Implementation of Artificial Intelligence in Diagnostic Practice as a Next Step after Going Digital: The UMC Utrecht Perspective

Flach, Rachel; Fransen, Nina L.; Sonnen, Andreas F.‐P.; Nguyen, Tri Q.; Breimer, Gerben E.; Veta, Mitko; Stathonikos, Nikolas; Dooijeweert, Carmen van; Diest, Paul J. van

doi:10.3390/diagnostics12051042

Cited by 15 publications

(13 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Only one of the participants was a cytopathologist (who also subspecialized in other areas of anatomic pathology), which may reflect the current underrepresentation of this subspecialty in computational pathology/AI due to the challenges of generating and storing z-stacked WSIs (which may resolve with the introduction of dedicated cytology whole-slide scanners in the near future). 42 Also, as most of the participants were practicing in North America and Europe, the results of this study may reflect a predominantly North American/European perspective that differs from the perspectives of those practicing in other parts of the world. Lastly, as our study was targeted toward a specific respondent demographic, individuals from other stakeholder groups (such as pathology trainees, pathologists with little or no experience in computational pathology/AI, pathology laboratory technicians, computer scientists and others working in the computational pathology/AI space with non-medical backgrounds, physicians in other specialties, and patients) were not represented in our results.…”

Section: Discussionmentioning

confidence: 83%

“…34 Numerous reviews have been published on the emerging and future applications of pathology AI to cancer diagnosis, prognostication, and treatment response prediction, metastasis detection in lymph nodes, single and multiplex biomarker quantification, tumor content/cellularity assessment for molecular testing, mutation status prediction, and a multitude of other tasks in pathology. [35][36][37][38][39][40][41][42][43][44][45] Despite the recent progress and enthusiasm surrounding the application of AI to pathology, few algorithms are currently in routine clinical use, 37,46 with a dearth of prospective multi-center, randomized trials present evaluating the impact of these algorithms in clinical settings. 47,48 Further, ethical concerns have been raised regarding potential patient data privacy breaches, biased datasets producing systemic algorithmic bias, potential harm related to erroneous or misleading AI-generated outputs, and exacerbation of healthcare disparities due to unequal access to AI.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

The future of computational pathology: expectations regarding the anticipated role of artificial intelligence in pathology by 2030

Berbís¹,

McClintock

Bychkov

et al. 2022

Preprint

View full text Add to dashboard Cite

Background: Artificial intelligence (AI) is rapidly fueling a fundamental transformation in the practice of pathology. However, the clinical integration of AI remains challenging, with no AI algorithms to date enjoying routine adoption within typical anatomic pathology (AP) laboratories. This survey gathered current expert perspectives and expectations regarding the role of AI in AP from those with first-hand computational pathology and AI experience. Methods: Perspectives were solicited using the Delphi method from 24 subject matter experts between December 2020 and February 2021 regarding the anticipated role of AI in pathology by the year 2030. The study consisted of three consecutive rounds: 1) an open-ended, free response questionnaire generating a list of survey items; 2) a Likert-scale survey scored by experts and analyzed for consensus; and 3) a repeat survey of items not reaching consensus to obtain further expert consensus. Findings: Consensus opinions were reached on 141 of 180 survey items (78.3%). Experts agreed that AI would be routinely and impactfully used within AP laboratory and pathologist clinical workflows by 2030. High consensus was reached on 100 items across nine categories encompassing the impact of AI on (1) pathology key performance indicators (KPIs) and (2) the pathology workforce and specific tasks performed by (3) pathologists and (4) AP lab technicians, as well as (5) specific AI applications and their likelihood of routine use by 2030, (6) role of AI in integrated diagnostics, (7) pathology tasks likely to be fully automated using AI, and (8) regulatory/legal and (9) ethical aspects of AI integration in pathology. Interpretation: This is the first systematic consensus study detailing the expected short/mid-term impact of AI on pathology practice. These findings provide timely and relevant information regarding future care delivery in pathology and raise key practical, ethical, and legal challenges that must be addressed prior to the successful clinical implementation of AI.

show abstract

Section: Discussionmentioning

confidence: 83%

Section: Introductionmentioning

confidence: 99%

The future of computational pathology: expectations regarding the anticipated role of artificial intelligence in pathology by 2030

Berbís¹,

McClintock

Bychkov

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…At UMC Utrecht, we employ three high throughput Hamamatsu XR scanners from Hamamatsu Photonics K.K. (Hamamatsu City, Japan) to digitize slides 27,28 . The Sectra Picture Archiving Communication System (PACS), provided by Sectra AB (Linkoping, Sweden) is used as the image management and workflow system.…”

