Artificial intelligence in radiology

Nature reviews. Cancer, Journal Year: 2018, Volume and Issue: 18(8), P. 500 - 510

Published: May 17, 2018

Language: Английский

---

Artificial intelligence in cancer imaging: Clinical challenges and applications

CA A Cancer Journal for Clinicians, Journal Year: 2019, Volume and Issue: 69(2), P. 127 - 157

Published: Feb. 5, 2019

Abstract Judgement, as one of the core tenets medicine, relies upon integration multilayered data with nuanced decision making. Cancer offers a unique context for medical decisions given not only its variegated forms evolution disease but also need to take into account individual condition patients, their ability receive treatment, and responses treatment. Challenges remain in accurate detection, characterization, monitoring cancers despite improved technologies. Radiographic assessment most commonly visual evaluations, interpretations which may be augmented by advanced computational analyses. In particular, artificial intelligence (AI) promises make great strides qualitative interpretation cancer imaging expert clinicians, including volumetric delineation tumors over time, extrapolation tumor genotype biological course from radiographic phenotype, prediction clinical outcome, impact treatment on adjacent organs. AI automate processes initial images shift workflow management whether or administer an intervention, subsequent observation yet envisioned paradigm. Here, authors review current state applied describe advances 4 types (lung, brain, breast, prostate) illustrate how common problems are being addressed. Although studies evaluating applications oncology date have been vigorously validated reproducibility generalizability, results do highlight increasingly concerted efforts pushing technology use future directions care.

Language: Английский

---

A radiomics approach to assess tumour-infiltrating CD8 cells and response to anti-PD-1 or anti-PD-L1 immunotherapy: an imaging biomarker, retrospective multicohort study

The Lancet Oncology, Journal Year: 2018, Volume and Issue: 19(9), P. 1180 - 1191

Published: Aug. 14, 2018

Language: Английский

---

Deep learning for lung cancer prognostication: A retrospective multi-cohort radiomics study

PLoS Medicine, Journal Year: 2018, Volume and Issue: 15(11), P. e1002711 - e1002711

Published: Nov. 30, 2018

Non-small-cell lung cancer (NSCLC) patients often demonstrate varying clinical courses and outcomes, even within the same tumor stage. This study explores deep learning applications in medical imaging allowing for automated quantification of radiographic characteristics potentially improving patient stratification.We performed an integrative analysis on 7 independent datasets across 5 institutions totaling 1,194 NSCLC (age median = 68.3 years [range 32.5-93.3], survival 1.7 0.0-11.7]). Using external validation computed tomography (CT) data, we identified prognostic signatures using a 3D convolutional neural network (CNN) treated with radiotherapy (n 771, age 68.0 1.3 We then employed transfer approach to achieve surgery 391, 69.1 37.2-88.0], 3.1 0.0-8.8]). found that CNN predictions were significantly associated 2-year overall from start respective treatment (area under receiver operating characteristic curve [AUC] 0.70 [95% CI 0.63-0.78], p < 0.001) (AUC 0.71 0.60-0.82], patients. The was also able stratify into low high mortality risk groups both (p 0.03) datasets. Additionally, outperform random forest models built parameters-including age, sex, node metastasis stage-as well as robustness against test-retest (intraclass correlation coefficient 0.91) inter-reader (Spearman's rank-order 0.88) variations. To gain better understanding captured by CNN, regions most contribution towards highlighted importance tumor-surrounding tissue stratification. present preliminary findings biological basis phenotypes being linked cell cycle transcriptional processes. Limitations include retrospective nature this opaque black box networks.Our results provide evidence networks may be used stratification based standard-of-care CT images motivates future research deciphering prospective data.

Language: Английский

---

AI applications to medical images: From machine learning to deep learning

Physica Medica, Journal Year: 2021, Volume and Issue: 83, P. 9 - 24

Published: March 1, 2021

Language: Английский

---

Predicting response to cancer immunotherapy using noninvasive radiomic biomarkers

