Tuberculosis Prevention & Artificial Intelligence!

Koushik Das

February 29

Tuberculosis & its pitfalls
Tuberculosis (TB), an infectious and communicable disease, is looming large over decades in the most populous country of India, having the highest burden of two deaths occurring every three minutes. The disease is caused by a bacterium affecting lungs, and spreads easily from an infected person to the other through droplets from the throat when one coughs, sneezes or spits. The burden is no less worldwide as an estimated 10.6 million people, including six million men, 3.4 million women and 1.2 million children, suffered from TB in 2021. Although the impact of this disease is felt mostly in low- and middle-income countries, with over 80% cases and deaths. In 2021, around 87% of new TB cases were recorded in 30 high risk countries, as more than two-thirds of the global total were registered in Bangladesh, China, the Democratic Republic of Congo, India, Indonesia, Nigeria, Pakistan and the Philippines.

Certain medical conditions increase the chances of contracting TB especially if a person suffers from diabetes, weak immune system and malnourishment, and those with heavy use of tobacco. Coughing with blood, chest pain, fatigue, weakness, weight loss, fever and night sweats are some common signs of the disease. TB can also occur in kidneys, brain, spine and skin. A deadly combination of TB with HIV (Human Immunodeficiency Virus) must have a mention here as many people living with HIV are at higher risk of death than those without HIV.

Traditional methods of TB prevention & cure
TB is preventable and curable, though sometimes not diagnosed where symptoms do not show up known as latent TB and then there is primary infection and active TB when the symptoms show up and make a person seriously ill. Having the traditional lab tests of sputum and lung fluid, chest X-ray and CT scans do not help reduce the number of cases and the deaths till date, giving rise to the understanding that the disease has been lurking for long. There are rapid tests, like Xpert MTB/RIF Ultra, for early detection of TB. However, diagnosing multi-drug resistant and HIV associated TB is a complex, as well as expensive, one and difficult to diagnose in children. There are various drugs, like Isoniazid, Rifampin, Ethambutol, Pyrazinamide and Rifapentine. However, they have various side effects, such as skin rashes, nausea, stomach upset, yellow skin or eyes, itchy skin and dark urine. Some of the usual preventable measures are washing hands thoroughly and often, sneezing into one's own elbow or covering the mouth while coughing, avoiding close contact with others, taking medication correctly and staying at home until s/he is cleared by a physician. Also, there is a vaccine, called Bacillus Calmette-Guerin, available in some countries. This vaccine is mostly given to children in order to prevent meningitis and a serious type of TB, called Military Tuberculosis. Unfortunately, the vaccine may make skin tests for TB less accurate.

Investments on innovations in TB prevention
The main source of funding research and development on TB cure and prevention is the Global Fund to Fight AIDS, Tuberculosis and Malaria. The US is the largest contributor. The World Health Organisation's (WHO) End TB strategy has mapped out the global response required between 2016 and 2035 for reducing the spread of TB by 90% and TB mortality by 95%. Electronic and mobile health apps, too, are becoming common in the healthcare industry in the Developed and Developing Nations, and this digital technology needs to be harnessed into TB care and prevention. TB care can be supported through digital technologies in four ways that of patient care, surveillance, programme management and e-learning. Digital technology methods of TB care and prevention can be augmented if it is rolled out extensively throughout the country and for this, some innovations are required by integrating Artificial Intelligence (AI) that is automation of human cognitive functions and machine learning. AI has already invaded our lives through various applications and in various domains, making our lives easier and better. The use of AI can accelerate the growth of existing digital procedures, e.g., treatment via video that involves the provider observing TB patients taking medication via video sent over a smartphone or tablet. Physicians require inputs from health workers before interpreting the signals sent by the patients. However, the process involves a privacy concern, although data encryption can reduce the risk. AI can also step in by adding value to the available tools set to the medical practitioners. Video software VDOT is one such tool, having unique features of a particular person taking a specific medication. It can be identified when combined with the timestamp of the event and the mobile phone number. A distinct signature can also be generated and transmitted as a small data packet to the caregiver. Like patient care, AI can add value in reviewing disaggregated patient records and as a means of detecting TB early. For instance, it can help in tracing people with the possibility of contracting the disease or its relapse along with adverse drug reactions, and also monitoring the mortality and its linkage with HIV.

AI can be useful in health programme management and e-learning courses. In case of management of TB programmes, connected diagnostics is an important concept that helps to monitor diagnostic machines remotely and to consolidate the results. With the help of AI, the next generation products can interpret results and match them with those from other diagnostic processes. Artificial neural networks have been applied to clinical situations that still create a problem for practitioners, such as microscopy sputum smear-negative TB and pleural TB in settings where diagnostic options are limited.

E-learning module preparation requires a multidisciplinary set of skills to ensure that the content is relevant, correct and up to date, and also attractive in format as well. Here, AI can automate the laborious classification of texts and images needed to create the content of an app. Hence, speeding up the process and lowering the cost are crucial. It can also increase the value of online courses by making the grading of student assessments automated.

AI: The way ahead
With the advent of smartphones, healthcare facilities and its services came to the fingertips of the common people in the past decade. Now, the generation Z is all set to witness the medical innovations with AI-enhanced software that can monitor patients and issue alerts, thus strengthening home-based care. AI is still at its nascent stage where the technology can be considered as a supportive tool in human decision-making. Also, it is better for humans to remain the watchdogs and control the AI for the very survival of the human race. Although AI can take a large share in medical decision-making, it is important to not turn it into a God.

References:
1) https://www.who.int/india/health-topics/tuberculosis
2) https://www.who.int/news-room/fact-sheets/detail/tuberculosis#:~:text=Around%2087%25%20of%20new%20TB,%2C%20Pakistan%2C%20and%20the%20Philippines
3) https://www.who.int/news-room/fact-sheets/detail/tuberculosis#:~:text=Around%2087%25%20of%20new%20TB,%2C%20Pakistan%2C%20and%20the%20Philippines
4) https://my.clevelandclinic.org/health/diseases/11301-tuberculosis
5) https://my.clevelandclinic.org/health/diseases/11301-tuberculosis
6) https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5478795/
7) https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5478795/
8) https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5478795/
9) https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5478795/