Biologically informed deep neural network for prostate cancer discovery

Elmarakeby, Haitham; Hwang, Justin; Arafeh, Rand; Crowdis, Jett; Gang, Sydney; Liu, David; AlDubayan, Saud H.; Salari, Keyan; Kregel, Steven; Richter, Camden; Arnoff, Taylor E.; Park, Jihye; Hahn, William C.; Allen, Eliezer M. Van

doi:10.1038/s41586-021-03922-4

Cited by 220 publications

(208 citation statements)

References 49 publications

Supporting

Mentioning

175

Contrasting

Order By: Relevance

“…For example, in the area of lung related complications, deep neural networks have been explored to great effect for aiding clinicians in the detection of tuberculosis ( 8 , 9 ), pulmonary fibrosis ( 10 , 11 ), and lung cancer ( 12 – 14 ). Similar works have been done for prostate cancer ( 15 , 15 , 16 ) and breast cancer ( 17 , 18 ).…”

Section: Introductionsupporting

confidence: 55%

MEDUSA: Multi-Scale Encoder-Decoder Self-Attention Deep Neural Network Architecture for Medical Image Analysis

et al. 2022

View full text Add to dashboard Cite

Medical image analysis continues to hold interesting challenges given the subtle characteristics of certain diseases and the significant overlap in appearance between diseases. In this study, we explore the concept of self-attention for tackling such subtleties in and between diseases. To this end, we introduce, a multi-scale encoder-decoder self-attention (MEDUSA) mechanism tailored for medical image analysis. While self-attention deep convolutional neural network architectures in existing literature center around the notion of multiple isolated lightweight attention mechanisms with limited individual capacities being incorporated at different points in the network architecture, MEDUSA takes a significant departure from this notion by possessing a single, unified self-attention mechanism with significantly higher capacity with multiple attention heads feeding into different scales in the network architecture. To the best of the authors' knowledge, this is the first “single body, multi-scale heads” realization of self-attention and enables explicit global context among selective attention at different levels of representational abstractions while still enabling differing local attention context at individual levels of abstractions. With MEDUSA, we obtain state-of-the-art performance on multiple challenging medical image analysis benchmarks including COVIDx, Radiological Society of North America (RSNA) RICORD, and RSNA Pneumonia Challenge when compared to previous work. Our MEDUSA model is publicly available.

show abstract

Section: Introductionsupporting

confidence: 55%

MEDUSA: Multi-Scale Encoder-Decoder Self-Attention Deep Neural Network Architecture for Medical Image Analysis

et al. 2022

View full text Add to dashboard Cite

show abstract

“…The result indicates that the ensemble performance of gene-mWCA SVM (EGmWS) was regarded as effective methodology compared to related methodologies in terms of accuracy and solving the uncertainty problems. Elmarakeby et al [ 15 ] designed a P-NET—a biologically informed DL method—for stratifying patients with PCa by treatment resistance state and gauging molecular driver of treatment resistance to therapeutic target via method interpretability. They demonstrated that P-NET could forecast cancer state by utilizing molecular information with performances, i.e., better than other modeling techniques.…”

Section: Literature Reviewmentioning

confidence: 99%

Optimal Deep Learning Enabled Prostate Cancer Detection Using Microarray Gene Expression

