James E. Davis

James E. Davis

Auto-generated profile, based on publicly available information

James E. Davis

Department of Pathology, SUNY Stony Brook University Hospital, Stony Brook, NY, USA

Research Focus & Metrics

Academic Career Length

36Years (1991-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 June 2016
Verified
Conference Paper
Closed access
LHLe Hou
Le Hou
DSDimitris Samaras
Dimitris Samaras
TKTahsin M. Kurc
Tahsin M. Kurc
YGYi Gao
Yi Gao
JDJames E. Davis
James E. Davis
JSJoel H. Saltz
Joel H. Saltz
Convolutional Neural Networks (CNN) are state-of-theart models for many image classification tasks. However, to recognize cancer subtypes automatically, training a CNN on gigapixel resolution Whole Slide Tissue Images (WSI) is currently computationally impossible. The differentiation of cancer subtypes is based on cellular-level visual features observed on image patch scale. Therefore, we argue that in this situation, training a patch-level classifier on image patches will perform better than or similar to an image-level classifier. The challenge becomes how to intelligently combine patch-level classification results and model the fact that not all patches will be discriminative. We propose to train a decision fusion model to aggregate patch-level predictions given by patch-level CNNs, which to the best of our knowledge has not been shown before. Furthermore, we formulate a novel Expectation-Maximization (EM) based method that automatically locates discriminative patches robustly by utilizing the spatial relationships of patches. We apply our method to the classification of glioma and non-small-cell lung carcinoma cases into subtypes. The classification accuracy of our method is similar to the inter-observer agreement between pathologists. Although it is impossible to train CNNs on WSIs, we experimentally demonstrate using a comparable non-cancer dataset of smaller images that a patch-based CNN can outperform an image-based CNN.

12 July 2018 in Annals of Surgical Oncology
Verified
Journal Article
Closed access
PDF file
MGMei Gao
Mei Gao
MLMiranda Lin
Miranda Lin
MRManisha Rao
Manisha Rao
HTHannah Thompson
Hannah Thompson
KHKelsi Hirai
Kelsi Hirai
JKJoseph Kim
Joseph Kim

1 April 2015 in The American Journal of Pathology
Verified
Journal Article
Closed access
VKVarvara Kirillov
Varvara Kirillov
JSJonathan T. Siler
Jonathan T. Siler
MRMahalakshmi Ramadass
Mahalakshmi Ramadass
LGLingyin Ge
Lingyin Ge
JDJames Davis
James Davis
GTGlenda Trujillo
Glenda Trujillo

1 January 2016 in Applied Immunohistochemistry & Molecular Morphology
Verified
Journal Article
Closed access
JDJames E. Davis
James E. Davis
BNBarbara Nemesure
Barbara Nemesure
SMSaira Mehmood
Saira Mehmood
VNVipul Nayi
Vipul Nayi
SBStephanie Bs Burke
Stephanie Bs Burke
MSMeenakshi Singh
Meenakshi Singh
Background: A subset of patients with ductal carcinoma in situ (DCIS) experience recurrence or progression to invasive cancer. Current clinical practice is not reliably guided by DCIS recurrence prediction, although recurrence risk for invasive breast cancer can now be assessed. We analyzed a panel of biomarkers (estrogen receptor, Her2, Ki67, p53, cyclin D1, COX-2, caveolin-1, survivin, and PPAR-γ) and DCIS histologic and clinical features to determine associations with DCIS recurrence. Materials and Methods: Seventy DCIS cases diagnosed between 1995 and 2010 were divided into 2 groups: 52 had DCIS without known recurrence after excision and 18 had DCIS with subsequent recurrence after excision as DCIS or invasive carcinoma in the ipsilateral or contralateral breast. Tissue microarrays were prepared, immunohistochemistry performed, and expression of the biomarkers scored semiquantitatively. Variables analyzed included age, tumor size, margin status, DCIS grade, necrosis, histologic type, and immunohistochemistry scores. Differences between groups were evaluated using t tests for continuous variables and Fisher exact tests for categorical variables. Results: Intraductal necrosis was associated with increased recurrence risk: 46% of nonrecurrent cases showed necrosis compared with 83% of those who recurred (P=0.007). Her2 (human epidermal growth factor receptor 2) and Ki67 expression distributions were significantly different between nonrecurrent and recurrent cases. Her2 was overexpressed in 14% of nonrecurrent cases compared with 50% in the recurrent cases (P=0.03). A total of 87% of nonrecurrent cases had low Ki67 staining (0% to 10%) compared with 50% among the recurrent cases (P=0.002). Conclusion: Our results suggest that Her2 and Ki67 immunohistochemistry and the presence of intraductal necrosis aid in DCIS risk stratification.

