Each year, 18 million people worldwide are diagnosed with cancer, a dangerous disease that can cause a large number of patients to die and costly medical systems. Patients often face multiple treatment options that are confusing and headaches. At the same time, oncologists need to screen a large number of medical literature and genomic data to develop the best treatment plan for each patient. Researchers have also been plagued with trials, often failing to recruit enough patients.
Studies have shown that the Watson AI system has value in supporting evidence-based treatment decisions, enhancing patient confidence in treatment planning, labeling genomic variation, and identifying clinical interventions, helping multidisciplinary oncology teams make smarter based on integrated scientific evidence. Decision making, providing important insights and information that cannot be discovered through manual identification, and improving patient satisfaction by providing a comprehensive treatment plan.
Based on information provided by Watson's oncology solution, the multidisciplinary oncology team adjusted its treatment plan for 13.6% of cases. In the Blinded evaluation of blind breast cancer, lung cancer and colorectal cancer in India, Manny The Paar Hospital Multidisciplinary Oncology Team changed the treatment decisions for 13.6% of cases based on information provided by the Watson Cancer Solution. Researchers based on Watson's new treatment options (55%), more personalized alternatives (30%), or new insights (15%) from genotypes, phenotypic data, and evolving clinical experience The treatment plan for the case was adjusted. The decision support provided by artificial intelligence can alleviate the cognitive burden of oncologists. This cognitive burden is an important factor that causes doctors to be overwhelmed. At the same time, artificial intelligence helps to reduce the irregularity of treatment.
The Watson Gene Solution provides new insights for oncologists in patients with hematologic malignancies. In a study of 54 patients with hematologic malignancies, the Watson Gene Solutions expert opinion on genome annotation and artificial formation of sequencing results Very well matched (randomly screened genomic annotation consensus rate of 90%), and the solution also found clinical insights that were not discovered by manual interpretation in 33% of the cases tested. This suggests that the review of gene sequencing results that require a lot of manpower can be improved with tools like the Watson Gene Solutions.
Watson's oncology solution boosts cancer patients' confidence in disease: Yu Zhonghe, director of the oncology department at Beijing Chaoyang Integrated Chinese and Western Medicine Emergency Rescue Center, said that by integrating Watson's oncology solution into a seven-step patient engagement and consultation process It can help patients better understand the disease and treatment plan, thus enhancing their confidence in the treatment plan.
“When patients don’t understand their disease and treatment options, we often see their lack of confidence in treatment, which may reduce patient acceptance of treatment options. Introduced in the development of treatment options and interactions with patients. Watson has helped us significantly improve patient satisfaction, increase patient trust in treatment options, and make them more willing to optimize the treatment options screened by Watson AI-supported tumor consultations," Yu Zhonghe said.
A total of more than 70 peer-reviewed research reports, posters, and abstracts support the products and services of the Watson Health family of solutions in oncology and genomics. Nathan Levitan, chief medical officer, MD, and MBA at IBM Watson Health Oncology and Genomics, said: "Artificial intelligence is at an early stage of medical health decision making. Watson Health's research at the annual meeting of the American Society of Clinical Oncology provides Convincing evidence that this technology will play an important role in helping oncologists improve cancer treatment options for each patient."
During the annual meeting of the American Society of Clinical Oncology, IBM Watson Health also released data on a new method by using the abstract texts of papers cited in three expert resources (NCCN, NCI-PDQ, and Hemonc.org). Learn to automatically identify high-quality scientific publications related to the clinic. The model has 93% accuracy in classification, 95% sensitivity and 91% specificity. This means that machine learning can be used to automatically identify relevant clinical medical publications, and can reduce the time it takes for clinicians to find relevant evidence for a patient's treatment plan.
IBM Watson Health improves the workflow experience of oncologists in critical healthcare markets based on feedback from physicians and insights from scientific data to customize products and solutions. Currently, Watson Health oncology and genomics products are supporting doctors and patients associated with cancer care in more than 15 healthcare markets worldwide.