Alshareef

Alsini

Alsieni

et al. 2022

Journal of Healthcare Engineering

View full text Add to dashboard Cite

Prostate cancer is the main cause of death over the globe. Earlier detection and classification of cancer is highly important to improve patient health. Previous studies utilized statistical and machine learning (ML) techniques for prostate cancer detection. However, several challenges that exist in the investigation process are the existence of high dimensionality data and less number of training samples. Metaheuristic algorithms can be used to resolve the curse of dimensionality and improve the detection rate of artificial intelligence (AI) techniques. With this motivation, this article develops an artificial intelligence based feature selection with deep learning model for prostate cancer detection (AIFSDL-PCD) using microarray gene expression data. The AIFSDL-PCD technique involves preprocessing to enhance the input data quality. In addition, a chaotic invasive weed optimization (CIWO) based feature selection (FS) technique for choosing an optimal subset of features shows the novelty of the work. Moreover, the deep neural network (DNN) model can be applied as a classification model to detect the existence of prostate cancer in the microarray gene expression data. Furthermore, the hyperparameters of the DNN model can be effectively adjusted by the use of RMSprop optimizer. The design of CIWO based FS technique helps for reducing the computational complexity and improve the classification accuracy. The experimental results highlighted the betterment of the AIFSDL-PCD approach on the other techniques with respect to distinct measures.

show abstract

“…MDM4 suppresses p53 transcriptional activity and facilitates p53 proteasomal degradation via binding MDM2's E3 ligase activity towards p53 [505] . Overexpression of MDM4 has recently been detected in PCa tissue and increases with disease progression [506][507][508] . Recent transcriptomic characterization of MEX isolated from cow, donkey, and goat milk identified MDM4 as a central node protein for all three species, indicating a conserved checkpoint with higher numbers of interconnections [509] .…”

Section: Microrna-155 and Microrna-223mentioning

confidence: 99%

The endocrine and epigenetic impact of persistent cow milk consumption on prostate carcinogenesis

Melnik¹,

John²,

Weiskirchen³

et al. 2022

jtgg

View full text Add to dashboard Cite

This review analyzes the potential impact of milk-induced signal transduction on the pathogenesis of prostate cancer (PCa). Articles in PubMed until November 2021 reporting on milk intake and PCa were reviewed. Epidemiological studies identified commercial cow milk consumption as a potential risk factor of PCa. The potential impact of cow milk consumption on the pathogenesis of PCa may already begin during fetal and pubertal prostate growth, critical windows with increased vulnerability. Milk is a promotor of growth and anabolism via activating insulin-like growth factor-1 (IGF-1)/phosphatidylinositol-3 kinase (PI3K)/AKT/mechanistic target of rapamycin complex 1 (mTORC1) signaling. Estrogens, major steroid hormone components of commercial milk of persistently pregnant dairy cows, activate IGF-1 and mTORC1. Milk-derived signaling synergizes with common driver mutations of the PI3K/AKT/mTORC1 signaling pathway that intersect with androgen receptor, MFG-E8, MAPK, RUNX2, MDM4, TP53, and WNT signaling, respectively. Potential exogenously induced drivers of PCa are milk-induced elevations of growth hormone, IGF-1, MFG-E8, estrogens, phytanic acid, and aflatoxins, as well as milk exosome-derived oncogenic microRNAs including miR-148a, miR-21, and miR-29b. Commercial cow milk intake, especially the consumption of pasteurized milk, which represents the closest replica of native milk, activates PI3K-AKT-mTORC1 signaling via cow milk’s endocrine and epigenetic modes of action. Vulnerable periods for adverse nutrigenomic impacts on prostate health appear to be the fetal and pubertal growth periods, potentially priming the initiation of PCa. Cow milk-mediated overactivation of PI3K-AKT-mTORC1 signaling synergizes with the most common genetic deviations in PCa, promoting PCa initiation, progression, and early recurrence.

show abstract

Biologically informed deep neural network for prostate cancer discovery

Cited by 220 publications

References 49 publications

MEDUSA: Multi-Scale Encoder-Decoder Self-Attention Deep Neural Network Architecture for Medical Image Analysis

MEDUSA: Multi-Scale Encoder-Decoder Self-Attention Deep Neural Network Architecture for Medical Image Analysis

Optimal Deep Learning Enabled Prostate Cancer Detection Using Microarray Gene Expression

The endocrine and epigenetic impact of persistent cow milk consumption on prostate carcinogenesis

Contact Info

Product

Resources

About