High-performance medicine: the convergence of human and artificial intelligence

Nature Medicine, Journal Year: 2018, Volume and Issue: 25(1), P. 44 - 56

Published: Dec. 28, 2018

Language: Английский

---

Resolving genetic heterogeneity in cancer

Nature Reviews Genetics, Journal Year: 2019, Volume and Issue: 20(7), P. 404 - 416

Published: March 27, 2019

Language: Английский

---

Translating cancer genomics into precision medicine with artificial intelligence: applications, challenges and future perspectives

Human Genetics, Journal Year: 2019, Volume and Issue: 138(2), P. 109 - 124

Published: Jan. 22, 2019

In the field of cancer genomics, broad availability genetic information offered by next-generation sequencing technologies and rapid growth in biomedical publication has led to advent big-data era. Integration artificial intelligence (AI) approaches such as machine learning, deep natural language processing (NLP) tackle challenges scalability high dimensionality data transform big into clinically actionable knowledge is expanding becoming foundation precision medicine. this paper, we review current status future directions AI application genomics within context workflows integrate genomic analysis for care. The existing solutions their limitations testing diagnostics variant calling interpretation are critically analyzed. Publicly available tools or algorithms key NLP literature mining evidence-based clinical recommendations reviewed compared. addition, present paper highlights adoption digital healthcare with regard requirements, algorithmic transparency, reproducibility, real-world assessment, discusses importance preparing patients physicians modern digitized healthcare. We believe that will remain main driver transformation toward medicine, yet unprecedented posed should be addressed ensure safety beneficial impact

Language: Английский

---

A Review of Deep Learning on Medical Image Analysis

Mobile Networks and Applications, Journal Year: 2020, Volume and Issue: 26(1), P. 351 - 380

Published: Nov. 5, 2020

Language: Английский

---

Tumor heterogeneity: preclinical models, emerging technologies, and future applications

Frontiers in Oncology, Journal Year: 2023, Volume and Issue: 13

Published: April 28, 2023

Heterogeneity describes the differences among cancer cells within and between tumors. It refers to describing variations in morphology, transcriptional profiles, metabolism, metastatic potential. More recently, field has included characterization of tumor immune microenvironment depiction dynamics underlying cellular interactions promoting ecosystem evolution. been found most tumors representing one challenging behaviors ecosystems. As critical factors impairing long-term efficacy solid therapy, heterogeneity leads resistance, more aggressive metastasizing, recurrence. We review role main models emerging single-cell spatial genomic technologies our understanding heterogeneity, its contribution lethal outcomes, physiological challenges consider designing therapies. highlight how dynamically evolve because leverage this unleash recognition through immunotherapy. A multidisciplinary approach grounded novel bioinformatic computational tools will allow reaching integrated, multilayered knowledge required implement personalized, efficient therapies urgently for patients.

Language: Английский

---

Integrating Artificial and Human Intelligence: A Partnership for Responsible Innovation in Biomedical Engineering and Medicine

OMICS A Journal of Integrative Biology, Journal Year: 2019, Volume and Issue: 24(5), P. 247 - 263

Published: July 17, 2019

Historically, the term "artificial intelligence" dates to 1956 when it was first used in a conference at Dartmouth College US. Since then, development of artificial intelligence has part been shaped by field neuroscience. By understanding human brain, scientists have attempted build new intelligent machines capable performing complex tasks akin humans. Indeed, future research into will continue benefit from study brain. While algorithms fast paced, actual use most (AI) biomedical engineering and clinical practice is still markedly below its conceivably broader potentials. This partly because for any algorithm be incorporated existing workflows stand test scientific validation, personal utility, application context, equitable as well. In this there much gained combining AI (HI). Harnessing Big Data, computing power storage capacities, addressing societal issues emergent applications, demand deploying HI tandem with AI. Very few countries, even economically developed states, lack adequate critical governance frames best understand steer innovation trajectories health care. Drug discovery translational pharmaceutical gain technology provided they are also informed HI. expert review, we analyze ways which applications likely traverse continuum life birth death, encompassing not only humans but all animal, plant, other living organisms that increasingly touched Examples include digital health, diagnosis diseases newborns, remote monitoring smart devices, real-time Data analytics prompt heart attacks, facial analysis software consequences on civil liberties. underscore need integration HI, note does replace medical specialists or rather, such Altogether, offer synergy responsible veritable prospects improving care prevention therapeutics while unintended automation should borne mind cultures, work force, society large.

