Risk Classification for Interstitial Cystitis/Bladder Pain Syndrome (IC/BPS) Using Machine Learning Based Predictions - Beyond the Abstract
May 22, 2024
To improve diagnosis of IC/BPS(IC) we hereby developed an improved IC risk classification using machine learning algorithms.
Biographies:
Michael B. Chancellor, MD, Professor of Urology, Oakland University William Beaumont School of Medicine, Rochester, MI
Biographies:
Michael B. Chancellor, MD, Professor of Urology, Oakland University William Beaumont School of Medicine, Rochester, MI
Read the Full Video Transcript
Michael Chancellor: Hi, I'm Dr. Michael Chancellor, Professor of Urology at the Oakland University William Beaumont School of Medicine.
Today I'd like to share with you, beyond the abstract on UroToday, a recently published paper in the Gold Journal, Urology on a Urine Protein-based Biomarker Diagnostic Tests for Interstitial Cystitis. I just have three quick takeaway points on what inspired me, and hopefully, it can help you with your research.
The first is the handling of urine samples. We have to freeze it, and it's hard to ship and we collaborate. So I found this urine preservative online that can preserve most urine proteins for up to several months. This allows us to send and collect urine from across the country.
The second is crowdsourcing. It's so hard, as we all know, to get enough clinical samples. And how many patients are there with interstitial cystitis that wants to participate in a trial at any city?
I was so frustrated getting enough sample for a biomarker study, so I started searching what to do about it on my iPhone. Then I found out crowdsourcing, like Apple Health, is beginning to be used more often for people from across the country and the world to collaborate.
So we reached out to the Interstitial Cystitis Association to initiate a crowdsourced research effort where patients with IC from across the country can participate by the social media network. We used Facebook, Instagram, YouTube, and yes, yeah, even TikTok. And it went very well. We collected over 1,000 samples in about three months. What the patient did is they collect their urine in this cup. Then they put it in this biologic shipping bag and send it to us via FedEx.
Then that leads us to the third point. Now, we collected over these thousand samples and we start analyzing protein in them. And we use the machine learning algorithm to develop a classifier: a score that improves upon what the patient's symptoms are to help the physician diagnose interstitial cystitis.
I'm not a computer expert. I don't have expertise in big data analytics or artificial intelligence machine learning. But with a good idea and capable collaborators, we were able to succeed together.
So I hope you find this paper and this little short video helpful in your research journey. I'd like to thank you for your time, and especially I'd like to thank my collaborators for working with me on this project together. Thank you.
Michael Chancellor: Hi, I'm Dr. Michael Chancellor, Professor of Urology at the Oakland University William Beaumont School of Medicine.
Today I'd like to share with you, beyond the abstract on UroToday, a recently published paper in the Gold Journal, Urology on a Urine Protein-based Biomarker Diagnostic Tests for Interstitial Cystitis. I just have three quick takeaway points on what inspired me, and hopefully, it can help you with your research.
The first is the handling of urine samples. We have to freeze it, and it's hard to ship and we collaborate. So I found this urine preservative online that can preserve most urine proteins for up to several months. This allows us to send and collect urine from across the country.
The second is crowdsourcing. It's so hard, as we all know, to get enough clinical samples. And how many patients are there with interstitial cystitis that wants to participate in a trial at any city?
I was so frustrated getting enough sample for a biomarker study, so I started searching what to do about it on my iPhone. Then I found out crowdsourcing, like Apple Health, is beginning to be used more often for people from across the country and the world to collaborate.
So we reached out to the Interstitial Cystitis Association to initiate a crowdsourced research effort where patients with IC from across the country can participate by the social media network. We used Facebook, Instagram, YouTube, and yes, yeah, even TikTok. And it went very well. We collected over 1,000 samples in about three months. What the patient did is they collect their urine in this cup. Then they put it in this biologic shipping bag and send it to us via FedEx.
Then that leads us to the third point. Now, we collected over these thousand samples and we start analyzing protein in them. And we use the machine learning algorithm to develop a classifier: a score that improves upon what the patient's symptoms are to help the physician diagnose interstitial cystitis.
I'm not a computer expert. I don't have expertise in big data analytics or artificial intelligence machine learning. But with a good idea and capable collaborators, we were able to succeed together.
So I hope you find this paper and this little short video helpful in your research journey. I'd like to thank you for your time, and especially I'd like to thank my collaborators for working with me on this project together. Thank you.