The project proposes a system for extracting and analysing information from the medical records of breast cancer patients, using artificial intelligence. This would support research, personalised medicine and health decision-making, benefiting clinicians and regional authorities. This provides an automated alternative to accelerate translational research, promote personalised medicine and improve the efficiency of health services.
Innovation Summary
Innovation Overview
The steady growth in the adoption of electronic health records (EHR) around the world has introduced the possibility of extracting information from the analysis of clinical notes. These notes contain useful and valuable information to support medical research and decision-making. However, the information in clinical notes is presented in written narrative form, which makes the task of extracting and structuring the data a difficult one. Extracting this information manually would not be a feasible task because it would be time-consuming and costly.
On the other hand, cancer remains one of the leading causes of death around the world. In particular, breast cancer is the most common cancer in women worldwide and the second most common cancer overall. Knowing patterns of behaviour and information about the evolution of this disease could help to improve its treatment. However, many of these patterns are hidden in medical records and detecting them manually is neither easy nor feasible.
Accordingly, this project addresses the following research questions.
- How can the natural history of breast cancer patients be automatically structured from the analysis of medical records using artificial intelligence techniques?
- What patterns associated with the diagnosis and treatment of breast cancer patients can be obtained from the analysis of medical records?
Artificial intelligence is of great help as it is useful to analyse large amounts of information automatically. Artificial intelligence systems are being used around the world to analyse oncology clinical notes and extract patterns that help to better understand the evolution of the disease. However, most proposals and projects have focused on the English language. In the case of automatic analysis of medical records written in Spanish, there are still many challenges to be solved. Moreover, in countries such as Colombia, artificial intelligence systems would be very useful because they would help to extract information that can be used by oncologists to learn more about the disease and create action plans to improve patients' conditions.
Innovation Description
What Makes Your Project Innovative?
The project stands out for its comprehensive approach that combines the intelligent adoption of existing technologies, such as Electronic Health Record, with the innovative implementation of specific AI techniques for the analysis of medical records in the Spanish language, since there are none in this language. This combination makes the project a pioneering and essential initiative to advance medical research and improve care for breast cancer patients globally.
What is the current status of your innovation?
By 2024, the project is ready to be launched for patients after 2022 (the date on which the data was taken). In addition to this, a description of the most important findings in the development of the proposal is provided. This document will include:
- A methodology to include the use of artificial intelligence techniques in the analysis of medical records.
- Challenges and opportunities in the use of artificial intelligence for the treatment of breast cancer.
- A list of recommendations for the development of public policies related to the use of artificial intelligence in the health sector.
Innovation Development
Collaborations & Partnerships
This research was created with the help of three sectors: the Public sector by the Government of Valle, the academic sector (the university hospital) and the private sector by the company Big Data & Analitics, all played a vital role in the creation. of this innovation project.
Users, Stakeholders & Beneficiaries
In 2018, more than 2 million people were diagnosed with breast cancer in the world. In the case of Cali and according to the population cancer registry, this disease represents the second most frequent case of cancer in the city. About 12 women die daily from breast cancer
Innovation Reflections
Results, Outcomes & Impacts
At the moment, the expected results will focus on:
- Tumor staging (stage), TNM.
- Treatments received
- Adverse effects found
- Most used medications
- Events associated with the patient (admissions, relapses, survival times).
- Characteristics of patients who behave similarly.
- Other variables established by oncologists can also be obtained.
Challenges and Failures
In 2022 there was little information about artificial intelligence and its benefits, this caused the procedures before the ethics committee to be delayed and the validation of data processing could be solved. The other drawback is the financing factor for these projects. There are no items or funds for cancer, the most relevant thing is the lack of opportunity to create business niches with these alternatives in Latin America.
Conditions for Success
The most complex thing has definitely been the financing.
Replication
There are many approaches to similar problems in English but nothing in Spanish. In addition, data extraction for the health sector and specifically for breast cancer is non-existent. The potential is enormous because artificial intelligence models have 92% certainty and can be trained for any type of cancer, according to the metrics given by the algorithm, and of course this is the future, it would save lives, optimize budgets and generate more demands on medications.
Lessons Learned
We thought that because it was Colombia, such a complex innovation could not be carried out. However, the coordination between entities was wonderful and the human talent found was exceptional. The sky is the limit!
Anything Else?
We want to continue growing and these spaces help show what we are capable of doing. Thank you, Special recognition to PhD. Oswaldo Solarte, Liliana Plazas, and Marizol Badier.
Status:
- Developing Proposals - turning ideas into business cases that can be assessed and acted on
- Implementation - making the innovation happen
Innovation provided by:
Date Published:
20 February 2025