A technological innovation can contribute to the search for the elimination of leprosy, one of the oldest diseases of humanity. An international team of scientists, led by the Oswaldo Cruz Institute (IOC/Fiocruz) in partnership with Microsoft AI for Health and the Novartis Foundation, has developed a diagnostic assistant, based on artificial intelligence, that can help identify suspected cases of the infection. The technology was called AI4Leprosy.
An article published in the scientific journal 'The Lancet Regional Health - Americas' shows that, from photos of the lesions on the skin of patients and clinical data observed by physicians, the diagnostic assistant indicates the probability of the disease, getting more than 90% correct. of the cases in the tests carried out. According to the scientists, the publication is a proof of concept for the method, which should serve as a basis for creating a mobile application for use by health professionals.
“Our study proves that it is possible to reach a suspected leprosy diagnosis with an artificial intelligence algorithm. This tool can support the doctor's decision to start the investigation, accelerating the confirmation of the diagnosis and the beginning of the treatment, which is essential to interrupt the transmission of the disease and prevent sequelae", says the head of the Leprosy Laboratory of the IOC, Milton Ozório Moraes.
“By taking advantage of artificial intelligence and machine learning, patients can find the right specialists at the right time, potentially improving the lives of millions of people around the world. The possibilities with this type of application are endless, especially for diseases. multifaceted as leprosy,” says Microsoft Chief Data Scientist Juan Lavista Ferres.
"These results are an exciting indication of the potential of AI4Leprosy. We are convinced that future validation and global launch can help cover the last few miles of leprosy elimination, using the newest technology available to end one of the oldest known scourges by man," says Novartis Foundation President Ann Aerts.
Delay in diagnosis is one of the biggest challenges for leprosy elimination. Since it was introduced in the 1980s, multidrug therapy – a treatment based on a combination of antibiotics – has cured about 18 million people, reducing the prevalence of the disease by 95%. The drugs are donated by Novartis through the World Health Organization (WHO) to be offered free of charge to patients worldwide. In addition to promoting healing, the treatment blocks the transmission of Mycobacterium leprae, which causes the infection.
However, because of the delay in identifying the condition, the bacteria continues to spread and many people still develop visible deformities, loss of movement of the feet or hands and vision problems due to leprosy. In 2019, more than 200 thousand new cases of the disease were registered worldwide, approximately 10 thousand with advanced lesions. In Brazil, the second most affected country on the planet, 27 new cases were detected, including 2,3 with advanced damage.
With the impact of the Covid-19 pandemic, which disrupted health services, leprosy diagnoses dropped, raising the concern of specialists. According to the WHO, the global drop in the detection of new cases was 37% in 2020. In Brazil, a survey by the Brazilian Society of Dermatology (SBD) pointed to a 35% reduction in records in 2020 and 45% in 2021, compared to 2019.
Considering the first results observed, scientists believe that artificial intelligence can contribute to achieving the goals set by the World Health Organization (WHO), such as reducing new cases of the infection by 70% by 2030 and, in the long term, interrupting transmission. of the grievance.
The starting point for the development of the virtual assistant for diagnosing leprosy was a type of image recognition algorithm that has been applied, for example, to support the diagnosis of melanoma, a form of skin cancer. One of the authors of the study, a doctor in oncology with an emphasis on bioinformatics and a postdoctoral fellow at the Leprosy Laboratory of the IOC, Paulo Thiago Souza Santos, explains that the technology is based on the computer's ability to distinguish subtle variations in images.
“Artificial intelligence can see more than the human eye. For the computer, each point of the image is a bit, translated into a number. An untrained person may not notice the difference between two colors that are very close together, but when the computer turns these colors into numbers, it 'sees' a clear difference. It is based on this that we can train the machine to try to make a differential diagnosis”, says the researcher.
