Explainable AI (XAI) in Healthcare

By Diya Mathew on

 

Explainable AI (XAI) in Healthcare



Artificial Intelligence has been an uprising technology since a decade ago and is still an emerging technology providing various levels of enhancements. But sometimes it is better to interpret the results of the solution in such a way that a user can understand it. Which is referred to as explainable or interpretable Artificial Intelligence. AI is widely used in medical and healthcare fields to help clinicians to make decisions, increase efficiency and reduce the mortality rate. However, the end-user cannot see the logic behind these outputs. Due to their black-box character, these models are difficult for physicians to comprehend, and as a result, they cannot be used in clinical practice. There comes the scope of explainable Artificial intelligence. The act of reliability on artificial intelligence is the next question. As said with a black box nature, AI systems are not widely accepted for many medicinal decision-making scenarios, especially during the detection of epilepsy and other symptoms using AI. But interpreting the logic in such a way a human can understand it will make the Artificial intelligence systems much more reliable as it explains why. Many techniques can be used to interpret logic. Multiple XAI techniques are used for interpreting the logic such as gradient, integrated gradient, LRP, Deeplift etc.

XAI Background In Healthcare

Artificial Intelligence (AI) systems have been widely used in the healthcare industry for several years for analyzing and diagnosing health data. Healthcare is becoming more reliant on AI with the help of smart wearable devices, and this improves the scope of personalized medicine. Curing or preventing diseases from progressing to worse stages is proportional to their early detection. AI along with smart wearable devices helps in the early detection of diseases since it facilitates real-time monitoring of health data. But the lack of explainability of the AI model’s predictions limits the acceptance of AI systems in healthcare since trusting AI systems requires deep technical and statistical knowledge. The lack of trust in the black box operation of AI systems as well as the difficulties in interpreting the results necessitates AI models that can be explained. Explanation of complex AI systems is not a new idea since expert systems supported reasoning architectures in the 1980s. In the healthcare industry, physicians must understand how and why the AI model made the decision before trusting the decision as it affects people’s lives. The introduction of Explainable AI (XAI) enhanced the user’s trust in AI systems since it describes the machine decisions and predictions made by the AI systems. XAI methods enhance the transparency of AI systems as well as helps to identify the factors that influenced the prediction made by an AI system. AI systems are learned from data and predictions are made based on it. It is possible for these predictions to be incorrect at times. XAI methods will allow the system to understand the learned rules and correct errors, thus improving the system’s accuracy. This helps healthcare professionals to make reasonable and data-driven decisions that improve the quality of healthcare services. Studies show that AI systems in the healthcare industry are more effective when XAI models are combined with clinical knowledge and thus improve their reliability in healthcare.

Significance of XAI methods to increase the reliability of AI Systems

Artificial intelligence (AI) is widely employed in the medical and healthcare areas to assist physicians in making choices, boost efficiency, and lower mortality rates. The end user, however, cannot perceive the reasoning underlying these outputs. These models are difficult for clinicians to understand due to their black-box nature, and as a result, they cannot be employed in clinical practice. The XAI methods increase the interpretability of the underlying decision-making process and which in turn makes the results convincing and thus increases the degree of user’s trust in AI systems.

Clinical feedback adds more conviction to the AI system output since the combination of clinical knowledge and AI systems provides better trustworthy results. In a nutshell, we could summarize that using an XAI to interpret the AI-based results significantly increases the reliability of an AI system.

References

Beddiar, D., Oussalah, M., Tapio, S., 2022. Explainability for Medical Image Captioning, in: 2022 Eleventh International Conference on Image Processing Theory, Tools and Applications (IPTA). Presented at the 2022 Eleventh International Conference on Image Processing Theory, Tools and Applications (IPTA), IEEE, Salzburg, Austria, pp. 1–6. https://doi.org/10.1109/IPTA54936.2022.9784146

Cavaliere, F., Cioppa, A., Marcelli, A., Parziale, A. and Senatore, R. (12AD). Parkinson’s Disease Diagnosis: Towards Grammar-based Explainable Artificial Intelligence.

Comments

Post a Comment

Popular posts from this blog

Transfer Learning Vs Fine Tuning

By Diya Mathew on
Image
Transfer Learning Transfer Learning involves adapting knowledge attained from one task to resolve other associated tasks we haven't faced before. When it comes to machine learning, it is possible to transfer knowledge from one domain/task to another.   Fine Tuning Transfer learning can be achieved through fine-tuning. The pretraining process starts with random weights and no prior knowledge, and the model is trained from scratch. A large amount of data is usually used for this pretraining. As a result, it requires a large amount of data, and training may  take several weeks. It also causes an environmental impact.  As opposed to pre-trained models, fine-tuned models are trained after they have been pretrained. The fine-tuning process begins by acquiring a pre-trained  language model, followed by additional training with the chosen dataset. The cost of fine-tuning a model is lower in terms of time, data, and finances. Training is also less constraining than full pretraining, so it i
Read more

AutoML

By Diya Mathew on
Image
                      AutoML Automated machine learning(AutoML) is a process to fully automates the end-to-end of the machine learning process and makes it more usable by intended users. AutoML is also related to AI and the introduction of AutoML in Robotics is a boom in technology. AutoML tools will outperform the manual machine learning (ML) modelling, with high robustness then only it has relevance. How to provide robustness while dealing with real-world datasets, with considerable time constraints is a relevant question in this area. TPOT, Auto-sklearn, Auto-keras and H2O-Automl are some of the open-source AutoML tools and Google cloud Automl and Microsoft AzureML are commercial AutoML Tools. To compare the robustness of the open-source AutoML tools many researchers take different datasets from openML and conduct studies based on different criteria. They use quantitative measurement based on time to understand the robustness of these AutoML tools. These experiments show, that diffe
Read more

Pre-trained Language Models (PTLM) in NLP

By Diya Mathew on
Image
  Pre-trained Language Models (PTLM) in NLP The main purpose behind Natural Language Processing(NLP) is to make the computer system capable to understand and communicate the language as a human. NLP tasks are challenging because it is not only important to understand individual words, but also to understand the context in which they are used. According to NLP, each word has a meaning, but the meaning is strongly influenced by the context in which it is used. The Language Model is a machine learning model that predicts the next word based on the parts of a sentence. A pre-trained language model has been trained on a large corpus of data. Through this training, the model learns the language’s general rules for word usage and how it is written. The model is then trained with a task-specific dataset.                                                        Pretrained language model concept Text-to-Text Transfer Transformer Model(T5) T5 (Raffel et al., 2019) is Google’s state-of-the-art (SOTA
Read more