Artificial intelligence and implementation of cancer research

Technological developments over the past few decades have revealed the use of computational methods in cancer laboratory research, clinical practice, and scientific infrastructure development (e.g. biobanks) and the use of artificial intelligence to support automated and continuous processes. innovation. The idea of using healthcare computing to manage, analyze, and document clinical trials is not new.

This question coincides with the earliest implementation of healthcare since the advent of computer systems in the 1960s.

However, until the early 2000s, these discussions were only partially conducted (except for administrative purposes). Leading international projects such as the Human Genome Project, UK Biobank and Cancer Moonshot have led to changes in our understanding and expectations of computational methods in health research conversion, followed by large investments to enable these methods.

Digital transformation of the broader healthcare

Digital transformation in healthcare should be seen as a collection of technologies that work together as one piece: IoT, blockchain, and artificial intelligence (AI). In the past, health data was collected and organized manually, often in a centralized manner, representing a complex, error-prone, and error-prone environment. The advent of connected devices with embedded sensors and lightweight software has changed the data collection game. However, the biggest challenge is ensuring data integrity (and sometimes anonymity) while maintaining controllability and traceability. Data science is a combination of design and statistical techniques as well as programming techniques to classify data, extract related information, clean data, and develop data association algorithms.

Artificial intelligence, a term commonly used in medical research, remains a vague and suggestive term to describe the capabilities of machines (i.e., more accurate and less objective than human experts and prone to human error). (1) Many definitions, few curses, No. As different technologies enhance existing methods with new high-tech skills and capabilities or add new methods, the use of terms and artificial intelligence is a connecting thread in healthcare and should be seen as a diverse and enriching event. For example, in the healthcare sector, AI is often used as clinical decision support software (CDS) (2) to inform the diagnosis, treatment, prevention, and recovery or amelioration of disease, or patient’s condition. The final decision depends on the experience of the specialized medical team. However, the intellectual component provides k eputus support, clinical diagnosis requires intelligence among many other factors. A good physician does not know a lot (a lot) of information, but one who uses that knowledge (interprets the information) and acumen, clinical experience, and background knowledge to diagnose guess.

The AI revolution is happening: sometimes quietly

Healthcare is evolving rapidly, and AI is slowly changing several key industries, such as improving breast cancer detection and dramatically improving the work of radiologists. (3) From a long-term economic perspective, AI will reduce the cost of high-volume repetitive healthcare tasks and thus have a significant impact on healthcare economics. In addition, if the use of artificial intelligence improves the speed of early disease detection, treatment will be easier, less invasive, and more successful. However, AI algorithms are based on long-term knowledge (disease-specific data sets), which helps to better understand the disease and reduce the risk of making wrong decisions, so the impact of implementing a claim will be felt in the long run. Finally, there is an ethical problem with the use of AI in the sensitive medical field: a wrong decision can be interpreted as negligent treatment by a physician. This can lead to questions like Who is responsible for the failures of AI? What mistakes do AI system designers make? Is this an implementation bug or an AI end-user fault?

Cancer research and the possibilities of artificial intelligence

In cancer research particularly, various initiatives have generated a large amount of data on cancer over the past decade. These datasets are derived from detailed analysis of characteristic tumor samples using high-throughput techniques and platforms. The Cancer Genome Atlas (TCGA) is the most comprehensive publicly available compilation of cancer profiling, including imaging, genomic, epigenetic, proteomic, and histopathological data types. (4) This publicly available and detailed information is used as an important resource for building predictive models and provides an opportunity to combine research data from local sources with data sets. High-quality reference material. Many studies have demonstrated the benefits of this integration, for example, training AI-based predictive models on different groups instead of a single data source has been shown to predict the target and Mechanism of action of small molecules against cancer and improve overall prediction accuracy. . Therefore, there is moderate optimism that AI-based research will improve cancer detection, classification, and classification; effectiveness and synergy of the drug; and ultimately significantly affect the outcome of treatment.

Cancer Research, AI and LMIC

However, the use of AI in cancer research also poses challenges beyond the immediate technical requirements of different research methods. As AI relies heavily on data that is large and abundant, this could be a limitation for low- and middle-income countries (LMICs) where quality data is not readily available. High. , or sometimes not at all. In addition, the availability of the necessary computing infrastructure and processing power, as well as the availability of appropriately trained personnel necessary to deploy and operate these applications, may be a challenge. As a result, AI-based health approaches that can be implemented in LMICs may differ significantly from those tested in high-income settings and their successful adoption. maybe less homogeneous and more context-sensitive.

In short, innovation is an important part of the growth and development of any organization or industry. The implementation of artificial intelligence is one of the major innovations in medicine and medical research and holds promise for the advancement of cancer research and ultimately cancer treatment. In a rapidly changing world where major advances from analog to digital and artificial intelligence may be required, the latter may require contextual (or regional) deployment to be applied with long-term success.