However, the scientists had to adapt the methodology to face one of the great challenges of leprosy: the diversity of forms of the disease. Considering only the skin manifestations, the infection can manifest itself with one or many lesions, small or large, flat or raised, whitish or reddish, concentrated in a region or disseminated throughout the body.
“Melanoma is a single, dark lesion. The differential diagnosis is made with few dermatological diseases of similar presentation. Leprosy has many faces, many forms of manifestation. The number of similar diseases is large. Since the beginning of the research, we thought that this would be a challenge and we looked for alternatives to increase the accuracy of the algorithm”, points out the dermatologist and researcher at Ambulatório Souza Araújo do IOC, Raquel Barbieri, who shares first authorship of the article with Microsoft's Senior Applied Research Scientist, Yixi Xu.
One of the assets of the project was having a large bank of images of lesions to train the system to differentiate leprosy from other skin diseases. In all, 1.229 photographs were taken of 585 lesions, including as many confirmed cases of leprosy as diseases with similar presentations. The images were obtained with the collaboration of 222 patients seen at the Souza Araújo Outpatient Clinic, a specialized service in the diagnosis, treatment and prevention of leprosy, maintained by the Leprosy Laboratory of the IOC.
The scientists also developed an artificial intelligence model capable of combining image recognition with the analysis of clinical data from patients. The tests carried out pointed out ten main characteristics to establish the probability of the disease. For example, loss of thermal sensitivity in the lesion and changes in sensitivity in the feet were associated with a high probability of leprosy, while itching, which is more present in other dermatological diseases, was associated with a lower chance of infection.
Analyzing the images separately, the system reached 70% of correct answers. Combining this analysis with the processing of clinical data, the index exceeded 90%. “Our research shows that with a combination of medical imaging and medical history, artificial intelligence modeling can provide predictive power to help diagnose leprosy,” says Yixi Xu.
In the next phase of the research, the scientists must train the algorithm by collecting images and data through a cell phone application, improving the system to operate with lower resolution images and in situations similar to the day-to-day of health services.
“Currently, in Brazil, the diagnosis of leprosy is made in basic health units, by general practitioners and family doctors, who are not specialists in dermatology. Our goal is to produce an intuitive application, in which professionals can load images of lesions and select some important data to estimate the probability of leprosy. This technology can provide more clarity for patient referrals”, comments Milton Moraes, who estimates that a beta version of the application could be available in two years.
The technology validation stage should further expand the geographic area of the research, covering urban and rural regions in Asia and Africa, in order to evaluate the system in different contexts.
Besides the IOC, from Microsoft AI for Health and the Novartis Foundation, the Universities of Basel, Switzerland, and Aberdeeen, Scotland, participated in the study. The research was funded by the Novartis Foundation in partnership with Microsoft. The study's image and database databases are available for use by other scientists in open access.
National reference center with the Ministry of Health, the Leprosy Laboratory of the IOC is at the forefront of pioneering research with the aim of contributing to the Unified Health System (SUS). Last year, after 15 years of studies, the scientists obtained registration from the National Health Surveillance Agency (Anvisa) for the first molecular diagnostic test for leprosy developed in Brazil. Called NAT Leprosy Kit, the product was developed by IOC in partnership with the Carlos Chagas Institute (Fiocruz-PR) and the Institute of Molecular Biology of Paraná (IBMP), linked to Fiocruz and the government of Paraná.
In November, the exam received a favorable opinion from the National Commission for the Incorporation of Technologies in the SUS (Conitec) and should form part of the new Clinical Protocol and Therapeutic Guidelines for Leprosy, being prepared by the Ministry of Health.
“Our goal is to expand the diagnosis of leprosy so that people no longer suffer from a disease that has treatment and can be cured. With the application, we seek to accelerate the suspicion of the disease, which allows referral to specialized services, where dermatologists, if necessary, can request the PCR, to arrive at the final diagnosis”, highlights Milton Moraes.