Annals of Oncology, Journal Year: 2019, Volume and Issue: 30(6), P. 998 - 1004

Published: March 19, 2019

IntroductionImmunotherapy is regarded as one of the major breakthroughs in cancer treatment. Despite its success, only a subset patients responds—urging quest for predictive biomarkers. We hypothesize that artificial intelligence (AI) algorithms can automatically quantify radiographic characteristics are related to and may therefore act noninvasive radiomic biomarkers immunotherapy response.Patients methodsIn this study, we analyzed 1055 primary metastatic lesions from 203 with advanced melanoma non-small-cell lung (NSCLC) undergoing anti-PD1 therapy. carried out an AI-based characterization each lesion on pretreatment contrast-enhanced CT imaging data develop validate machine learning biomarker capable distinguishing between responding nonresponding. To define biological basis biomarker, gene set enrichment analysis independent dataset 262 NSCLC patients.ResultsThe reached significant performance (up 0.83 AUC, P < 0.001) borderline lymph nodes (0.64 = 0.05). Combining these lesion-wide predictions patient level, response could be predicted AUC up 0.76 both types (P< 0.001), resulting 1-year survival difference 24% (P 0.02). found highly associations pathways involved mitosis, indicating relationship increased proliferative potential preferential immunotherapy.ConclusionsThese results indicate standard-of-care function immunotherapy, show utility improved stratification neoadjuvant palliative settings.

Language: Английский

---

Biomarkers in Lung Cancer Screening: Achievements, Promises, and Challenges

Journal of Thoracic Oncology, Journal Year: 2018, Volume and Issue: 14(3), P. 343 - 357

Published: Dec. 4, 2018

Language: Английский

---

The Biological Meaning of Radiomic Features

Radiology, Journal Year: 2021, Volume and Issue: 298(3), P. 505 - 516

Published: Jan. 5, 2021

Radiomic analysis offers a powerful tool for the extraction of clinically relevant information from radiologic imaging. Radiomics can be used to predict patient outcome through automated high-throughput feature extraction, using large training cohorts elucidate subtle relationships between image characteristics and disease status. However powerful, data-driven nature radiomics inherently no insight into biological underpinnings observed relationships. Early work was dominated by semantic, radiologist-defined features carried qualitative real-world meaning. Following rapid developments popularity machine learning approaches, field moved quickly toward agnostic analyses, resulting in increasingly sets. This trend took focus an increase predictive power further away understanding findings. Such disconnect predictor model meaning will limit broad clinical translation. Efforts reintroduce are gaining traction with distinct emerging approaches available, including genomic correlates, local microscopic pathologic textures, macroscopic histopathologic marker expression. These methods presented this review, their significance is discussed. The authors that following increasing pressure robust radiomics, validation become standard practice field, thus cementing role method decision making. © RSNA, 2021 An earlier incorrect version appeared online. article corrected on February 10, 2021.

Language: Английский

---

On the Interpretability of Artificial Intelligence in Radiology: Challenges and Opportunities

Radiology Artificial Intelligence, Journal Year: 2020, Volume and Issue: 2(3), P. e190043 - e190043

Published: May 1, 2020

As artificial intelligence (AI) systems begin to make their way into clinical radiology practice, it is crucial assure that they function correctly and gain the trust of experts. Toward this goal, approaches AI "interpretable" have gained attention enhance understanding a machine learning algorithm, despite its complexity. This article aims provide insights current state art interpretability methods for AI. review discusses radiologists' opinions on topic suggests trends challenges need be addressed effectively streamline in practice. Supplemental material available article. Keywords: Convolutional Neural Network (CNN), Informatics, Radiomics, Supervised learning, Technology Assessment © RSNA, 2020 See also commentary by Gastounioti Kontos issue.

Language: Английский

---

Transparency of deep neural networks for medical image analysis: A review of interpretability methods

Computers in Biology and Medicine, Journal Year: 2021, Volume and Issue: 140, P. 105111 - 105111

Published: Dec. 4, 2021

Artificial Intelligence (AI) has emerged as a useful aid in numerous clinical applications for diagnosis and treatment decisions. Deep neural networks have shown the same or better performance than clinicians many tasks owing to rapid increase available data computational power. In order conform principles of trustworthy AI, it is essential that AI system be transparent, robust, fair, ensure accountability. Current deep solutions are referred black-boxes due lack understanding specifics concerning decision-making process. Therefore, there need interpretability before they can incorporated into routine workflow. this narrative review, we utilized systematic keyword searches domain expertise identify nine different types methods been used learning models medical image analysis based on type generated explanations technical similarities. Furthermore, report progress made towards evaluating produced by various methods. Finally, discuss limitations, provide guidelines using future directions imaging analysis.

Language: Английский

---