How AI impacts clinical decision making

Healthcare is an area where advances in artificial intelligence (AI) have been both challenging and rapid, with tangible results and promising future. Today, artificial intelligence (AI) is increasingly used in the healthcare field to improve care and services. Indeed, it is possible to improve patient care by better collecting and leveraging data. For example, AI can be used to predict cardiovascular diseases and calculate the risk of a heart attack. Similarly, it can be used to determine the risk of Alzheimer’s disease or screening for breast or colon cancer.

How ai impacts clinical decision making

Data collection and analysis

The digitization of patient data and the development of health technologies are at the origin of a massive data production : healthcare data represents roughly 30% of the world’s data volume. These data sources represents a wealth of valuable information on patients’ pathologies and their treatments. The collection and analysis of data in the healthcare field can be used to assist in medical decision making and enhance the care pathway.

The data obtained from measurement tools for hospitals, health insurance cards, drug purchase history, can be utilized to determine which patient needs specific treatment or closer follow-up. It can also be used to optimize the planning of consultations. Finally, it can be used to improve prescribing practices in order to improve collaboration between patients and physicians.

Production of severity score

At the Chalon sur Saône Hospital Center, the papAI platform developed by Datategy allows IT & medical teams to quickly integrate their data into machine learning models and obtain exploitable results to enhance the care delivery. The goal is to provide physicians with an Explainable AI tool that will gather healthcare information from different sources (structured and unstructured), comprehend patient data to define a dedicated scoring reflecting patient’s health state and ultimately adapt the level and emergency of care needed. Eventually, even the data coming from the shared medical record could feed the predictive algorithm.

It improves the availability within the hospital on two major levels: 

  • The availability of information by collecting and structuring data that is usually scattered and unstructured, not only within the hospital but could also be lost between town medicine and homecare. 
  • The availability of healthcare professionals towards the patient by integrating an explainable AI in the hospital to automate certain tasks and offer better working conditions, management and resource planning

 For the patient, AI improves the security and quality of care. For hospitals, as the population ages and the occurrence of chronic diseases increase, AI is a precious daily assistance to tackle these challenges, as long as the solution is built and deployed with the consultation of every stakeholder. For healthcare professionals, AI can greatly help reduce mental load and reconnect with patients.

Early disease detection

Early detection of diseases is a major public health issue. The aging of the population and the multiplication of chronic pathologies pose new challenges to healthcare systems, particularly in terms of diagnosis. In this context, artificial intelligence (AI) technologies can offer interesting opportunities to strengthen diagnostic capabilities.  Several studies have shown the interest of AI in the early detection of diseases. 

The best known medical applications are the recognition of breast or colon cancer, but they can also be used to detect diabetes or osteoporosis. Similarly, image analysis, which uses artificial neural networks (ANNs), could allow doctors to anticipate and prevent chronic diseases such as type 2 diabetes or rheumatoid arthritis.

Oncology

Predicting the appearance of a breast tumor up to four years before it is detectable by imaging is what an algorithm developed by researchers at the Massachusets Institute of Technology (Cambridge, USA) has managed to do. Their work, published in the journal Radiology, confirms once again that artificial intelligence (AI) should play an important role in future efforts to detect cancer. A decisive issue since, for the record, five-year survival is 26% for metastasized breast cancer versus 99% if the tumor is detected early

Diagnostic accuracy

Medical diagnosis is changing. Precision medicine, based on artificial intelligence, is improving treatments for individual patients and increasing the quality of medical care. This practice can be used for diagnosis and prescription of care, but also for the monitoring and evaluation of the results. It allows to adapt treatments to the individual and to improve the therapeutic results for each patient. The information gathered by the machines is used to personalize the treatment of patients, which means reducing the number of unnecessary interventions.

This may seem like a self evident fact, but the personalization of medical treatments is also based on the needs of each patient and not only on physical criteria such as gender or age. It is now possible to take into account factors such as the patient’s health history, previous treatments and diet.

Neuro-rehabilitaiton to aid recovery of Covid-19 intensive care patients

Patients in intensive care often leave with significant brain damage: about 30 % to 50 % will not return to work within three years. These people need neuro-rehabilitation but most healthcare systems do not provide it due to the lack of cost-effective solutions. The CDAC project, contributed to the development and clinical validation of innovative technologies that have already been used for the rehabilitation of over 3 000 stroke patients across Europe. More than 30 % of intensive care patients suffer delirium and cognitive impairment, a figure that rises to 80 % among mechanically ventilated patients such as the thousands treated for COVID-19.

Improve medical image analysis

Medical image analysis is a very important field for research and diagnosis. Images are often misinterpreted because there is too much data to process, which can lead to diagnostic errors. 

Artificial intelligence (AI) technology is applied to medical image analysis to allow doctors to visualize data more clearly and quickly.It has proven to be a very effective and promising tool for improving medical treatments, especially in the early detection of diseases and optimizing patient care in general. This means that patients can be treated more effectively and quickly, resulting in a better quality of life for them and their families.

Deep learning to improve medical imaging

Medical imaging has revolutionized diagnosis, treatment and follow-up, providing fundamental information on anatomy and physiology with very high spatial resolution. However, the imaging process can be stressful for patients, and it is difficult in the presence of motion. The Deep4MI project’s objective is to advance and automate medical imaging so as to provide higher diagnostic and prognostic accuracy for clinical decision-making. Using machine and deep learning techniques, scientists will improve image acquisition, reconstruction and analysis to extract more clinical information from medical images and optimize results interpretation..

Conclusion

Healthcare is a strategic field for artificial intelligence. It is characterized by the complexity of the data to be analyzed, the large number of variables to be taken into account and the importance of accurate diagnoses to reduce medical errors and improve the quality of care. 

Artificial intelligence systems can be used to identify precursor or predictive signs of the development of a disease or syndrome, anticipating the possible consequences for the patient and adapting the level of care.  The objective is to enable healthcare specialists to anticipate the future needs of the patient and adapt their therapeutic strategy accordingly.

 

Interested in discovering papAI ?

 Our commercial team is at your disposal for any question

 
How AI impacts clinical decision making
Scroll to top