Three Fingers, One Wrist, and a Question Engineers Are Still Trying to Answer....
For ~800 years, Ayurvedic physicians have diagnosed patients by pressing 3 fingers, index, middle, ring onto the radial artery at the wrist, each finger meant to read a different one of the body's three doshas: vata under the index finger, pitta under the middle, kapha under the ring.
It is called Nadi Parikshan, and it traces back to the 13th-century physician Śārṅgadhara, the same scholar who first formally introduced pulse diagnosis into the Ayurvedic tradition.
For all 800 of those years, the entire diagnosis has depended on 1 thing: the trained sensitivity of a single physician's fingertips.
No instrument. No number. No second opinion except another physician's fingers.
A group of Indian scientists decided to actually try to fix that, not by dismissing the tradition, but by asking a genuinely engineering question: can a machine feel what a Vaidya's fingers feel, and turn it into a number?
Starting in the mid-2000s, an engineer named Aniruddha Joshi, trained at IIT Bombay, built a device called Nadi Tarangini, working alongside co-authors Anand Kulkarni, Sharat Chandran, V. K. Jayaraman, and B. D. Kulkarni, a senior scientist at Pune's National Chemical Laboratory who mentored and supported parts of the effort.
The device was instrumented with a strain-gauge sensor (commercial and some later research versions incorporate piezoelectric elements) pressed against the wrist, feeding into an amplifier and a 16 bit digitizer, capturing the radial pulse not as a subjective "feeling" but as an actual continuous waveform, logged as raw data over time.
Where a physician's fingers give an impression, the machine gives a graph, precise, repeatable, and comparable across patients, across time, across doctors.
Their published results, in a 2007 IEEE Engineering in Medicine and Biology Society paper, real, peer-reviewed academic ground showed the device could capture genuinely richer, more detailed waveform data than earlier attempts at pulse-sensing hardware.
Joshi went on to commercialize the tech through a company called Atreya Innovations, based in Pune, which continued refining the hardware.
Follow-up research groups have kept building on the underlying idea too, a 2024 paper, published through the ACM's Computational Biology and Bioinformatics conference, described a newer, independent device using a piezoelectric sensor and a modern STM32 microcontroller, specifically built to test whether Ayurvedic pulse principles could be captured with cleaner, more reliable signal acquisition.
Worth noting is the commercial device has since moved well beyond the prototype stage.
It received approval from India's drug regulator, the CDSCO, making it by that account, the first Ayurvedic pulse-diagnostic device of its kind to clear that specific regulatory bar, and it is reportedly now in use across a number of clinics.
The commercial version of the device is marketed today with claims of 85%+ accuracy, reportedly based on matching the device's readings against experienced Vaidyas' own assessments across a large dataset, and including data submitted as part of the CDSCO regulatory process.
That is a genuinely more substantial evidence base than pure marketing copy.
What's genuinely remarkable here is not a triumphant "science proved Ayurveda right" headline, it is that a diagnostic method built entirely on trained human touch, refined over eight centuries with zero instrumentation, was specific and consistent enough that engineers a century after the invention of the ECG machine looked at it and thought: that is not folklore, that is a signal worth trying to capture properly.
The jury on what the machine actually proves is still out. But the fact that serious biomedical engineers are still building new hardware to chase this exact question, 20 years and counting, tells you something about how seriously the underlying observation is being taken, not as legend, but as an open, unresolved measurement problem.