{"id":24115,"date":"2023-06-06T07:11:47","date_gmt":"2023-06-06T12:11:47","guid":{"rendered":"https:\/\/cdtrp.ca\/?p=24115"},"modified":"2023-06-13T11:48:56","modified_gmt":"2023-06-13T16:48:56","slug":"spotlight-paper-michael-chasse","status":"publish","type":"post","link":"https:\/\/cdtrp.ca\/en\/spotlight-paper-michael-chasse\/","title":{"rendered":"Spotlight paper in Nature Scientific Reports: Dr. Micha\u00ebl Chass\u00e9"},"content":{"rendered":"

<\/a><\/span><\/div>
<\/div>

Spotlight paper in Nature Scientific Reports: Dr. Micha\u00ebl Chass\u00e9<\/strong><\/h4>\n

The CDTRP would like to extend a huge congratulations to Dr. Micha\u00ebl Chass\u00e9<\/strong> and his team for their recent publication in Nature Scientific Reports<\/strong>: on automated screening of potential organ donors using a temporal machine learning model. This groundbreaking research has the potential to make a significant impact in the field of organ donation and transplantation. We are proud to have Dr. Micha\u00ebl Chass\u00e9 as a member of our community and are grateful for their contributions to the field. Keep up the great work!<\/p>\n

We have asked Dr. Chass\u00e9 a series of questions about the article, that you can read below.<\/p>\n<\/div>

<\/div><\/div>
How does the machine learning model compare to traditional methods of organ donor screening?<\/strong><\/h6>\n
\n

The machine learning model demonstrated promising performance, potentially surpassing that of a simpler logistic regression model. Although not flawless, the neural network achieved high accuracy in identifying potential organ donors using routinely collected medical data. Notably, the model’s performance exhibited robustness across donor subgroups and maintained stability in a prospective simulation.<\/p>\n<\/blockquote>\n

How can healthcare providers effectively incorporate the machine learning model into their organ procurement practices?<\/strong><\/h6>\n
\n

This study suggests the potential for integrating such models into existing electronic health record systems. In practice, the model could operate in the background, flagging potential organ donors as they meet the defined criteria. Consequently, this may empower healthcare providers to initiate organ donation discussions and processes earlier. However, the acceptability, ethical implications, and feasibility of such practices necessitate further investigation.<\/p>\n<\/blockquote>\n

What implications does the use of this machine learning model have for organ transplantation rates and patient outcomes?<\/strong><\/h6>\n
\n

By enhancing the identification of potential organ donors, the model could potentially augment the availability of organs for transplantation, independent of organ management practices. Consequently, this might reduce wait times for organ transplants and potentially improve patient outcomes. Nonetheless, it’s essential to emphasize that these are potential outcomes and warrant further study.<\/p>\n<\/blockquote>\n

What are the next steps and how could the CDTRP support the future directions of this work?<\/strong><\/h6>\n
\n

Should the Canadian Donation and Transplantation Research Program (CDTRP) find this approach worthwhile, they could potentially support future research endeavors to validate similar models in diverse hospital or healthcare settings. They might also back efforts to refine the model for improved accuracy, or to develop implementation strategies for integrating the model into healthcare systems. Importantly, at this juncture, these kinds of models are not ready for clinical implementation. While this paper provides preliminary feasibility evidence, further prospective multicenter validation is required.<\/p>\n<\/blockquote>\n<\/div><\/div><\/div><\/div><\/div>

Abstract<\/strong><\/span><\/h5>\n

Organ donation is not meeting demand, and yet 30\u201360% of potential donors are potentially not identified. Current systems rely on manual identification and referral to an Organ Donation Organization (ODO). We hypothesized that developing an automated screening system based on machine learning could reduce the proportion of missed potentially eligible organ donors. Using routine clinical data and laboratory time-series, we retrospectively developed and tested a neural network model to automatically identify potential organ donors. We first trained a convolutive autoencoder that learned from the longitudinal changes of over 100 types of laboratory results. We then added a deep neural network classifier. This model was compared to a simpler logistic regression model. We observed an AUROC of 0.966 (CI 0.949\u20130.981) for the neural network and 0.940 (0.908\u20130.969) for the logistic regression model. At a prespecified cutoff, sensitivity and specificity were similar between both models at 84% and 93%. Accuracy of the neural network model was robust across donor subgroups and remained stable in a prospective simulation, while the logistic regression model performance declined when applied to rarer subgroups and in the prospective simulation. Our findings support using machine learning models to help with the identification of potential organ donors using routinely collected clinical and laboratory data.<\/span><\/p>\n

Read the full paper here.<\/strong><\/a><\/p>\n<\/div><\/div><\/div><\/div><\/div>

\n
(Re)Watch CDTRP Research Connect<\/strong><\/h5>\n

On April 4, we were pleased to have\u00a0Dr. Micha\u00ebl Chass\u00e9<\/strong>, and Dr.<\/strong>\u00a0Nicolas Sauthier<\/strong>, Anesthesiology Resident and Master\u2019s student in biomedical sciences at Universit\u00e9 de Montr\u00e9al, to present on the topic: \u201cThe challenge of missed organ donors: Can machine learning be used for early identification of potential donors?\u201d as part of the CDTRP Theme 1 \u2013 Improve a Culture of Donation.<\/p>\n<\/div>\n<\/div>

<\/div>