19 June 2018 in Journal of Biomedical Optics
Verified
Journal Article
Open access
PDF file
DBDavid R Busch
David R Busch
JDJames Davis
James Davis
AKAngela Kogler
Angela Kogler
RGRobert M. Galler
Robert M. Galler
APAshwin B. Parthasarathy
Ashwin B. Parthasarathy
TFThomas F. Floyd
Thomas F. Floyd
Spinal cord injuries can occur as a result of disease (e.g., tumors), trauma (e.g., accidents), or complications from surgeries, such as scoliosis correction,1 spinal cord tumor resection,2 and aortic aneurysm repair.3,4 The prevention and treatment of spinal cord injury are focused upon the maintenance and/or restoration of spinal cord blood flow, yet no technology exists to monitor spinal cord ischemia. This is a critical gap in the field of neurocritical care monitoring. Recently, we developed and demonstrated continuous monitoring for spinal cord ischemia using a thin, flexible fiber-optic probe employing near-infrared (NIR) technology.5,6 Furthermore, testing is needed prior to clinical adoption of this concept, including a clear demonstration that deployment of the required laser light in the proximity of the spinal cord is safe. This type of safety study is especially important because unlike many tissues in the body, even minimal damage to spinal tissues, as might arise in a “burn,” can lead to paralysis or paraparesis.

1 December 2017 in Pediatric Critical Care Medicine
Verified
Journal Article
Closed access
JHJody N. Huber
Jody N. Huber
BHBrieanna M. Bs Hilkin
Brieanna M. Bs Hilkin
JHJessica S. Hook
Jessica S. Hook
PBPatrick D. Brophy
Patrick D. Brophy
TDTina L. Arnp Davenport
Tina L. Arnp Davenport
JMJessica Moreland
Jessica Moreland
Objectives: Infants with congenital heart disease frequently require cardiopulmonary bypass, which causes systemic inflammation. The goal of this study was to determine if neutrophil phenotype and activation status predicts the development of inflammatory complications following cardiopulmonary bypass. Design: Prospective cohort study. Setting: Tertiary care PICU with postoperative cardiac care. Patients: Thirty-seven patients 5 days to 10 months old with congenital heart disease requiring cardiopulmonary bypass. Interventions: None. Measurements and Main Results: Laboratory and clinical data collected included length of mechanical ventilation, acute kidney injury, and fluid overload. Neutrophils were isolated from whole blood at three time points surrounding cardiopulmonary bypass. Functional analyses included measurement of cell surface protein expression and nicotinamide adenine dinucleotide phosphate oxidase activity. Of all patients studied, 40.5% displayed priming of nicotinamide adenine dinucleotide phosphate oxidase activity in response to N-formyl-Met-Leu-Phe stimulation 24 hours post cardiopulmonary bypass as compared to pre bypass. Neonates who received steroids prior to bypass demonstrated enhanced priming of nicotinamide adenine dinucleotide phosphate oxidase activity at 48 hours. Patients who displayed priming post cardiopulmonary bypass were 8.8 times more likely to develop severe acute kidney injury as compared to nonprimers. Up-regulation of neutrophil surface CD11b levels pre- to postbypass occurred in 51.4% of patients, but this measure of neutrophil priming was not associated with acute kidney injury. Subsequent analyses of the basal neutrophil phenotype revealed that those with higher basal CD11b expression were significantly less likely to develop acute kidney injury. Conclusions: Neutrophil priming occurs in a subset of infants undergoing cardiopulmonary bypass. Acute kidney injury was more frequent in those patients who displayed priming of nicotinamide adenine dinucleotide phosphate oxidase activity after cardiopulmonary bypass. This pilot study suggests that neutrophil phenotypic signature could be used to predict inflammatory organ dysfunction.