A technological innovation can contribute to the search for the elimination of leprosy, one of the oldest diseases of humanity. An international team of scientists, led by the Oswaldo Cruz Institute (IOC/Fiocruz) in partnership with Microsoft AI for Health and the Novartis Foundation, has developed a diagnostic assistant, based on artificial intelligence, that can help identify suspected cases of the infection. The technology was called AI4Leprosy.
An article published in the scientific journal 'The Lancet Regional Health - Americas' shows that, from photos of the lesions on the skin of patients and clinical data observed by physicians, the diagnostic assistant indicates the probability of the disease, getting more than 90% correct. of the cases in the tests carried out. According to the scientists, the publication is a proof of concept for the method, which should serve as a basis for creating a mobile application for use by health professionals.
“Our study proves that it is possible to reach a suspected leprosy diagnosis with an artificial intelligence algorithm. This tool can support the doctor's decision to start the investigation, accelerating the confirmation of the diagnosis and the beginning of the treatment, which is essential to interrupt the transmission of the disease and prevent sequelae", says the head of the Leprosy Laboratory of the IOC, Milton Ozório Moraes.
“By taking advantage of artificial intelligence and machine learning, patients can find the right specialists at the right time, potentially improving the lives of millions of people around the world. The possibilities with this type of application are endless, especially for diseases. multifaceted as leprosy,” says Microsoft Chief Data Scientist Juan Lavista Ferres.
"These results are an exciting indication of the potential of AI4Leprosy. We are convinced that future validation and global launch can help cover the last few miles of leprosy elimination, using the newest technology available to end one of the oldest known scourges by man," says Novartis Foundation President Ann Aerts.
Delay in diagnosis is one of the biggest challenges for leprosy elimination. Since it was introduced in the 1980s, multidrug therapy – a treatment based on a combination of antibiotics – has cured about 18 million people, reducing the prevalence of the disease by 95%. The drugs are donated by Novartis through the World Health Organization (WHO) to be offered free of charge to patients worldwide. In addition to promoting healing, the treatment blocks the transmission of Mycobacterium leprae, which causes the infection.
However, because of the delay in identifying the condition, the bacteria continues to spread and many people still develop visible deformities, loss of movement of the feet or hands and vision problems due to leprosy. In 2019, more than 200 thousand new cases of the disease were registered worldwide, approximately 10 thousand with advanced lesions. In Brazil, the second most affected country on the planet, 27 new cases were detected, including 2,3 with advanced damage.
With the impact of the Covid-19 pandemic, which disrupted health services, leprosy diagnoses dropped, raising the concern of specialists. According to the WHO, the global drop in the detection of new cases was 37% in 2020. In Brazil, a survey by the Brazilian Society of Dermatology (SBD) pointed to a 35% reduction in records in 2020 and 45% in 2021, compared to 2019.
Considering the first results observed, scientists believe that artificial intelligence can contribute to achieving the goals set by the World Health Organization (WHO), such as reducing new cases of the infection by 70% by 2030 and, in the long term, interrupting transmission. of the grievance.
The starting point for the development of the virtual assistant for diagnosing leprosy was a type of image recognition algorithm that has been applied, for example, to support the diagnosis of melanoma, a form of skin cancer. One of the authors of the study, a doctor in oncology with an emphasis on bioinformatics and a postdoctoral fellow at the Leprosy Laboratory of the IOC, Paulo Thiago Souza Santos, explains that the technology is based on the computer's ability to distinguish subtle variations in images.
“Artificial intelligence can see more than the human eye. For the computer, each point of the image is a bit, translated into a number. An untrained person may not notice the difference between two colors that are very close together, but when the computer turns these colors into numbers, it 'sees' a clear difference. It is based on this that we can train the machine to try to make a differential diagnosis”, says the researcher.
However, the scientists had to adapt the methodology to face one of the great challenges of leprosy: the diversity of forms of the disease. Considering only the skin manifestations, the infection can manifest itself with one or many lesions, small or large, flat or raised, whitish or reddish, concentrated in a region or disseminated throughout the body.
