Publications
Department of Medicine faculty members published more than 3,000 peer-reviewed articles in 2022.
2015
BACKGROUND AND STUDY AIMS
The real-time optical diagnosis of colorectal polyps with high confidence predictions can achieve high levels of accuracy. Increasing the rates of high confidence optical diagnosis can improve the clinical application of real-time optical diagnosis in routine practice. The primary aim of this prospective study was to evaluate whether high magnifying endoscopy improves the rates of high confidence narrow-band imaging (NBI) - based optical diagnosis for differentiating between neoplastic and non-neoplastic colorectal lesions according to the NBI international colorectal endoscopic (NICE) classification.
PATIENTS AND METHODS
Consecutive adult patients undergoing colonoscopy with a high magnifying (maximum, × 80) colonoscope between April and August 2012 were recruited. The optical diagnosis for each polyp was evaluated during colonoscopy in two consecutive stages by the same endoscopist, who first used NBI with non-magnifying endoscopy (NBI-NME), then NBI with magnifying endoscopy (NBI-ME). A level of confidence was assigned to each prediction.
RESULTS
The analysis included 124 patients (mean age, 56.4 years; male-to-female ratio, 72:52) with 248 polyps smaller than 10 mm. Of the 248 polyps, 210 were 1 to 5 mm in size and 38 were 6 to 9 mm in size; 77 polyps were hyperplastic, 4 were sessile serrated adenomas/polyps, 160 were low grade adenomas, 5 were high grade adenomas, and 2 were deep submucosal invasive carcinomas. The rate of high confidence optical diagnosis when NBI-ME was used was significantly higher than the rate when NBI-NME was used for diminutive (1 - 5 mm) polyps (92.9 % vs 79.5 %, P < 0.001) and for small (6 - 9 mm) polyps (94.7 % vs 84.2 %, P = 0.048).
CONCLUSION
High magnifying endoscopy significantly improved the rates of high confidence NBI-based optical diagnosis of diminutive and small colorectal polyps.
STUDY REGISTRATION
UMIN 000007608.
View on PubMed2015
BACKGROUND AND SIGNIFICANCE
Sparsity is often a desirable property of statistical models, and various feature selection methods exist so as to yield sparser and interpretable models. However, their application to biomedical text classification, particularly to mortality risk stratification among intensive care unit (ICU) patients, has not been thoroughly studied.
OBJECTIVE
To develop and characterize sparse classifiers based on the free text of nursing notes in order to predict ICU mortality risk and to discover text features most strongly associated with mortality.
METHODS
We selected nursing notes from the first 24h of ICU admission for 25,826 adult ICU patients from the MIMIC-II database. We then developed a pair of stochastic gradient descent-based classifiers with elastic-net regularization. We also studied the performance-sparsity tradeoffs of both classifiers as their regularization parameters were varied.
RESULTS
The best-performing classifier achieved a 10-fold cross-validated AUC of 0.897 under the log loss function and full L2 regularization, while full L1 regularization used just 0.00025% of candidate input features and resulted in an AUC of 0.889. Using the log loss (range of AUCs 0.889-0.897) yielded better performance compared to the hinge loss (0.850-0.876), but the latter yielded even sparser models.
DISCUSSION
Most features selected by both classifiers appear clinically relevant and correspond to predictors already present in existing ICU mortality models. The sparser classifiers were also able to discover a number of informative - albeit nonclinical - features.
CONCLUSION
The elastic-net-regularized classifiers perform reasonably well and are capable of reducing the number of features required by over a thousandfold, with only a modest impact on performance.
View on PubMed2015
UNLABELLED
Staphylococcus aureus is a Gram-positive, commensal bacterium known to asymptomatically colonize the human skin, nares, and gastrointestinal tract. Colonized individuals are at increased risk for developing S. aureus infections, which range from mild skin and soft tissue infections to more severe diseases, such as endocarditis, bacteremia, sepsis, and osteomyelitis. Different virulence factors are required for S. aureus to infect different body sites. In this study, virulence gene expression was analyzed in two S. aureus isolates during nasal colonization, bacteremia and in the heart during sepsis. These models were chosen to represent the stepwise progression of S. aureus from an asymptomatic colonizer to an invasive pathogen. Expression of 23 putative S. aureus virulence determinants, representing protein and carbohydrate adhesins, secreted toxins, and proteins involved in metal cation acquisition and immune evasion were analyzed. Consistent upregulation of sdrC, fnbA, fhuD, sstD, and hla was observed in the shift between colonization and invasive pathogen, suggesting a prominent role for these genes in staphylococcal pathogenesis. Finally, gene expression data were correlated to the roles of the genes in pathogenesis by using knockout mutants in the animal models. These results provide insights into how S. aureus modifies virulence gene expression between commensal and invasive pathogens.
IMPORTANCE
Many bacteria, such as Staphylococcus aureus, asymptomatically colonize human skin and nasal passages but can also cause invasive diseases, such as bacteremia, pneumonia, sepsis, and osteomyelitis. The goal of this study was to analyze differences in the expression of selected S. aureus genes during a commensal lifestyle and as an invasive pathogen to gain insight into the commensal-to-pathogen transition and how a bacterial pathogen adapts to different environments within a host (e.g., from nasal colonization to invasive pathogen). The gene expression data were also used to select genes for which to construct knockout mutants to assess the role of several proteins in nasal colonization and lethal bacteremia. These results not only provide insight into the factors involved in S. aureus disease pathogenesis but also provide potential therapeutic targets.
View on PubMed2015
2015
2015
2015
BACKGROUND
Busulfan and etoposide have been used as myeloablative therapy for autologous hematopoietic stem cell transplantation (HSCT) in adults with acute myeloid leukemia (AML) for > 20 years. The use of targeted intravenous (I.V.) busulfan has significantly improved the tolerability and efficacy of this regimen. We designed a dose-escalation study to examine the maximum tolerated dose (MTD) of targeted I.V. busulfan with bolus etoposide as preparative therapy for autologous HSCT in AML.
PATIENTS AND METHODS
In this single-center, phase I study, adult AML patients received I.V. busulfan targeted to either an area under the curve (AUC) of 1250 (cohort 1) or 1400 (cohort 2) μmol/min over 16 doses. Dose adjustments based on plasma pharmacokinetics occurred before doses 2 and 11. Etoposide 60 mg/kg I.V. was administered 24 hours after the last busulfan dose and 3 days before stem cell infusion.
RESULTS
Twelve patients with intermediate-risk AML in first complete remission were treated. All patients in cohort 1 and 5 patients (83%) in cohort 2 were within 10% of the target AUC. The MTD was not reached, although Grade ≥ 3 mucositis occurred in 3 patients (50%) in cohort 1 and in 4 patients (66%) in cohort 2, limiting further dose escalation. Two-year relapse-free survival was 33% in cohort 1 versus 67% in cohort 2 (P = .08).
CONCLUSION
Etoposide and targeted, dose-escalated I.V. busulfan as myeloablative therapy for autologous HSCT in AML is safe, with mucositis being the most significant toxicity. A phase II study is warranted to further evaluate the activity and safety of busulfan targeted to AUC 1400 μmol/min.
View on PubMed