A Novel Breakthrough In Wrist-Worn Transdermal Troponin-I-Sensor Assessment For Acute Myocardial Infarction

On March 4-6, 2023, the Annual Conference of the American College of Cardiology (ACC) was held in New Orleans. One of the major guidelines and latest studies was released, including a study of a Wrist-Worn Transdermal Troponin-I-Sensor using artificial intelligence technology to rapidly diagnose acute myocardial infarction without blood samples. Professor Partho P. Sengupta, author of the study from West Virginia University in the United States, is invited to provide details.

International Circulation: Would you please talk about the principle of the Wrist-Worn Transdermal Troponin-I-Sensor in the evaluation of AMI？

Professor Partho P. Sengupta: Thank you very much for the chance to explain a little bit more about the technology in which we use an infrared sensor that is embedded in a wrist-worn device. From that, it is possible for us to measure different molecules, and in this instance, troponin, from the skin, and not requiring blood. In this particular study that was performed at five sites in India, we investigated whether this technology can allow detection of the blood biomarker, troponin, using the infrared sensor. That can be utilized for arriving at a diagnosis of myocardial infarction. The technology was developed using an AI algorithm. From the spectrophotometric sensor recordings, we used artificial intelligence technology to first train the data at three sites, and then after the training was done and the machine learning model was built, it was externally validated in two sites, and we were able to develop a method for arriving at a diagnosis of elevated troponin without requiring blood samples, and coming to the diagnosis of myocardial infarction.

International Circulation: What is the accuracy and effect of Wrist-Worn Transdermal Troponin-I-Sensor in assessing AMI compared to conventional test?

Professor Partho P. Sengupta: With regards to the accuracy of the results, there were two different methods we used. The first thing to do was a head-to-head comparison between the troponin infrared sensor spectrophotometric recording and the machine learning model for predicting the presence of elevated high-sensitivity troponin levels. The second was to look at the relationship with the coronary angiograms and the wall motion abnormalities that were observed on invasive catheterization and on cardiac ultrasound techniques. With regards to the ability to detect elevated high-sensitivity troponin values, which was measured at a central laboratory, the technology had an accuracy of 0.90 in the validation cohort, and an accuracy of 0.92 in the external validation cohort. So the precision and the accuracy were quite high. With regard to the ability to detect the presence of obstructive coronary artery disease, the presence of elevated troponin-I as indicated by the machine learning model was associated with increased odds, which was significantly important and statistically significant for the prediction of obstructive coronary artery disease. It was also associated with a statistical prediction of the presence of wall motion abnormalities. Obstructive coronary artery disease and wall motion abnormalities as endpoints are important phenotypic presentations that we were able to validate with this technology.

International Circulation: Currently, what needs to be improved in the application of Wrist-Worn Transdermal Troponin-I-Sensor?

Professor Partho P. Sengupta: With regards to further work that has to be done with the wrist-worn infrared sensor, first of all, this is just the beginning. We have presented a feasibility study showing that it is possible to diagnose the presence of elevated troponin levels. There are several levels of work that need to be done. Number one, we need to make the wrist-worn sensor more high-fidelity with regards to the noise that comes from motion artifacts, or because of the sensor coming into contact with the skin. There is already work in progress whereby the machine learning tool can tell once the wrist-worn test has been recorded, whether the recording has been accurately registered or not. It is very sensitive to motion artifacts, and also the ability to touch the skin. Those principle changes in the wrist-worn device are going to evaluated in the near future. Secondly, we need to test more across different ethnicities, different skin colors and different circumstances related to melanin content, and see how generalizable this particular technology is. Number three, we need to test this in real life scenarios in the emergency room. The current project evaluated patients who were diagnosed with myocardial infarction. We need to take this emergency room or ambulatory chest pain context and find out how best it performs in real world settings in emergency rooms. Finally, the most important is, right now, the intended purpose is for the presence or absence of elevated troponin levels. We need to develop a test or machine learning model that can determine the exact troponin level value. We need to develop a regression algorithm using a higher number of sample sizes.

International Circulation: What is your opinion about the future of Wrist-Worn Transdermal Troponin-I-Sensor?

Professor Partho P. Sengupta: The future of the wrist-worn troponin sensor creates an opportunity to address the concerns we currently have with overcrowding in emergency rooms. If you look around the world, chest pain is a very common symptom that patients present with. Perhaps there will be opportunity for applying the troponin test using the wrist-based sensor technique in order to streamline the flow of patients through emergency rooms. Number two, taking a blood sample and waiting for the results for an hour is sometimes cumbersome, and in high traffic situations, there may be shortages of people available to take blood samples. We may be able to streamline the number of patients who would actually require blood sampling - at least reduce the number - and those for people who need an earlier confirmation of test results, they could be accelerated towards this infrared wrist-worn technology. It is also possible that we could utilize this technique in the field, in ambulances and emergency services where blood sampling may not be available. Maybe in the near future, we could implement this in community settings. However, this is more challenging, and would require more testing, validation and maturation of the technology.