“Melanoma is a single, dark lesion. The differential diagnosis is made with few dermatological diseases of similar presentation. Leprosy has many faces, many forms of manifestation. The number of similar diseases is large. Since the beginning of the research, we thought that this would be a challenge and we looked for alternatives to increase the accuracy of the algorithm”, points out the dermatologist and researcher at Ambulatório Souza Araújo do IOC, Raquel Barbieri, who shares first authorship of the article with Microsoft's Senior Applied Research Scientist, Yixi Xu.
One of the assets of the project was having a large bank of images of lesions to train the system to differentiate leprosy from other skin diseases. In all, 1.229 photographs were taken of 585 lesions, including as many confirmed cases of leprosy as diseases with similar presentations. The images were obtained with the collaboration of 222 patients seen at the Souza Araújo Outpatient Clinic, a specialized service in the diagnosis, treatment and prevention of leprosy, maintained by the Leprosy Laboratory of the IOC.
The scientists also developed an artificial intelligence model capable of combining image recognition with the analysis of clinical data from patients. The tests carried out pointed out ten main characteristics to establish the probability of the disease. For example, loss of thermal sensitivity in the lesion and changes in sensitivity in the feet were associated with a high probability of leprosy, while itching, which is more present in other dermatological diseases, was associated with a lower chance of infection.
Analyzing the images separately, the system reached 70% of correct answers. Combining this analysis with the processing of clinical data, the index exceeded 90%. “Our research shows that with a combination of medical imaging and medical history, artificial intelligence modeling can provide predictive power to help diagnose leprosy,” says Yixi Xu.
In the next phase of the research, the scientists must train the algorithm by collecting images and data through a cell phone application, improving the system to operate with lower resolution images and in situations similar to the day-to-day of health services.
“Currently, in Brazil, the diagnosis of leprosy is made in basic health units, by general practitioners and family doctors, who are not specialists in dermatology. Our goal is to produce an intuitive application, in which professionals can load images of lesions and select some important data to estimate the probability of leprosy. This technology can provide more clarity for patient referrals”, comments Milton Moraes, who estimates that a beta version of the application could be available in two years.
The technology validation stage should further expand the geographic area of the research, covering urban and rural regions in Asia and Africa, in order to evaluate the system in different contexts.
Besides the IOC, from Microsoft AI for Health and the Novartis Foundation, the Universities of Basel, Switzerland, and Aberdeeen, Scotland, participated in the study. The research was funded by the Novartis Foundation in partnership with Microsoft. The study's image and database databases are available for use by other scientists in open access.
National reference center with the Ministry of Health, the Leprosy Laboratory of the IOC is at the forefront of pioneering research with the aim of contributing to the Unified Health System (SUS). Last year, after 15 years of studies, the scientists obtained registration from the National Health Surveillance Agency (Anvisa) for the first molecular diagnostic test for leprosy developed in Brazil. Called NAT Leprosy Kit, the product was developed by IOC in partnership with the Carlos Chagas Institute (Fiocruz-PR) and the Institute of Molecular Biology of Paraná (IBMP), linked to Fiocruz and the government of Paraná.
In November, the exam received a favorable opinion from the National Commission for the Incorporation of Technologies in the SUS (Conitec) and should form part of the new Clinical Protocol and Therapeutic Guidelines for Leprosy, being prepared by the Ministry of Health.
“Our goal is to expand the diagnosis of leprosy so that people no longer suffer from a disease that has treatment and can be cured. With the application, we seek to accelerate the suspicion of the disease, which allows referral to specialized services, where dermatologists, if necessary, can request the PCR, to arrive at the final diagnosis”, highlights Milton Moraes.
The non-profit reproduction of the text is allowed as long as the source is cited (Comunicação / Instituto Oswaldo Cruz)
 |
 |
