Explainability and artificial intelligence in medicine

Reddy, Sandeep

doi:10.1016/s2589-7500(22)00029-2

Cited by 91 publications

(62 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The explainability of AI models is gaining importance in clinical practices. Transparent algorithms or explanatory approaches create trust and can make adopting AI systems less risky for clinical practitioners [ 67 ].…”

Section: Discussionmentioning

confidence: 99%

Future of Artificial Intelligence Applications in Cancer Care: A Global Cross-Sectional Survey of Researchers

et al. 2023

View full text Add to dashboard Cite

Cancer significantly contributes to global mortality, with 9.3 million annual deaths. To alleviate this burden, the utilization of artificial intelligence (AI) applications has been proposed in various domains of oncology. However, the potential applications of AI and the barriers to its widespread adoption remain unclear. This study aimed to address this gap by conducting a cross-sectional, global, web-based survey of over 1000 AI and cancer researchers. The results indicated that most respondents believed AI would positively impact cancer grading and classification, follow-up services, and diagnostic accuracy. Despite these benefits, several limitations were identified, including difficulties incorporating AI into clinical practice and the lack of standardization in cancer health data. These limitations pose significant challenges, particularly regarding testing, validation, certification, and auditing AI algorithms and systems. The results of this study provide valuable insights for informed decision-making for stakeholders involved in AI and cancer research and development, including individual researchers and research funding agencies.

show abstract

Section: Discussionmentioning

confidence: 99%

Future of Artificial Intelligence Applications in Cancer Care: A Global Cross-Sectional Survey of Researchers

et al. 2023

View full text Add to dashboard Cite

show abstract

“…In addition, only data and image balance processing were performed for benign and malignant cases in this project. Finally, the interpretability of AI 46 could increase clinicians’ understanding of the results and reduce the risks of using AI, which requires further study.…”

Section: Discussionmentioning

confidence: 99%

Towards precision medicine based on a continuous deep learning optimization and ensemble approach

Jin

Wang

et al. 2023

npj Digit. Med.

View full text Add to dashboard Cite

We developed a continuous learning system (CLS) based on deep learning and optimization and ensemble approach, and conducted a retrospective data simulated prospective study using ultrasound images of breast masses for precise diagnoses. We extracted 629 breast masses and 2235 images from 561 cases in the institution to train the model in six stages to diagnose benign and malignant tumors, pathological types, and diseases. We randomly selected 180 out of 3098 cases from two external institutions. The CLS was tested with seven independent datasets and compared with 21 physicians, and the system’s diagnostic ability exceeded 20 physicians by training stage six. The optimal integrated method we developed is expected accurately diagnose breast masses. This method can also be extended to the intelligent diagnosis of masses in other organs. Overall, our findings have potential value in further promoting the application of AI diagnosis in precision medicine.

show abstract

“…One major discussion in the field of data sciences, particularly in the health care area, is model interpretability (Lipton, 2016;Reddy, 2022). Dealing with complex models complicates the user capacity to understand and learn from the predictions, which, in the biomedicine field, can limit the impact that such models can have on disease diagnosis or treatment (Vellido, 2020;Quinn et al, 2022).…”

Section: Discussionmentioning

confidence: 99%

“…Nevertheless, recent advances in single-cell genomics are expected to dramatically expand the catalogue of cell types and developmental stages in which gene expression profiles and enhancer maps can be explored (Abe et al, 2022; Eraslan et al, 2022; Tabula Sapiens Consortium* et al, 2022). Therefore, as single-cell datasets are generated in human embryos and incorporated to POSTRE, the tool will be capable of handling additional congenital abnormalities and also its sensitivity will increase. One major discussion in the field of data sciences, particularly in the health care area, is model interpretability (Lipton, 2016; Reddy, 2022). Dealing with complex models complicates the user capacity to understand and learn from the predictions, which, in the biomedicine field, can limit the impact that such models can have on disease diagnosis or treatment (Vellido, 2020; Quinn et al, 2022).…”

Section: Discussionmentioning

confidence: 99%

POSTRE: a tool to predict the pathological effects of human structural variants

Sánchez-Gaya

Rada-Iglesias

2022

Preprint

View full text Add to dashboard Cite

Understanding the pathological impact of non-coding genetic variation is a major challenge in medical genetics. Accumulating evidences indicate that a significant fraction of genetic alterations, including structural variants (SVs), can cause human disease by altering the function of non-coding regulatory elements, such as enhancers. In the case of SVs, described pathomechanisms include changes in enhancer dosage and long-range enhancer-gene communication. However, there is still a clear gap between the need to predict and interpret the medical impact of non-coding variants, and the existence of tools to properly perform these tasks. To reduce this gap, we have developed POSTRE (Prediction Of STRuctural variant Effects), a computational tool to predict the pathogenicity of SVs implicated in a broad range of human congenital disorders. By considering disease-relevant cellular contexts, POSTRE identifies SVs with either coding or long range pathological consequences with high specificity and sensitivity. Furthermore, POSTRE not only identifies pathogenic SVs, but also predicts the disease-causative genes and the underlying pathological mechanism (e.g, gene deletion, enhancer disconnection, enhancer adoption, etc.). POSTRE is available at https://github.com/vicsanga/Postre.

show abstract

Explainability and artificial intelligence in medicine

Cited by 91 publications

References 9 publications

Future of Artificial Intelligence Applications in Cancer Care: A Global Cross-Sectional Survey of Researchers

Future of Artificial Intelligence Applications in Cancer Care: A Global Cross-Sectional Survey of Researchers

Towards precision medicine based on a continuous deep learning optimization and ensemble approach

POSTRE: a tool to predict the pathological effects of human structural variants

Contact Info

Product

Resources

About