Language: Английский

---

Clonal architecture in mesothelioma is prognostic and shapes the tumour microenvironment

Nature Communications, Journal Year: 2021, Volume and Issue: 12(1)

Published: March 19, 2021

Abstract Malignant Pleural Mesothelioma (MPM) is typically diagnosed 20–50 years after exposure to asbestos and evolves along an unknown evolutionary trajectory. To elucidate this path, we conducted multi-regional exome sequencing of 90 tumour samples from 22 MPMs acquired at surgery. Here show that exomic intratumour heterogeneity varies widely across the cohort. Phylogenetic tree topology ranges linear highly branched, reflecting a steep gradient genomic instability. Using transfer learning, detect repeated evolution, resolving 5 clusters are prognostic, with temporally ordered clonal drivers. BAP1 /−3p21 FBXW7 /-chr4 events always early clonal. In contrast, NF2 /−22q events, leading Hippo pathway inactivation predominantly late clonal, positively selected, when subclonal, exhibit parallel evolution indicating constraint. Very somatic alteration /22q occurred in one patient 12 Clonal architecture dictate MPM inflammation immune evasion. These results reveal potentially drugable bottlenecking MPM, impact on shaping landscape, potential clinical response checkpoint inhibition.

Language: Английский

---

Translating insights into tumor evolution to clinical practice: promises and challenges

Genome Medicine, Journal Year: 2019, Volume and Issue: 11(1)

Published: March 29, 2019

Accelerating technological advances have allowed the widespread genomic profiling of tumors. As yet, however, vast catalogues mutations that been identified made only a modest impact on clinical medicine. Massively parallel sequencing has informed our understanding genetic evolution and heterogeneity cancers, allowing us to place these mutational into meaningful context. Here, we review methods used measure tumor heterogeneity, potential challenges for translating insights gained achieve cancer therapy, monitoring, early detection, risk stratification, prevention. We discuss how can guide therapy by targeting clonal subclonal both individually in combination. Circulating DNA circulating cells be leveraged monitoring efficacy tracking emergence resistant subclones. The evolutionary history tumors deduced late-stage either directly sampling precursor lesions or leveraging computational approaches infer timing driver events. This approach identify recurrent represent promising avenues future detection strategies. Emerging evidence suggests processes complex dynamics are active even normal development aging. will make discriminating developing malignant neoplasms from aging cell lineages challenge. Furthermore, insight signatures may allow cancer-prevention approaches. Research studies incorporate an appreciation patterns not produce more data, but also better exploit vulnerabilities cancer, resulting improved treatment outcomes.

Language: Английский

---

Normal tissue architecture determines the evolutionary course of cancer

Nature Communications, Journal Year: 2021, Volume and Issue: 12(1)

Published: April 6, 2021

Abstract Cancer growth can be described as a caricature of the renewal process tissue origin, where architecture has strong influence on evolutionary dynamics within tumor. Using classic, well-studied model tumor evolution (a passenger-driver mutation model) we systematically alter spatial constraints and cell mixing rates to show how structure influences functional (driver) mutations genetic heterogeneity over time. This approach explores key mechanism behind both inter-patient intratumoral heterogeneity: competition for space. Time-varying leads an emergent transition from Darwinian premalignant subsequent invasive neutral growth. Initial determine mode (Darwinian neutral) without change in cell-specific rate or fitness effects. Driver acquisition during precancerous stage may modulated en route by combination two factors: limited cellular mixing. These factors occur naturally ductal carcinomas, branching topology network dictates rates.

Language: Английский

---

Artificial Intelligence and Cardiovascular Genetics

Life, Journal Year: 2022, Volume and Issue: 12(2), P. 279 - 279

Published: Feb. 14, 2022

Polygenic diseases, which are genetic disorders caused by the combined action of multiple genes, pose unique and significant challenges for diagnosis management affected patients. A major goal cardiovascular medicine has been to understand how variation leads clinical heterogeneity seen in polygenic diseases (CVDs). Recent advances emerging technologies artificial intelligence (AI), coupled with ever-increasing availability next generation sequencing (NGS) technologies, now provide researchers unprecedented possibilities dynamic complex biological genomic analyses. Combining these may lead a deeper understanding heterogeneous CVDs, better prognostic guidance, and, ultimately, greater personalized medicine. Advances will likely be achieved through increasingly frequent robust characterization patients, as well integration data other data, such cardiac imaging, coronary angiography, biomarkers. This review discusses current opportunities limitations genomics; provides brief overview AI; identifies applications, limitations, future directions AI genomics.

Language: Английский

---