93.9%; His, 76.7 vs. 88.7%; Ile, 76.4 vs. 87.5%; Leu, 77.8 vs. 86.8%; Lys, 75.4 vs. 88.4%; Met, 82.8 vs. 88.4%; Phe, 77.9 vs. 87.3%; Thr, 72.5 vs. 83.5%; Trp, 85.1 vs. 91.0%; Val, 73.3 vs. 84.3%). The DE and ME in HP-DDG(Buhler) and in corn were measured in Exp. 2 using 16 growing barrows (24.6 +/- 1.66 kg of BW). A corn-based diet and a diet containing 50% corn and 48.2% HP-DDG(Buhler) were formulated. The total collection method and the difference procedures were used. The concentrations of DE and ME in HP-DDG(Buhler) were greater (P < 0.001) than in corn (5,043 vs. 4,002 kcal/kg of DM and 4,690 vs. 3,921 kcal/kg of DM, respectively). Experiment 3 was a 9-wk growth assay using 40 pigs (initial BW: 58.2 +/- 2.28 kg) allotted to 5 dietary treatments, with 8 replicates of individually housed pigs per treatment. Treatments included a control diet based on corn and SBM and 4 diets in which HP-DDG(Buhler) replaced 33, 66, 66, or 100% of the SBM in the control diet. All HP-DDG(Buhler) diets contained supplemental Lys and Thr to provide similar concentrations of SID Lys and Thr in all diets; one of the diets in which HP-DDG(Buhler) replaced 66% of the SBM and the diet in which HP-DDG(Buhler) replaced 100% of the SBM also contained crystalline Trp. Dietary treatments had no effect on ADG (1.15, 1.13, 1.16, 1.12, and 1.14 kg), ADFI (3.33, 3.35, 3.39, 3.30, and 3.33 kg), or G:F (0.35, 0.34, 0.34, 0.34, and 0.34 kg/kg). Carcass traits of pigs fed the diet in which HP-DDG(Buhler) replaced 100% of the SBM were not different from those of pigs fed the control diet. In conclusion, HP-DDG(Buhler) contains more DE and ME than corn, but has decreased SID values for AA compared with SBM. Soybean meal can be replaced by HP-DDG(Buhler) in diets fed to finishing pigs without any effect on growth performance or carcass characteristics, provided that diets are adequate in indispensable AA.

Differential gene expression and biomarkers in zebrafish (Danio rerio) following exposure to produced water components.

The main effluent from oil and gas production is produced water (PW), a waste that contains low to moderate concentrations of oil-derived substances such as polycyclic aromatic hydrocarbons (PAHs) and alkylphenols (APs). PW components may be present in seawater at low concentrations over large areas in the vicinity of oil and gas production facilities. In this study, zebrafish (Danio rerio) were exposed to control and three treatments (high-, pulsed-, low-dose) of a synthetic PW mixture for 1, 7 and 13 weeks. The aim was to investigate the development of transcriptome and biomarker responses as well as relationships between early responses and population-relevant effects. The synthetic PW contained a mixture of low-molecular-weight PAHs (<5 ring) and short-chain APs (C1-C4). The water-borne exposure levels (sum PAH) ranged from 0.54 ppb (low dose) to 5.4 ppb (high dose). Bile pyrene metabolites ranged from 17-133 ng g(-1) bile in the control group to 23-1081 ng g(-1) bile in the high exposure group. Similar levels have been observed in wild fish, confirming an environmentally relevant exposure. The expression of mRNAs of hepatic genes was investigated in the high exposure group using the Zebrafish OligoLibrary from Compugen. Functional clustering analysis revealed effects in the reproductive system, the nervous system, the respiratory system, the immune system, lipid metabolism, connective tissue and in a range of functional categories related to cell cycle and cancer. The majority of differentially expressed mRNAs of genes were down-regulated, suggesting reduction in gene transcription to be as relevant as up-regulation or induction when assessing biological responses to PW exposure. Biomarkers for effects of PAHs (cytochrome P450 1A) and environmental estrogens (vitellogenin) did not appear to be affected by the chronic exposure to low concentration of PW components. Effects at the population level included a reduction in condition factor in male fish from all exposed groups and spinal column deformations in the F1 generation of exposed groups. The different exposure regimes did not produce any significant differences in reproduction or recruitment. The results from this study demonstrate that environmentally relevant concentrations of PW affect gene expression and population-relevant endpoints in zebrafish, although links between the two were not obvious.

Novel procedure for estimating endogenous losses and measurement of apparent and true digestibility of phosphorus by growing pigs.

An experiment was conducted to evaluate a novel procedure for estimating endogenous losses of P and for measuring the apparent total tract digestibility (ATTD) and true total tract digestibility (TTTD) of P in 5 inorganic P sources fed to growing pigs. The P sources were dicalcium phosphate (DCP), monocalcium phosphate (MCP) with 50% purity (MCP50), MCP with 70% purity (MCP70), MCP with 100% purity (MCP100), and monosodium phosphate (MSP). A gelatin-based, P-free basal diet was formulated and used to estimate endogenous losses of P. Five P-containing diets were formulated by adding 0.20% total P from each of the inorganic P sources to the basal diet. A seventh diet was formulated by adding 0.16% P from MCP70 to the basal diet. All diets were fed to 7 growing pigs in a 7 x 7 Latin square design, and urine and feces were collected during 5 d of each period. The endogenous loss of P was estimated as 139 +/- 18 mg/kg of DMI. The ATTD of P in MSP was greater (P < 0.05) than in DCP, MCP50, and MCP70 (91.9 vs. 81.5, 82.6, and 81.7%, respectively). In MSP, the TTTD of P was 98.2%. This value was greater (P < 0.05) than the TTTD of P in DCP, MCP50, and MCP70 (88.4, 89.5, and 88.6%, respectively). The ATTD and the TTTD for MCP70 were similar in diets formulated to contain 0.16 and 0.20% total P. Results from the current experiment demonstrate that a P-free diet may be used to measure endogenous losses of P in pigs. By adding inorganic P sources to this diet, the ATTD of P can be directly measured and the TTTD of P may be calculated for each source of P.

Development of an ELISA for vitellogenin in whole body homogenate of zebrafish (Danio rerio).

The yolk protein, lipovitellin (Lv) was purified from ovaries of mature female zebrafish (Danio rerio) by gel filtration and anion exchange chromatography. Polyclonal antibodies against Lv were raised in rabbits. Anti-Lv IgG was purified by affinity chromatography. SDS-PAGE followed by Western blotting was performed to analyse the specificity of the antibody and the immunological similarities between Lv and vitellogenin (Vtg). Anti-Lv IgG was used to develop a direct non-competitive sandwich ELISA to measure Vtg concentrations of whole body homogenate (WBH) in zebrafish. The intra- and interassay variabilities were 5.8% and 10.4%, respectively. The sensitivity was 0.2 ng Vtg x ml(-1) and the practical detection limit was 40 ng Vtg x g(-1) fish (wet weight). Adult male zebrafish were exposed to a nominal water concentration of 10 ng x l(-1) of ethinylestradiol (EE2) in a semi-static exposure system for 7 days. Compared with the control group, exposure to 10 ng EE2 x l(-1) induced a 200-fold increase in Vtg levels.