Artificial Intelligence and Early Detection of Pancreatic Cancer: 2020 Summative Review.

Published on Mar 1, 2021in Pancreas2.92
· DOI :10.1097/MPA.0000000000001762
Barbara J. Kenner5
Estimated H-index: 5
Suresh T. Chari103
Estimated H-index: 103
(University of Texas MD Anderson Cancer Center)
+ 31 AuthorsBrian M. Wolpin64
Estimated H-index: 64
(Harvard University)
ABSTRACT Despite considerable research efforts, pancreatic cancer is associated with a dire prognosis and a 5-year survival rate of only 10%. Early symptoms of the disease are mostly nonspecific. The premise of improved survival through early detection is that more individuals will benefit from potentially curative treatment. Artificial intelligence (AI) methodology has emerged as a successful tool for risk stratification and identification in general health care. In response to the maturity of AI, Kenner Family Research Fund conducted the 2020 AI and Early Detection of Pancreatic Cancer Virtual Summit ( in conjunction with the American Pancreatic Association, with a focus on the potential of AI to advance early detection efforts in this disease. This comprehensive presummit article was prepared based on information provided by each of the interdisciplinary participants on one of the 5 following topics: Progress, Problems, and Prospects for Early Detection; AI and Machine Learning; AI and Pancreatic Cancer-Current Efforts; Collaborative Opportunities; and Moving Forward-Reflections from Government, Industry, and Advocacy. The outcome from the robust Summit conversations, to be presented in a future white paper, indicate that significant progress must be the result of strategic collaboration among investigators and institutions from multidisciplinary backgrounds, supported by committed funders.
#1Bo Yuan (Broad Institute)H-Index: 2
#2Ciyue Shen (Broad Institute)H-Index: 5
Last. Chris Sander (Broad Institute)H-Index: 184
view all 7 authors...
Summary Systematic perturbation of cells followed by comprehensive measurements of molecular and phenotypic responses provides informative data resources for constructing computational models of cell biology. Models that generalize well beyond training data can be used to identify combinatorial perturbations of potential therapeutic interest. Major challenges for machine learning on large biological datasets are to find global optima in a complex multidimensional space and mechanistically interp...
13 CitationsSource
#1Adam Yala (MIT: Massachusetts Institute of Technology)H-Index: 8
#2Peter G. Mikhael (MIT: Massachusetts Institute of Technology)H-Index: 4
Last. Regina Barzilay (MIT: Massachusetts Institute of Technology)H-Index: 73
view all 10 authors...
Improved breast cancer risk models enable targeted screening strategies that achieve earlier detection and less screening harm than existing guidelines. To bring deep learning risk models to clinical practice, we need to further refine their accuracy, validate them across diverse populations, and demonstrate their potential to improve clinical workflows. We developed Mirai, a mammography-based deep learning model designed to predict risk at multiple timepoints, leverage potentially missing risk ...
5 CitationsSource
#1Jaelim Cho (University of Auckland)H-Index: 3
#1Jaelim Cho (University of Auckland)H-Index: 15
Last. Maxim S. Petrov (University of Auckland)H-Index: 28
view all 4 authors...
It is well established that individuals with diabetes mellitus (DM) may develop exocrine pancreatic dysfunction (EPD) requiring pancreatic enzyme replacement therapy, whereas the converse relationship has been poorly studied. Pancreatitis is a disease that is well suited to investigate the latter as it is often characterized by the development of EPD and/or new-onset DM. The aim was to investigate the association between EPD and the risk of new-onset DM in individuals after the first attack of p...
8 CitationsSource
#1Noura S. Abul-HusnH-Index: 23
#2Emily R Soper (ISMMS: Icahn School of Medicine at Mount Sinai)H-Index: 4
Last. Eimear E. KennyH-Index: 37
view all 12 authors...
BACKGROUND: Pathogenic variants in BRCA1 and BRCA2 (BRCA1/2) lead to increased risk of breast, ovarian, and other cancers, but most variant-positive individuals in the general population are unaware of their risk, and little is known about prevalence in non-European populations. We investigated BRCA1/2 prevalence and impact in the electronic health record (EHR)-linked BioMe Biobank in New York City. METHODS: Exome sequence data from 30,223 adult BioMe participants were evaluated for pathogenic v...