Section: Methodsmentioning

confidence: 99%

“…(Hamamatsu City, Japan) to digitize slides. 27,28 The Sectra Picture Archiving Communication System (PACS), provided by Sectra AB (Linkoping, Sweden) is used as the image management and workflow system. With PACS, images of HE slides can be viewed and annotated using the integrated ruler tool, which is useful for measuring.…”

Section: Histopathologymentioning

confidence: 99%

A closer look at the resection margins of buccal mucosa cancer: Their influence on local recurrence free survival

et al. 2023

Self Cite

View full text Add to dashboard Cite

Background: The adequate surgical margin for local control of buccal mucosa squamous cell carcinoma (BMSCC) is under debate. This study investigates surgical margins and other factors associated with local recurrence free survival (LRFS) in a large cohort of BMSCC patients.Methods: Multiple factors were evaluated retrospectively in 97 patients with BMSCC. Cox-regression and Kaplan-Meier curves were used for analysis.Results: The local recurrence rate was 23%. The tumor-free margin was <5.0 mm in 89% of the patients and the deep margin was significantly more often inadequate. Multivariate analysis associated pT3-classification, former smokers, tumor-free margin status, and postoperative (chemo)radiation (PO (ch)RT) with local recurrence. Re-resections did not improve LRFS in patients with <5.0 mm tumor-free margins.Conclusions: Adequate tumor-free margins are pivotal for LRFS of BMSCC. PO(ch)RT, not re-resection, can improve LRFS in patients with <5.0 mm tumor-free margins.

show abstract

“…The high incidence of breast cancer (BC) cases poses a substantial burden on healthcare providers, including pathologists. To alleviate some of the clinical burden, decision support systems based on computer vision and machine learning can be utilised [2], including systems designed for characterization and diagnosis of BC from routine digital histopathology whole slide images (WSI) from tissue sections stained with Haematoxylin-eosin (H&E).…”

Section: Introductionmentioning

confidence: 99%

Ensembles for improved detection of invasive breast cancer in histological images

Solorzano

Robertson

Hartman

et al. 2023

Preprint

View full text Add to dashboard Cite

Accurate detection of invasive breast cancer (IC) can provide decision support to pathologists as well as improve downstream computational analyses, where detection of IC is a first step. Tissue containing IC is characterized by the presence of specific morphological features, which can be learned by convolutional neural networks (CNN). Here, we compare the use of a single CNN model versus an ensemble of several base models with the same CNN architecture, and we evaluate prediction performance as well as variability across ensemble based model predictions.Two in-house datasets comprising 587 WSI are used to train an ensemble of ten InceptionV3 models whose consensus is used to determine the presence of IC. A novel visualization strategy was developed to communicate ensemble agreement spatially. Performance was evaluated in an internal test set with 118 WSIs, and in an additional external dataset (TCGA breast cancer) with 157 WSI.We observed that the ensemble-based strategy outperformed the single CNN-model alternative with respect to accuracy on tile level in 89% of all WSIs in the test set. The overall accuracy was 0.92 (DICE coefficient, 0.90) for the ensemble model, and 0.85 (DICE coefficient, 0.83) for the single CNN alternative in the internal test set. For TCGA the ensemble outperformed the single CNN in 96.8% of the WSI, with an accuracy of 0.87 (DICE coefficient 0.89), the single model provides an accuracy of 0.75 (DICE coefficient 0.78)The results suggest that an ensemble-based modeling strategy for breast cancer invasive cancer detection consistently outperforms the conventional single model alternative. Furthermore, visualization of the ensemble agreement and confusion areas provide direct visual interpretation of the results. High performing cancer detection can provide decision support in the routine pathology setting as well as facilitate downstream computational analyses.

show abstract

Implementation of Artificial Intelligence in Diagnostic Practice as a Next Step after Going Digital: The UMC Utrecht Perspective

Cited by 15 publications

References 32 publications

The future of computational pathology: expectations regarding the anticipated role of artificial intelligence in pathology by 2030

The future of computational pathology: expectations regarding the anticipated role of artificial intelligence in pathology by 2030

A closer look at the resection margins of buccal mucosa cancer: Their influence on local recurrence free survival

Ensembles for improved detection of invasive breast cancer in histological images

Contact Info

Product

Resources

About