CogniDiagnose, an innovative AI-driven platform, is revolutionizing diagnostic solutions with cutting-edge technology. It employs EEG headset-based diagnostics to monitor brain health through 21 channels and other critical biomarkers. Moreover, its AI-enabled software streamlines the process by automatically generating comprehensive reports, enabling scalable diagnostic services with precision and efficiency. This groundbreaking approach earned CogniDiagnose the prestigious Best Healthtech Startup Award at NBHC 2024, hosted by GLF Business School.
In an exclusive conversation with The Interview World on the sidelines of NBHC 2024, Devyansh Gupta, CEO and Founder of CogniDiagnose Pvt. Ltd., sheds light on the platform’s transformative impact on brain health diagnostics. He elaborates on the innovative methodologies implemented by CogniDiagnoseAI, emphasizes the biomarkers’ role in detecting neurodegenerative diseases, and shares insights into the platform’s exceptional diagnostic accuracy. Additionally, he outlines the organization’s long-term vision, underscoring its commitment to advancing global brain health.
Here are the key highlights from this engaging and insightful discussion.
Q: What innovative approaches is CogniDiagnoseAI implementing to advance brain health diagnostics?
A: CogniDiagnose AI is transforming brain health diagnostics in India by addressing the critical need for affordable and accessible objective screening tests. To achieve this, we have developed advanced EEG hardware—electroencephalography technology designed for precision and ease of use.
For those unfamiliar, EEG functions for the brain much like ECG does for the heart. Our EEG headset features 21 channels and follows a 30-minute data collection protocol. This technology is specifically designed for individuals over the age of 45 who may experience symptoms such as memory fog, difficulty recalling daily details, or general cognitive decline. These issues often leave people questioning whether they are normal signs of aging or early indicators of conditions like dementia or Alzheimer’s.
Our testing identifies early biomarkers for neurodegenerative diseases, including Alzheimer’s, Parkinson’s, dementia, and cognitive decline. By detecting these conditions at their earliest stages, we empower individuals and healthcare providers to take proactive steps toward better brain health management.
Q: What biomarkers does CogniDiagnoseAI utilize for detecting neurodegenerative diseases?
A: Our system processes data from 21 EEG channels, leveraging advanced frequency-domain analysis to decode brain activity. To explain, think of heart rate: a normal range is around 82 beats per minute. If it rises to 100 or drops to 70, it signals an anomaly. Similarly, we analyze the “brain beat” across these channels, focusing on frequencies between 1 Hz and 50 Hz, where critical brain health information resides.
For each channel, we calculate the power of every frequency band, generating over 1,000 features (21 channels × 50 frequency bands). Our data collection protocol includes two key phases: first, the patient alternates between resting with eyes open and closed in 7-minute intervals. Next, they perform cognitive tasks displayed on a screen. This comprehensive 30-minute protocol yields over 2,000 unique features for analysis.
Processing this data requires addressing inevitable noise. We eliminate electrical interference, such as the 50 Hz line noise common in India, and artifacts caused by muscle movements, like blinking or head turns. Once the data is cleaned, we extract the 2,000 features and input them into our AI-driven machine-learning models. These models then generate a predictive brain health score, classifying the individual as healthy, at risk for cognitive decline, or already experiencing cognitive decline.
Initially, this process involved multiple complex and time-consuming steps, including raw data preprocessing, cleaning, feature extraction, and amplification. These tasks were manually executed across five or six separate workflows, making the process both tedious and inefficient.
We have now streamlined and automated the entire operation. With a single click, our graphical user interface (GUI) processes the recorded data, cleans it, extracts features, and runs it through machine learning algorithms. In minutes, it generates an automated, comprehensive brain health report—turning a once labour-intensive task into an efficient, user-friendly solution.
Q: With many leveraging AI and machine learning for data analysis, what is the accuracy rate of the results derived from your data sets?
A: To ensure accuracy, we trained our models using a dataset of 100 individuals. While this is not a large dataset, it serves as a valuable starting point for a startup like ours to validate the technology during its early stages.
The results so far have been promising, with accuracy rates ranging from 80% to 85%. However, there is significant room for improvement, particularly given the critical precision required in medical science. Achieving higher accuracy is not optional; it is essential.
Currently, this 80% to 85% accuracy represents a strong proof of concept. However, we fully recognize that medical devices demand near-perfect accuracy—between 95% and 99%. To bridge this gap, we are actively collaborating with clinical and research institutions. These partnerships will facilitate pilot runs of our technology, allowing us to refine and validate it further.
Q: Is this product already launched in the market, or are you planning its launch soon?
A: We are currently in the research and development phase, with our minimum viable product (MVP) fully developed and ready. At this stage, we are actively conducting clinical trials to refine and validate our technology.
While we have not officially launched commercially, we welcome individuals interested in brain health diagnostics to connect with us. We can arrange personalized, one-on-one testing on a case-by-case basis. However, our commercial launch is still pending as we focus on ensuring the highest standards through rigorous testing and validation.
Q: What is your long-term vision for the growth of this company?
A: CogniDiagnose AI’s long-term vision is to develop objective brain health screening tests that will revolutionize healthcare diagnostics. Initially, we plan to roll out our product in Delhi and the National Capital Region (NCR). From there, we aim to expand gradually across India, with the ultimate goal of positioning ourselves as a “Make in India” provider serving global markets.
Our ambition is to build and deliver these innovative services from India, catering to the needs of a global audience. In addition to brain health diagnostics, we envision developing a comprehensive suite of healthcare products. These will include blood-based biomarker testing, continuous glucose monitoring, and more.
Furthermore, we see significant opportunities for applying our neurotechnology in industries beyond healthcare. For example, in defense, our solutions can be used to monitor focus and consciousness levels in troops, enabling more informed decision-making. Additionally, there is potential to extend our technology to sectors like marketing, where it can provide enhanced analytics for more effective decision-making.
