REDUCE

Reduce Logo

At the center of the AI-4-AI Lab is the REDUCE project which captures video and audio recordings and satisfaction scores of patients and providers during their clinical visit to gain an inside look into the clinic, improve the quality of their interactions, and reduce provider documentation workload. If you are a healthcare provider and are interested in participating, please reach out to us at observerproject@pennmedicine.upenn.edu.

Timeline

 

Undertaking a thorough qualitative analysis to identify potential computational insight from the clinical video and audio recordings. 

 

Developing methods to detect signs of cognitive impairment in patient-provider interactions using the recorded data.

 

 

Publications/Projects: 

  1. Alasaly B, Jang KJ, Mopidevi S, Johnson KB. Enabling REDCap as a Tool for Multimodal Video-Labeling Research. Poster presented at: REDCapCon 2024; September 2024; St. Petersburg, FL. doi:10.13140/RG.2.2.10338.54723

 

Developing methods to detect and assess the frequency of patient signs and symptoms.

 

 

 

Publications: 

1. Kaur A, Budko A, Liu K, Eaton E, Steitz BD, Johnson KB. Automating Responses to Patient Portal Messages Using Generative AI. Appl Clin Inform. 2025;16(03):718-731. doi:10.1055/a-2565-9155

2. Iannone S, Kaur A, Johnson KB. Artificial Intelligence in Outpatient Primary Care: A Scoping Review on Applications, Challenges, and Future Directions. Health Informatics. Preprint posted online May 13, 2025. doi:10.1101/2025.05.12.25327223

3. Amogh Ananda Rao, Chen PL, Pugh S, Johnson KB. Developing an LLM-based conversational agent for Primary-Care Pre-visit Planning (PCP-Bot). Published online May 2025. doi:10.13140/RG.2.2.20413.17125

4. Johnson KB, Cohen DL, Alasaly B, et al. Observer: Creation of a Novel Multimodal Dataset for Outpatient Care Research. Health Informatics. Preprint posted online May 19, 2025. doi:10.1101/2025.05.18.25327837

5. Mopidevi S, Jang KJ, Alasaly B, et al. MedVidDeID: Protecting privacy in clinical encounter video recordings. J Biomed Inform. Published online August 2025:104901. doi:10.1016/j.jbi.2025.104901

6. Pugh S, Hill M, Hwang S, et al. WATCH-SS: A Trustworthy and Explainable Modular Framework for Detecting Cognitive Impairment from Spontaneous Speech. Neurology. Preprint posted online August 8, 2025. doi:10.1101/2025.08.06.25333047

 

Assessing the accuracy and usefulness of REDUCE algorithms generated in previous years. 

 

Testing REDUCE algorithms in clinical workflows with integration into the EHR as appropriate.