Explaining the Effectiveness of Multi-Task Learning for Efficient Knowledge Extraction from Spine MRI Reports

Sehanobish, Arijit; Sandora, McCullen; Abraham, Nabila; Pawar, Jayashri; Torres, Danielle; Das, Anasuya; Becker, Murray; Herzog, Richard J.; Odry, Benjamin L.; Ron, Vianu,

doi:10.18653/v1/2022.naacl-industry.16

Cited by 2 publications

(1 citation statement)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The multi-source heterogeneity problem of grid data improves the security and utilisation of grid data [16,17]. The smart grid data management platform eliminates the heterogeneity of multiple sources of grid data through data support and service optimisation, thereby improving the data application and operation efficiency of the smart grid [18,19].…”

Section: Introductionmentioning

confidence: 99%

Knowledge graph construction and Internet of Things optimisation for power grid data knowledge extraction

Sun

Hao

Li³

2022

Applied Mathematics and Nonlinear Sciences

View full text Add to dashboard Cite

Problems exist in power grid data management that have unclear relationships, weak security and low accuracy. By analysing the knowledge graph construction characteristics of smart grid data information and knowledge extraction, the grid data management platform is reshaped architecturally, and the knowledge graph construction technology is embedded in the grid data management framework. For the aforementioned problems, the knowledge graph construction and Internet of Things optimisation framework of power grid data knowledge extraction are proposed in this article. Firstly, the semantic search (KGSS) algorithm based on the knowledge graph is used. The KGSS algorithm extracts knowledge from structured, semi-structured and unstructured grid data through the massively parallel processing acquisition model and Hadoop database, and constructs knowledge entities, attributes and inter-entity relationships. Then, it optimises and predicts through the knowledge graph construction and Internet of Things optimisation framework extracted from power grid data knowledge. Finally, the experimental results show that the accuracy rate of the KGSS algorithm is 92%. The experimental results also show that it provides new ideas and research directions for power grid data under big data in the future.

show abstract

Section: Introductionmentioning

confidence: 99%

Knowledge graph construction and Internet of Things optimisation for power grid data knowledge extraction

Sun

Hao

Li³

2022

Applied Mathematics and Nonlinear Sciences

View full text Add to dashboard Cite

show abstract

Challenges of AI driven diagnosis of chest X-rays transmitted through smart phones: a case study in COVID-19

Antony,

Kakileti,

Shah

et al. 2023

Sci Rep

View full text Add to dashboard Cite

Healthcare delivery during the initial days of outbreak of COVID-19 pandemic was badly impacted due to large number of severely infected patients posing an unprecedented global challenge. Although the importance of Chest X-rays (CXRs) in meeting this challenge has now been widely recognized, speedy diagnosis of CXRs remains an outstanding challenge because of fewer Radiologists. The exponential increase in Smart Phone ownership globally, including LMICs, provides an opportunity for exploring AI-driven diagnostic tools when provided with large volumes of CXRs transmitted through Smart Phones. However, the challenges associated with such systems have not been studied to the best of our knowledge. In this paper, we show that the predictions of AI-driven models on CXR images transmitted through Smart Phones via applications, such as , suffer both in terms of Predictability and Explainability, two key aspects of any automated Medical Diagnosis system. We find that several existing Deep learning based models exhibit prediction instability–disagreement between the prediction outcome of the original image and the transmitted image. Concomitantly we find that the explainability of the models deteriorate substantially, prediction on the transmitted CXR is often driven by features present outside the lung region, clearly a manifestation of Spurious Correlations. Our study reveals that there is significant compression of high-resolution CXR images, sometimes as high as 95%, and this could be the reason behind these two problems. Apart from demonstrating these problems, our main contribution is to show that Multi-Task learning (MTL) can serve as an effective bulwark against the aforementioned problems. We show that MTL models exhibit substantially more robustness, 40% over existing baselines. Explainability of such models, when measured by a saliency score dependent on out-of-lung features, also show a 35% improvement. The study is conducted on WaCXR dataset, a curated dataset of 6562 image pairs corresponding to original uncompressed and compressed CXR images. Keeping in mind that there are no previous datasets to study such problems, we open-source this data along with all implementations.

show abstract

Explaining the Effectiveness of Multi-Task Learning for Efficient Knowledge Extraction from Spine MRI Reports

Cited by 2 publications

References 2 publications

Knowledge graph construction and Internet of Things optimisation for power grid data knowledge extraction

Knowledge graph construction and Internet of Things optimisation for power grid data knowledge extraction

Challenges of AI driven diagnosis of chest X-rays transmitted through smart phones: a case study in COVID-19

Contact Info

Product

Resources

About