<aside> 📌 Hey, I’m Pavi! Growing up, my second home was a doctor’s office (I’m in perfect condition, I just always got the flu!). But, going from doctor’s offices to hospitals, I developed a sense of gratitude for going to these facilities and magically feeling better within weeks; until I realized not everyone has access to them. And so, for the past two years I’ve been working on increasing healthcare accessibility through technology.

I’ve explored everything from nanotechnology to machine learning and realized that this is just the beginning.

Aside from reading research papers and exploring random math topics, I’ve explored many different domains over the past few years. From interning at the world’s largest retailers innovation Incubator, Walmart Blue Labs, to speaking on some of the world’s largest stages, WebSummit, to building out technically intensive projects, like QuantumTags, I’ve discovered a lot about myself.

The main thing I’ve discovered, is my love for impact, and my hate for world problems. These two feelings led me to spending the past 10 months on Parkinson’s Disease.

</aside>

Diagnosis and Prognosis of Parkinson’s Disease Through Your Smartphone

https://www.loom.com/share/8a99e00d2e4d4772b1cc4062340f182a

Tangible Results:

  1. Previously: Built a machine learning algorithm for voice based biomarkers: diagnosing Parkinson’s Disease 2-5 years prior to clinical diagnosis with a 10-second voice recording.
  2. Real World Companies:
  3. In Progress: Now building MonitorPD — a progression monitoring app for Parkinson’s Disease through five assessments, running on machine learning and deep learning.

Overview:

Part 1: Voice Biomarkers Diagnosis

Voice Based Biomarkers for Neurodegenerative Diseases

Part 2: Progression Monitoring Front-End

[Figma Mockup](https://www.loom.com/share/baa6fb9dda3448309c04babee37dccf9)

Figma Mockup

<aside> 📌

My ask: There’s a couple things which could really help the progression of this project, if you can help with any one of these, I’d love to talk!

  1. Institute level connections (everyone): know anyone at Parkinson’s Canada? Any healthcare startups or companies who would be interested in them? I would love to get in contact with you and them to find an intersection that works.

  2. Data (health-tech experts): Do you have or know of anyone that might have data on people with Parkinson’s? Whether it be the voice, bradykinesia statistics, gait, etc. I’d love to chat.

</aside>

The Problem: Parkinson’s diagnosis and prognosis sucks.

General: From the diagnosis to the prognosis, Parkinson’s Disease (PD) is simply a mess. There’s no definite diagnosis for PD: normally, patients go through MRI, CT or SPECT scans but these are used to exclude other syndromes, not for a PD diagnosis. On top of that, when patients are diagnosed with PD, they meet with episodically (3-6x a year) and the specialist asks the patient to perform numerous tasks (there’s not only one biomarker to assess) and the patients are given a score, overall a very subjective process.

Diagnosis

Parkinson’s Disease (PD) is a degenerative neurological disorder caused by decreased dopamine levels in the brain. PD leads to the deterioration of movement, tremors, stiffness, changes in mood and changes in speech. In speech particularly, PD patients inherit dysarthria (a disorder where there’s difficulty articulating sounds), hypophonia (lowered volume) and monotone.

The status quo of diagnosing PD is a specialist will take the neurological history of the patient and observe the patient’s motor skills in various situations. However, there is no definitive laboratory test to diagnose PD because there’s multiple biomarkers, which makes the diagnosis even more difficult.

In other words, the key here is the difficulty in finding one biomarker to diagnose the disease because of how similar many other biomarkers of similar diseases are. So, due to PD’s broad sequela, a diagnosis of the disease can be non-trivial — particularly in the early stages — and progression rates are variable across patients.