24 CitationsSource
#1Amanda L. Blackford (Johns Hopkins University)H-Index: 56
#2Marcia I. Canto (Johns Hopkins University)H-Index: 73
Last. Michael Goggins (Johns Hopkins University)H-Index: 118
view all 5 authors...
BACKGROUND: Rapid access to pancreatic imaging and regular pancreatic surveillance may help identify Stage I pancreatic cancer. We investigated recent trends in the stage of newly-diagnosed PDACs, age at diagnosis, and survival. METHODS: Trends in age-adjusted incidence of Stage IA pancreatic ductal adenocarcinoma (PDAC) between 2004 and 2016 were determined from the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) database. All tests were two-sided. RESULTS: The inc...
21 CitationsSource
#1Gina GheorgheH-Index: 3
#2Simona BungauH-Index: 17
Last. Camelia Cristina Diaconu (UMFCD: Carol Davila University of Medicine and Pharmacy)H-Index: 12
view all 10 authors...
Pancreatic cancer (PC) is one of the most aggressive forms of cancer. Negative prognosis is mainly due to the late diagnosis in advanced stages, when the disease is already therapeutically overcome. Studies in recent years have focused on identifying biomarkers that could play a role in early diagnosis, leading to the improvement of morbidity and mortality. Currently, the only biomarker widely used in the diagnosis of PC is carbohydrate antigen 19-9 (CA19.9), which has, however, more of a progno...
5 CitationsSource
#1Enrico Santus (MIT: Massachusetts Institute of Technology)H-Index: 16
#2Tal Schuster (MIT: Massachusetts Institute of Technology)H-Index: 9
Last. Kevin S. Hughes (Harvard University)H-Index: 52
view all 12 authors...
PURPOSELiterature on clinical note mining has highlighted the superiority of machine learning (ML) over hand-crafted rules. Nevertheless, most studies assume the availability of large training sets...
2 CitationsSource
#1Troels Siggaard (UCPH: University of Copenhagen)H-Index: 2
#2Roc Reguant (UCPH: University of Copenhagen)H-Index: 1
Last. Søren Brunak (UCPH: University of Copenhagen)H-Index: 107
view all 10 authors...
We present the Danish Disease Trajectory Browser (DTB), a tool for exploring almost 25 years of data from the Danish National Patient Register. In the dataset comprising 7.2 million patients and 122 million admissions, users can identify diagnosis pairs with statistically significant directionality and combine them to linear disease trajectories. Users can search for one or more disease codes (ICD-10 classification) and explore disease progression patterns via an array of functionalities. For ex...
7 CitationsSource
#1Talal Seoud (Stony Brook University Hospital)H-Index: 1
#2Aslam Syed (Stony Brook University Hospital)
Last. Shyam Thakkar (CMU: Carnegie Mellon University)H-Index: 9
view all 9 authors...
OBJECTIVES The aim of this study was to evaluate depression in pancreatic cancer (PC) patients before and after a cancer diagnosis using a US-based healthcare database. We also sought to study the impact of treatment of depression in PC patients on all-cause mortality. METHODS Pancreatic cancer patients with comorbid depression in Explorys (1999-2019) were compared with controls using odds ratios with 95% confidence intervals. Rates of depression diagnosed within 6 months, 1 year, and 3 years be...
2 CitationsSource
#1Ayuko Hoshino (TITech: Tokyo Institute of Technology)H-Index: 14
#2Han Sang Kim (Yonsei University)H-Index: 16
Last. David Lyden (Cornell University)H-Index: 64
view all 117 authors...
Summary There is an unmet clinical need for improved tissue and liquid biopsy tools for cancer detection. We investigated the proteomic profile of extracellular vesicles and particles (EVPs) in 426 human samples from tissue explants (TEs), plasma, and other bodily fluids. Among traditional exosome markers, CD9, HSPA8, ALIX, and HSP90AB1 represent pan-EVP markers, while ACTB, MSN, and RAP1B are novel pan-EVP markers. To confirm that EVPs are ideal diagnostic tools, we analyzed proteomes of TE- (n...
120 CitationsSource
Cited By0