The prognostic impact of the tumour stroma fraction: A machine learning-based analysis in 16 human solid tumour types

EBioMedicine, Journal Year: 2021, Volume and Issue: 65, P. 103269 - 103269

Published: March 1, 2021

The development of a reactive tumour stroma is hallmark progression and pronounced generally considered to be associated with clinical aggressiveness. variability between types regarding fraction, its prognosis associations, have not been systematically analysed.Using an objective machine-learning method we quantified the in 16 solid cancer from 2732 patients, representing retrospective tissue collections surgically resected primary tumours. Image analysis performed segmentation into stromal epithelial compartment based on pan-cytokeratin staining autofluorescence patterns.The fraction was highly variable within across types, kidney showing lowest pancreato-biliary type periampullary highest proportion (median 19% 73% respectively). Adjusted Cox regression models revealed both positive (pancreato-biliary oestrogen negative breast cancer, HR(95%CI)=0.56(0.34-0.92) HR(95%CI)=0.41(0.17-0.98) respectively) (intestinal HR(95%CI)=3.59(1.49-8.62)) associations survival.Our study provides quantification major cancer. Findings strongly argue against commonly promoted view general high abundance poor prognosis. results also suggest that full exploitation prognostic potential requires analyses go beyond determination abundance.The Swedish Cancer Society, Lions Foundation Uppsala, Government Grant for Clinical Research, Mrs Berta Kamprad Foundation, Sweden, Sellanders foundation, P.O.Zetterling Sjöberg Sweden.

Language: Английский

---

Clinlabomics: leveraging clinical laboratory data by data mining strategies

BMC Bioinformatics, Journal Year: 2022, Volume and Issue: 23(1)

Published: Sept. 24, 2022

Abstract The recent global focus on big data in medicine has been associated with the rise of artificial intelligence (AI) diagnosis and decision-making following advances computer technology. Up to now, AI applied various aspects medicine, including disease diagnosis, surveillance, treatment, predicting future risk, targeted interventions understanding disease. There have plenty successful examples using data, such as radiology pathology, ophthalmology cardiology surgery. Combining become a powerful tool change health care, even nature screening clinical diagnosis. As all we know, laboratories produce large amounts testing every day laboratory combined may establish new treatment attracted wide attention. At present, concept radiomics created for imaging AI, but definition lacked so that many studies this field cannot be accurately classified. Therefore, propose omics (Clinlabomics) by combining AI. Clinlabomics can use high-throughput methods extract feature from blood, body fluids, secretions, excreta, cast test data. Then statistics, machine learning, other read more undiscovered information. In review, summarized application medical fields. Undeniable, is method assist fields still requires further validation multi-center environment laboratory.

Language: Английский

---

Integrating artificial intelligence in pathology: a qualitative interview study of users' experiences and expectations

Modern Pathology, Journal Year: 2022, Volume and Issue: 35(11), P. 1540 - 1550

Published: Aug. 4, 2022

Abstract

Recent progress in the development of artificial intelligence (AI) has sparked enthusiasm for its potential use pathology. As pathology labs are currently starting to shift their focus towards AI implementation, a better understanding how tools can be optimally aligned with medical and social context daily practice is urgently needed. Strikingly, studies often fail mention ways which should integrated decision-making processes pathologists, nor do they address this achieved an ethically sound way. Moreover, perspectives pathologists other professionals within concerning integration remains underreported topic. This article aims fill gap literature presents first in-depth interview study professionals' on possibilities, conditions prerequisites explicated. The results have led formulation three concrete recommendations support integration, namely: (1) foster pragmatic attitude toward development, (2) provide task-sensitive information training health care working departments (3) take time reflect upon users' changing roles responsibilities.

Language: Английский

---

Translation of tissue-based artificial intelligence into clinical practice: from discovery to adoption

Oncogene, Journal Year: 2023, Volume and Issue: 42(48), P. 3545 - 3555

Published: Oct. 24, 2023

Abstract Digital pathology (DP), or the digitization of images, has transformed oncology research and cancer diagnostics. The application artificial intelligence (AI) other forms machine learning (ML) to these images allows for better interpretation morphology, improved quantitation biomarkers, introduction novel concepts discovery diagnostics (such as spatial distribution cellular elements), promise a new paradigm biomarkers. AI tissue analysis can take several conceptual approaches, within domains language modelling image analysis, such Deep Learning Convolutional Neural Networks, Multiple Instance risk scores their ML. use different approaches solves problems workflows, including assistive applications detection grading tumours, quantification delivery established image-based biomarkers treatment prediction prognostic purposes. All formats, applied digital are also beginning transform our approach clinical trials. In parallel, novelty DP/AI devices related computational science pipeline introduces requirements manufacturers build into design, development, regulatory post-market processes, which may need be taken account when using tissues in discovery. Finally, represents challenge way we accredit diagnostic tools with applicability, understanding will allow patients have access generation complex

Language: Английский

---

Society of Toxicologic Pathology Digital Pathology and Image Analysis Special Interest Group Article*: Opinion on the Application of Artificial Intelligence and Machine Learning to Digital Toxicologic Pathology

Toxicologic Pathology, Journal Year: 2019, Volume and Issue: 48(2), P. 277 - 294

Published: Oct. 23, 2019

Toxicologic pathology is transitioning from analog to digital methods. This transition seems inevitable due a host of ongoing social and medical technological forces. Of these, artificial intelligence (AI) in particular machine learning (ML) are globally disruptive, rapidly growing sectors technology whose impact on the long-established field histopathology quickly being realized. The development increasing numbers algorithms, peering ever deeper into histopathological space, has demonstrated scientific community that AI platforms now poised truly future precision personalized medicine. However, as with all great advances, there implementation adoption challenges. review aims define common relevant ML terminology, describe data generation interpretation, outline current potential business cases, discuss validation regulatory hurdles, most importantly, propose how overcoming challenges this burgeoning may shape toxicologic for years come, enabling pathologists contribute even more effectively answering questions solving global health issues. [Box: see text]

Language: Английский

---